DUKE UNIVERSITY LIBRARY Digitized by the Internet Archive in 2018 with funding from Duke University Libraries r. https://archive.org/details/seventhreportwit01roya '' nu ■r:.-ins —. hinifr:\-s. CAEKLAVEKOCK CASTLE. Frontispiece. Kclinburgh : Published hy His Majesty's Stationery Office. To be purchased through any bookseller or directly from H.M. Stationery Office at the following addresses: —23 Forth Street, Edinburgh; Imperial House, Kingsway, London, M'.C.2, and 28 Abingdon Street, London, S.W. i ; 37 Peter Street, Manchester; 1 St .Andrew's Crescent, Cardiffi; or from E. Ponsonby, Ltd., it6 Grafton Street. Dublin. Price Two Pounds Net. CONTENTS Page. Seventh Report ......... iii List of Ancient and Historical Monuments in the County of Dumfries which THE Commissioners deem most worthy of Preservation . . . . v List of Illustrations ........ viii List of Parishes ......... xii Bibliography ......... xii Introduction to Inventory of Ancient and Historical Monuments and Constructions in the County of Dumfries ..... xvii Inventory .......... i Appendix : Report on the Ruthwell Cross . . . . .219 Glossary ... ...... 287 Index .......... 292 Map of the County of Dumfries, indicating the Position of Monuments, etc., BY Numbers referable to the Inventory . . . . .at end SEVENTH REPORT OF THE ROYAL COMMISSION ON THE ANCIENT AND HISTORICAL MONUMENTS OF SCOTLAND. To THE King’s Most Excellent Majesty. May it please your Majesty,— We, your Majesty’s Commissioners, appointed to make an Inventory of the Ancient and Historical Monuments and Constructions connected with or illustrative of the contemporary culture, civilisation, and conditions of life of the people in Scotland from the earliest times to the year 1707, and to specify those which seem most worthy of preservation, humbly present to your Majesty this our seventh Report. In doing so, we must refer with deepest regret to the death of our esteemed colleague. Lord Guthrie, upon whose counsel we had become accustomed to rely and whose historical knowledge we have found invaluable in the discharge of our duties. We regret also that we have since lost another colleague in the death of Mr Francis C. Buchanan. Appended to the Report is a list of the monuments and constructions of Dumfries¬ shire, which, in the opinion of 3mur Commissioners, seem most worthy of preservation, divided into two classes, viz. {a) those which appear to be specially in need of pro¬ tection, and [h) those worthy of preservation but not in imminent risk of demolition or decay. Your Commissioners have found it desirable to adopt a different format for their Reports and Inventories in order to present the material, and particular!}^ the illustra¬ tions, in a more adequate manner, and the present volume is the first in this new style. On the eve of its issue, in the early summer of 1916, a fire in the printers’ works totally destroyed the whole material, which had to be assembled afresh for publication. Your Commissioners have again to express their thanks to proprietors and others for affording facilities and assistance in the prosecution of their work ; particularly to E. J. Brook, Esq., of Hoddom Castle, Dumfriesshire, and Mr James MTvillop, formerly of the Hoddom Estates Office, Ecclefechan, also to IMr G. W. Shirley of the Ewart Public Library, Dumfries. In the preparation of the Inventor}' they have to thank George Macdonald, Esq., LL.D., C.B., F.B.x\., for assistance in the field of Roman antiquities ; George Neilson, Esq., LL.D., for a contribution and other material ; the Rev. J. King Hewison, D.D., for the use of blocks; the Society of Antiquaries of Scotland for the use of illustrations; and Professor Halliday, Liverpool, for the illustration of the Bruce stone, which stone is in the possession of his family. \Yt. jiKj- n/->i .— n. & co., i.t.i. Gp. 3. 35.'5lin HISTORICAL MONUMENTS (SCOTLAND) C0M:\IISSI0N. The Report on the Riithwell Cross occupies a considerable part of this volume, and is so wide in its scope that your Commissioners think it desirable to preface it with a few words of explanation. This famous monument is an object of quite exceptional interest, attracting much attention not only among British but also among Continental and American scholars. In the three years igi2 to 1914, no fewer than three books, and at least nine articles or pamphlets, appeared on the subject iu England and the United States, and since then these numbers have been materiallv increased. In all these publications arguments regarding the date and provenance of the monument were based on the figure and ornamental sculpture and on the inscriptions in Runic and Latin characters, as well as on the historical and geographical probabilities for or against this or that theor}^ of origin. Such being the case, it has seemed to ^mur Commissioners that, while it is the first part of their dut\" to describe with as much fulness and accuracy as possible the Ruthwell Cross in all its aspects, it is incumbent on them also to supply the available in¬ formation, archaeological, linguistic, and historical, without which no reasoned opinion can be formed as to the date and provenance of this remarkable specimen of mediaeval art. ^^Tth this purpose in view, the necessary references have been made to the similar monument at Bewcastle in Cumberland, of which illustrations have been added for comparison. The Commissioners have further availed themselves of the aid of ^Ir A. Blvth Webster, formerly Lecturer in English in the University of Edinburgh, now Professor of English Literature in the Universit}^ of St Andrews, who has furnished them with an examination of the language and literar}/ content of the poem inscribed on the Ruthwell Cross. In this connection they desire also to acknowledge the services of Mr Ritchie Girvan, Lecturer on the English Language in the University of Glasgow. During the summer of 1915 the archaeological survey of Skye and the Outer Hebrides was carried through, and considerable progress was made with the archi¬ tectural Survey of Midlothian, of which county the prehistoric survey had already been completed. The work of the Commission was suspended in March 1916 for the duration of the M'ar, but since its resumption in 1919 the survey of the monuments of East Lothian has been finished and that of Midlothian is expected to be completed in the current year. Your Commissioners regret that many instances have been brought to their notice of the serious decay of historical buildings owing to neglect. The publica¬ tion of County Inventories, however, having already served to bring some such cases to the attention of proprietors and others, it is hoped that a continuation of the series will not be without further effect in causing more care to be bestowed upon other buildings worthy of preservation. During the W'ar the staff of the Commission was emplo3'^ed in different services relating thereto, both the architects receiving commissions in the Ro^^al Engineers. HERBERT MAXWELL, Chairman. G. BALDWIN BROM'N. THOMAS H. BRYCE. W. T. OLDRIEVE. THOMAS ROSS. ALEXE O. CURLE. W. iAIACKAY MACKENZIE, Secretary. Edinburgh, December 1920. iv LIST OFiiANCIENT AND HISTORICAL MONUMENTS AND CONSTRUCTIONS IN THE COUNTY OF DUMFRIES WHICH THE COMMISSIONERS DEEM MOST WORTHY OF PRESERVATION. I.—MONUMENTS AND CONSTRUCTIONS SPECIALLY IN NEED OF PROTECTION. Parish. Ecclesiastical Structures. Durisdeer Kirkbride Church (No. 155). Caerlaverock Canonbie Dunscore Lochmaben. Moffat Tinwald Torthorwald Durisdeer Eskdalemuir }> Holywood . Hutton and Corrie Tundergarth Wamphray Canonbie Glencairn . Keir . Kirkpatrick-Juxta Castellated and Domestic Structures. Caerlaverock Castle (No. 33(2) ). Hollows Tower (No. 43). Lag Tower (No. 136). Lochmaben Castle (No. 445(2) ). Frenchland Tower (No. 480). Amisfield Tower (No. 578). Torthorwald Castle (No. 590). Fort. Earthwork, Durisdeer (No. 162). Stone Circles. “ Girdle Stanes ” (No. 198). “ Loupin’ Stanes,” near Hartmanor (No. 199). ” Twelve Apostles,” Holywood (No. 284). Whitcastles (No. 307). Whiteholm Rig (No. 603). Kirkhill (No. 625). Long Cairns Windy Edge (No. 47). “ White Cairn,” Fleuchlarg (No. 249) Capenoch Moor (No. 329). Stiddrig (No. 415). V b 353119 INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. Parish. Canonbie Glencairn . Gretna Kirkconnel . Penpont IT—MONUMENTS NOT IN Parish. Cummertrees Uurisdeer . Holywood . Kirkmahoe . Kirkpatrick-Juxta Lochmaben. f} Morton RuthweU Sanquhar . Applegarth Canonbie Dalton Dryfesdale . Dunscore Durisdeer . Eskdalemuir Glencairn . Hoddom Hutton and Corrie Kirkmahoe . f f Kirkmichael Lochmaben. Middlebie . Moffat Sanquhar . Tinwald Tundergarth Tynron Westerkirk . Miscellaneous. Scots Dike (No. 48). Cross-shaft (portion ol),Hastings Hall, Moniaive (No. 250). Roman Altar (No. 266). Cross-Socket, Orchard (No. 333). Cross, Nith Bridge (No. 531). AND CONSTRUCTIONS DESERVING PROTECTION BUT IMMINENT RISK OF DEMOLITION OR DECAY. Castellated and Domestic Structures. Repentance Tower (No. 89). Tibbers Castle (No. 157). Fourmerkland Tower (No. 280). Dalswinton Old House (No. 338). Auchen Castle (No. 384). Spedlin’s Tower (No. 446). Elshieshields Tower (No. 447). Morton Castle (No. 510). Comlongon Castle (No. 537). Sanquhar Castle (No. 551). Forts. Dalmakethar Burn (No. 20). Roman Camp, Gilknockie (No. 45). “ Range Castle,” Holmains (No. g8). Gallaberry, Dryfeholm (No. 115). Springfield Hill (No. 141). Kirk Burn, Durisdeer (No. 163). Roman Camp, Raeburnfoot (No. 172). Castle O’er (No. 177). “ Mote,” The Orchard, Snade (No. 237). Fortifications, Birrenswark (No. 272). Carthur Hill (No. 291). Vitrified Fort, Mullach (No. 339). Stone Fort, the Belt, High Townhead (No. 342). . ” Wallace’s House,” Kirkland Hill, Burrance Bridge (No. 358). Woodycastle (No. 450). Roman Camp, Birrens (No. 462). Birrens Hill, Carruthers (No. 464). Ericstane (No. 486). ” Kemp’s Castle,” Euchan Water, Sanquhar (No. 557). Barr’s Hill (No. 581). Crawthat Cottage (No. 595). Tynron Doon (No. 609). . ” Bogle Walls ” (No. 638). Camp Hill, Bailiehill (No. 640). VI HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. Parish. Motes. Closebury .... Glencairn .... ) y • • • Dinning (No. 65). Lower Mote, Ingleston (No. 238). Maxwelton (No. 241). Hutton and Corrie Johnstone .... Kirkpatrick-Juxta Lochmaben.... Moffat .... Mote of Hutton (No. 296). Lochwood (No. 316). Coats Hill (No. 395). Rockhall (No. 448). Auldton, Moffat (No. 483). Dunscore . . . . . Lake Dwelling. Rough Island, Loch Urr (No. 144). Closeburn .... Cairns. Threip Moor (No. 72). Gawin Moor (No. 75). Canonbie . . . . . Sanquhar . . . . . Miscellaneous. Spiral-marked Slab, Hollows Tower (No. 43). Cross, Mennock Pass (No. 564). Note .—The following monuments, which are under the charge of H.M. Office of Works, are not included in the foregoing lists : Kirkpatrick-Fleming a Gravestone of Adam Fleming (No. 373). Merkland Cross (No. 378). Ruthwell . . . . . Ruthwell Cross (No. 538). ILLUSTRATIONS. INTRODUCTION. Figure. Name. Page. 1 Annan, c. 1560, showing Mote and Tower ..... xxxii 2 Spiral-marked slab, Hollows Tower ...... 1 3 Motes, and Bruce Stone ....... Iviii 4 Towers ......... Ixi 5 Castlemilk, c. 1547 ........ Ixii 6 Map showing the situation of castles and fortihed houses in the i6th century (from map in the British Museum) ..... Ixiii 7 Crosses ......... Ixvii INVENTORY. No. IN Parish. Figure. Name. Inventory, Annan 8 Bonshaw Tower I 9 Mote of Annan 3 Applegarth 10 Fort, Dalmakethar Burn 20 Caerlaverock II Caerlaverock Castles : block plan 33 (i) 12 Old Castle of Caerlaverock : plan 33 (i) 13 Do. ; splayed base 33 (i) Frontispiece. Caerlaverock Castle 33 (2) 14 Do. ground and first floor plans 33 (2) 15 Do. north front and Gatehouse 33 (2) 16 Do. east wall 33 (2) 17 Do. sections 33 (2) 18 Do. west curtain and base tower 33 (2) 19 Do. elevations 33 (2) 20 Do. second, third, and fourth floor plans 33 (2) 21 Do. interior from the south 33 (2) 22 Do. fireplaces and details 33 (2) 23 Do. elevation of east wing 33 (2) 24 Do. entrance to hall 33 (2) 2.5 Fort, Wardlaw 35 Canonbie . . 26 Hollows Tower 43 27 Do. ; plan 43 28 Roman Camp, Gilnockie 45 29 Long Cairns , etc.. Windy Edge 47 Closeburn . 30 Closeburn Castle : plan 59 31 Mote, Dinning viii 65 Parish. Cummertrees Dalton Dornock . Dumfries . Dunscore . Durisdeer . Eskdalemuir Glencairn . Gretna Hoddom . INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. Figure. Name. No. IN Inventory. 32 Repentance Tower ; plan 89 33 Hoddom Castle 90 34 Do. : plan 90 35 Dalton Church : plan 96 36 Little Dalton Church : plan 97 37 Do. ; window 97 38 Fort, ‘‘ Range Castle ” 98 39 Stapleton Tower : plan 106 40 Robgill Tower : plan 107 41 Coped Stone, Dornock 109 42 Do. do. 109 43 Dumfries Midsteeple 127 44 Do. ; inscribed stone 127 45 Dumfries Bridge 131 46 Do. : plan 131 47 Lake Dwelling, Loch Urr 144 48 Do. do. ; plan 144 49 Drumlanrig Castle ; plan 156 50 Do. ; principal entrance 156 51 Do. : stairs to garden 156 52 Do. ; sundials 156 53 Tibbers Castle ; plan 157 62 63 6d 65 66 66a 67 68 69 70 71 72 73 74 75 76 77 78 54 Do. do. 55 Fort, Kirk Burn 56 Roman Camp, Raeburnfoot 57 Fort, ” Castle O’er Fort ” ; key sketch 58 Do. : plan 59 Do. ; Trench 60 Do. : Ramparts 61 Do. : Entrance Fort, The Knowe Stone Circle, “ Girdle Stanes ” Do. : “ Loupin’ Stanes ” Old Crawfordton : plan “ Mote,” The Orchard, Snade Mote, Ingleston ” Lochmaben Stane ” Churchyard and Church Foundations, Hoddom Bridge : plan Roman Stone in Church Foundations, Hoddom Bridge Birrenswark or Burnswork Do. Do. Do. Do. : plan of Fortihca- tions Roman Glandes from South Camp interior of redoubt Bridle-Bit from 157 163 172 177 177 177 177 177 178 198 199 233 237 238 263 271 271 272 272 272 272 272 Sculptured fragments of Celticcross from Knockhill 273 Pedestal of Roman Altar from do. 273 Other sculptured fragments from do. 273 Cross-shaft from Hoddom 273 IX HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. No. IN Parish. Figure. Name. Inventory. Hoddom . 79 Crosses, Hoddom Church37ard 274 Holywood So Fourmerkland Tower 280 8i Do. : plan 280 82 Stone circle, “ Twelve Apostles,” Holywood 284 Hutton and Corrie 83 Gillesbie Tower : plan 287 Johnstone 84 Lochwood Tower and Mote ; plans 315 and 316 Kirkmahoe 85 Isle Tower 337 86 Do. : plan 337 87 Do. : Heraldic panel 337 88 Dalswinton Old House 338 Kirkpatrick-FIeming . 89 “ Adam Fleming ” stone, Kirkconnel Churchyard 373 Kirkpatrick-Juxta 90 Auchencass or Auchen Castle 384 91 Structure and Incised Cross, Kinnelhead 385 and 386 92 Lochhouse Tower : plan 388 93 Mote, Coats Hill 395 94 ” Camp,” Garpol Water 396 95 Mote, do. 397 96 Fort, Beattock Hill (summit) 401 97 Fort, Stanshiel Rig 403 98 Enclosure, Beattock Hill 412 Langholm 99 Barntalloch Mote 431 Lochmaben 100 Lochmaben Old Castle 445 (i) lOI Lochmaben Castle : plan showing outworks 445 (2) 102 Do. : plan 445 (2) 103 Do. : from a sketch by John Clerk of Eldin 445 (2) 104 Do. 445 (2) 105 Spedlin’s Tower : plan 446 106 Do. : prison 446 107 Elshieshields Tower 447 108 Do. ; plan 447 109 Fort, Woodycastle 450 Middlebie . no Blackwood or Blacket House 460 III Roman Camp, Birrens 462 112 Dedicatory Tablet from Birrens 462 113 Altar to Discipline of Augustus from do. 462 114 Altar to Mars, etc., from do. 462 Altar to Fortune from do. 462 116 Dedicatory slab to Brigantia from do. 462 117 Altar to Harimella from do. 462 118 Fort, Birrens Hill, Carruthers 464 Moffat 119 Breckonside Tower 475 120 Frenchland Tower 480 121 Mote, Auldton 483 122 Fort, Auchencat Burn 485 Morton 123 Morton Castle : plan 510 124 Do. ; Gatehouse tower 510 125 Do. : interior 510 126 Do. : doorway at first floor level 510 127 Cross-shaft, Grierson Museum 514 Mouswald . 128 Grave-slabs at Ruthwell U.F. Church 518 X INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. No. IN Parish. Figure. Name. Inventory. Penpont . 129 Free-standing Cross, Nith Bridge 531 Ruthwell . 130 Comlongon Castle 537 131 Do. : plan 537 132 Do. : S.W. corner of Hall 537 133 Do. : “ yett ” 537 134 Do. : prison 537 Sanquhar . 135 Effigy, Sanquhar Church 549 136 Sanquhar Castle : plan 551 137 Do. : entrance to inner courtyard 551 138 Do. : tower 551 139 Fort, “ Kemp’s Castle ” 557 Tinwald 140 Amish eld Tower : plan, section, and elevation 578 141 Do. ; north face 578 142 Do. : south face 578 143 Do. ; Coloured plaster frieze in Hall 578 I43A Do. ; Carved oak door in National Museum of Antiquities 578 144 Fort, Barr’s Hill 581 Torthorwald 145 Torthorwald Castle 590 146 Do. : plan 590 Tundergarth 147 Fort, Crawthat Cottage 595 Tynron 148 Tynron Doon 609 Wamphrav Westerkirk 149 Sculptured Stone, Wamphray Church 628 150 Fort, “ Bogle Walls ” 638 151 Cist, “ King Schaw’s Grave ” 648 APPENDIX. The Ruthwell Cross (No. 538). Figure. Page. 152. The Ruthwell Cross as now set up in the Parish Church . . . 219 153. The Cross as it stood before its removal to the Parish Church in 1887 . . 222 154. The Bewcastle Cross, showing the four sides ..... 224 155. The Ruthwell Cross, showing the four sides ..... 225 156. The Christ, on the north face of the Cross ..... 227 157. The Flight into Egypt ........ 228 158. Portions of the western and southern faces of the Cross . . 230 159. The Annunciation ........ 231 160. The Upper Arm of the Cross-head ...... 232 161. Table of Runic Futhorcs ....... 236 162. Table of Letters ........ 243 163. Various illustrative pieces ....... 246 164. Cross-forms ......... 247 165. Cross-slab at Hoddom ........ 248 166. Figure of Christ, from Alexandria ...... 252 167. Map of Bernicia and Strathclyde ...... 257 168. Runes on the sides of the Cross ...... 269 XI HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. LIST OF PARISHES. PAGE Annan ...... i Applegarth ..... 3 Caerlaverock ..... 10 Canonbie ..... 26 Closebum ..... 30 Cummertrees ..... 37 Dalton ...... 41 Domock ...... 44 Drvfesdale ..... 45 Dumfries ..... 48 Dunscore . . . . . 55 Durisdeer ..... 59 Eskdalemuir ..... 68 Ewes ...... 80 Glencaim ..... 84 Gretna ...... 92 Hoddom ..... 93 Holywood ..... 104 Hutton and Corrie .... 107 Johnstone ..... 114 Keir . . . . . .119 PAGE Kirkconnel . . . . .120 Kirkmahoe . . . . .121 Kirkmichael ..... 126 Kirkpatrick-Fleming . . . 128 Kirkpatrick-Juxta .... 131 Langholm ..... 146 Lochmaben ..... 148 Middlebie . . . . .159 Moffat.169 Morton ...... 176 Mouswald . . . . .180 Penpont ..... 182 Ruthwell ..... 185 St. Mungo ..... 188 Sanquhar ..... 189 Tinwald ..... 195 Torthorwald ..... 200 Tundergarth ..... 203 Tynron ...... 207 Wamphray ..... 209 Westerkirk . . . . .213 BIBLIOGRAPHY. Abbreviated Reference. Accounts of the Lord High Treasurer of Scotland. Ancient Cross-Shafts at Bewcastle and Ruthwell, by the Right Rev. G. F. Browne, D.D., D.C.L., LL.D. (Cambridge : University Press, 1916). Ancient Scottish Lake Dwellings or Crannogs, by Robert Munro, M.D. (Edinburgh : David Douglas, 1879). Anglo-Saxon Chronicle. ri. 5 . Chronicle. Annals of the Solway until a.d. 1307, by George Neilson, LL.D. (Glasgow : 1900). Annals of Ulster. (Record publications.) Annandale Family Book of the Johnstones, Earls and Marquises of Annandale, by Sir William Fraser (2 vols. Edinburgh : 1894). Antiquities of Scotland, by Francis Grose, F.A.S. (London ; Grose’s Antiquities. 1789). Archceologia Scotica, or Transactions of the Society of Antiquaries of Scotland. Armstrong MSS., in the possession of the Society of Antiquaries of Scotland. Bede. Historice Ecclesiasticce gentis Angloruni. xii Eccl. Hist., or Bede, INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. Birrel’s Diary {See Fragments of Scottish History). Birrens and its Antiquites, by James Macdonald and J. Barbour (Dumfries: 1897). Book of Caerlaverock: Memoirs of the Maxwells, Earls of Nithsdale, Lords Maxwell and Herries, by Sir William Fraser (2 vols. Edinburgh ; 1873). Border Laws (See Leges Marchiarum). Buccleuch and Queensberry MSS., Historical MSS. Commission, 15th Report, Appendix, Part VIII. Caledonia, by George Chalmers, F.R.S., F.S.A. (Paisley : Alexander Gardner. New edition, 1887-1894). Calendar of Documents relating to Scotland, edited by Joseph Bain, F.S.A. Scot. Calendar of Letters and Papers relating to the Affairs of the Borders of England and Scotland, edited by Joseph Bain, F.S.x\. Scot. Calendar of Papal Registers. Calendar of State Papers relating to Scotland and Mary, Queen of Scots, 1547-1603, edited by (i) Joseph Bain, (2) William K. Boyd. Castellated and Domestic Architecture of Scotland, by David Macgibbon and Thomas Ross, Architects (Edinburgh : David Douglas, 1887-1892). Celtic Scotland, by W. F. Skene (3 vols. Edinburgh ; Edmonston & Douglas, 1876). Chronica Gcntis Scotorurn, John de Fordun (Edinburgh ; 1871). Chronicles of the Piets and Scots, and other early Memorials of Scottish History, edited by William F. Skene, LL.D. (Edin¬ burgh : 1867). Closeburn (Dumfriesshire), Reminiscent, Historic, and Traditional, by R. M. F. Watson (Glasgow : 1901). Date of the Ruthwell and Bewcastle Crosses, by Albert S. Cook, Professor of the English Language and Literature in Yale University (Yale University Press, New Haven, Connecticut, 1912). De Origine, Moribus, et Rebus Gestis Scotorurn, John Leslie, Rome : 1378- Documents illustrative of the History of Scotland from the Death of King Alexander the Third to the Accession of Robert Bruce, edited by the Rev. Joseph Stevenson (2 vols. : 1870). Douglas Book, by Sir V'illiam Fraser (4 vols. Edinburgh : 1885). Drumlanrig Castle and The Douglases, with the Early History and Ancient Remains of Durisdeer, Closeburn, and Morton, by Cranford Tait Ramage, LL.D. (Dumfries : J. Anderson & Son, 1876). Early Christian Monuments of Scotland, by J. Romilly Allen and Joseph Anderson (Edinburgh : Neill & Co., 1903)- Early Fortifications in Scotland, by Dr David Christison (Edin¬ burgh and London : Blackwoods, 1898). xiii Abbreviated Reference. Buccleuch,ov Bucc.MSS. Bain’s Calendar,or Bain. Calendar of Border Papers, or Border Papers. Cal. Scottish Papers, or Scottish Papers. Cast, and Dom. Arch. Chron. P. & S. De Origine, etc., Scot- oriim. Stevenson’s Documents. Early Christ. Mon. HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. Early Scottish Charters prior to 1153, collected, with Notes and an Index, by Sir Archibald C. Lawrie (Glasgow ; James MacLehose N Sons, 1905). Ecclesiastical Architecture of Scotland, by David Macgibbon and Thomas Ross, Architects (Edinburgh : David Douglas, 1S96-1897). Exchequer Rolls of Scotland. Fcrdcra, etc., by Thomas Rymer (London : 1704-35). Fragments of Scottish Histor}', by Sir John Graham Dalyell. State of Ancient Scotland—Birrel’s Diary—Expedition in Scotland (Edinburgh ; 1798). Glencairn, Dumfriesshire: The Annals of an Inland Parish, by J. Corrie (Dumfries : 1911). Glenriddell MSS., in the possession of the Society of Antiquaries of Scotland. Growth of a Scottish Burgh, by J. W. Shirley (Dumfries: 1915)- Hamilton Papers. Letters and Papers illustrating the Political Relations of England and Scotland in the i6th century, edited by Joseph Bain. Hisioria Ecclesice Dimelmensis, by Symeon of Durham (London : 1732). Historical Families of Dumfriesshire, and the Border Wars, by C. L. Johnstone, 1889. Historical Manuscripts Commission Reports. Historical Memoirs of the Reign of Mary, Queen of Scots, and a portion of the Reign of King James the Sixth, by Lord Herries, edited by Robert Pitcairn (Edinburgh : 1836). History and Antiquities of the Counties of Westmorland and Cumberland, by Joseph Nicolson and Richard Burn, 2 vols. (London ; 1777). History and Antiquities of Scotland from the Earliest Account of Time, etc., by William Maitland (2 vols. London; 1757)- History of the Affairs of Church and State in Scotland from the beginning of the Reformation in the Reign of King James V. to the Retreat of Queen Mary into England, Anno 1568, by the Rev. Robert Keith, Bishop of the Episcopal Church in Scotland. (Edinburgh ; 1734). History of the Burgh of Dumfries, with notices of Nithsdale, Annan- dale, and the Western Border, by W. M'Dowall (Edinburgh : 3rd edition, 1906). History of the Church of Scotland, by John Spottiswoode, Arch¬ bishop of St Andrews, with Biographical Sketch and Notes by the Right Rev. M. Russell, LL.D., 3 vols. (Edinburgh : 1847-51—Spottiswoode Society). History of Dumfries and Galloway, by the Right Hon. Sir Herbert Maxwell, Bart. (Edinburgh and London : Blackwoods, 1896). History of the House of Douglas, by the Right Hon. Sir Herbert Maxwell, Bart. (2 vols. London : Fremantle & Co., 1902). Abbreviated Reference. Eccles. Arch. Birrel’s Diary. Symeon of Durham. Hist. MSS. Comm. Herries’ Memoirs. History of Westmorland, etc. Maitland. History of Affairs, etc., or Keith. Spottiswoode’s History, or Spottiswoode. Dumfries and Galloway. XIV INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. History of the Johnstones, with descriptions of Border Life, 1191- 1909, by C. L. Johnstone (Edinburgh ; 1909). History of the Kirk of Scotland (1514-1625), by the Rev. David Calderwood, A.M., edited from the original MS. preserved in the British Museum by the Rev. Thomas Thomson, 8 vols. — Woodrow Society—(Edinburgh : 1842-49). History of Liddesdale, Eskdale, Ewesdale, Wauchopedale, and the Debateable Land, by Robert Bruce Armstrong (Part 1 . Edin¬ burgh : 1883). History of Moffat, with frequent notices of Moffatdale and Annan- dale, by W. R. Turnbull (Edinburgh : 1871). History of Sanquhar, by James Brown (Dumfries ; J. Anderson & Son, 1891). History of Scotland from the accession of the House of Stewart to that of Mary, with Appendices of original papers, by John Pinkerton, 2 vols. (London : 1797). Itinerariiini Septcntrionale, by Alexander Gordon (London ; 1726-32). Langholm as it was, by John and Robert Hyslop (Langholm : Robert Scott, 1912). Leges Marchiarum, or Border Laws, by William (Nicolson) Lord Bishop of Carli(s)le (London ; 1705). Letters and Papers, Foreign and Domestic, of the Reign of Henry VIII. Liber Gardrobee, Rolls Series. Lochmaben Five Hundred Years Ago ; or Selections, Historical and Antiquarian, from Papers collected by John Parker, by the Rev. William Graham (Edinburgh : 1865). Lord Wardens of the Marches of England and Scotland, by Howard Pease, M.A., F.S.A. (London : Constable & Co., Ltd., 1913)- Macfarlane’s Geographical Collections, edited by Sir Arthur Mitchell, K.C.B. — Scottish History Society—(Edinburgh : T. & A. Constable, 1906-1908). Manuscripts of J. J. Hope Johnstone, Esq., of Annandale, — Historical Manuscripts Commission — 15th Report, Appendix, Part IX. (London : 1897). Memorials of St Michael’s Churchyard, Dumfries, by Win. M'Dowall (Edinburgh ; A. & C. Black, 1876). Military Antiquities of the Romans in North Britain, by Major- General Roy (London ; 1793). Minstrelsy of the Scottish Border, etc., by Sir Walter Scott (Kelso and Edinburgh : 1802-3). Origines Parochiales Scotice, edited by Cosmo Innes — Bannatyne Club—(Edinburgh ; 1850-55). Peel ; its Meaning and Derivation, by George Neilson (Glasgow; 1893). Proceedings of the Society of Antiquaries of Scotland. Register of the Privy Council of Scotland. Registrum Magni Sigilli Regum Scotorum. The Register of the Great Seal of Scotland. Abbreviated Reference. History of the Kirk, or Calderwood. Armstrong or Arm¬ strong’s Liddesdale. Gordon. Johnstone MSS. Roy. Orig. Paroch. Proc. Soc. Ant. Scot., or A ntiquaries. Reg. P. C. Reg. Mag. Sig. XV HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. Rcgisinon Episcopatus Glasguensis —BannatjTie Club— -2 vols. (1843). Repentance Tower and its Tradition, b}’ George Neilson (Glasgow ; iSQo)- A Roman Frontier Post and its People : The Fort of Newstead, in the Parish of Melrose, by James Curie, F.S.A. Scot., F.S.A. (Glasgow : James MacLehose & Sons, 1911). Runic Roods of Ruthwell and Bewcastle, by James King Hewison, i\LA., D.D., F.S.A. (Scot.) (Glasgow : John Smith & Son, Ltd., 1914). Scoiichronicon, edited by W. Goodall (2 vols. Edinburgh : 1759). Scottish Historical Review (Glasgow : MacLehose). Scots Lore (Glasgow ; 1895). Series of Etchings, chiefly of Views in Scotland, by John Clerk, of Eldin—Bannatyme Club—(Edinburgh : 1855). Siege de Karlaverok, edited by Sir Nicolas Harris Nicolas. State Papers. Henry VIII.—vols. iv. and v. Statistical Account of Scotland (1797). Statistical Account of Scotland, New (1845). Tour in Scotland in 1769 and 1772, by Thomas Pennant. Tours in Scotland in 1747, 1750, and 1760, by Bishop Pococke, edited by D. W. Kemp—Scottish History Society—(Edinburgh : T. & A. Constable, 1887). Transactions of the Dumfries and Galloway Natural History and Antiquarian Society. Transactions of the Glasgow Archaeological Society. Vita St Kentigerni. The Historians of Scotland, vol. v. (Edinburgh : Edmonston & Douglas, 1874). Vita Sancti ColumbcB —(Adamnan ed. by Reeves)—Bannatyne Club (Dublin ; 1857). Vitruvius Britannicus (Colin Campbell; 1767). Wamphray : Pages from the History and Traditions of a famous parish in Upper Annandale, illus. : by J. Paterson. (Lockerbie ; 2nd edition, 1907). Abbreviated Reference. Registrum Epis. Glasg. Scot. Hist. Rev. Stat. Acet. New Stat. Acet. Pennant. Pococke. Trans. Dumf. and Gall. Antiq. Soc. Trans. Glasgow Arch. Soc. Vita St Kentig. Vita ColumhcB. XVI INTRODUCTION TO INVENTORY OF ANCIENT AND HISTORICAL MONUMENTS AND CONSTRUCTIONS IN THE COUNTY OF DUMFRIES HISTORICAL INTRODUCTION. 1 . Historical Geography. Dumfriesshire is virtually the West March ^ of old Border days, Galloway proper being an outlying district in history as in geography. Its northern region is part of the Silurian upland of southern Scotland, and is deeply trenched on the west side by the valley of the Nith, which valley also marks a division between the more monotonous high land to the east and the massive and boldly outlined hills of Galloway. On the east side, too, a mountainous country extends between the basins of the Esk and the Teviot. Southwards, towards England, Dumfries¬ shire inclines hrst to a gently undulating country and then to a great flat, which, along the shores of the Solway, offers a “ Merse ” or marshy tract, a mere fringe of waste, however, in comparison with its nominal counterpart in Berwickshire. Superficially, indeed, these two counties have much in common. Both pass without serious obstruction into the north of England plain on either- side of the Pennine range. Both offer an open road round an extremity of the Cheviots, which so effectually cover the intermediate shire of Roxburgh. But the western gate, if flatter than the eastern, is also narrower. On the other hand, the lower Esk was an even less serious obstacle than the Tweed. It offered no difficulties of fording. In 1745 the Jacobite army in its retreat from England passed across this river in a column one hundred men broad, when “ the water was big and took most of the men breast-high.” ^ At low tide on the Solway there were crossings also far down the channel of the river, where it is subject to overflow by the water of the firth, one below the town of Annan to Bowness in Cumberland—where a railway line is now carried over to the southern shore—and another from Dornock to Drumburgh. The latter was known as the " Sandywathe,” ^ while the regular ford on the Esk, the nearest to the mouth and on that account the most important of the river fords, was of old ^ Ewesdale and even Eskdale are sometimes referred to as in the Middle March. Cf. Ma.wvell’s Dumfries and Galloway, p. 158. * Lord G. Murray’s Journal. ^ Chronicon de Lanercost, Bann. Club, p. 272. xvii HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. known as the “ Siilwath,” a name signifying the “ muddy ” {sol—mud), in contrast to the “ sandv,” watli or ford and later transferred, in the form “ Solway,” to the lirth as a whole.^ The English chronicler Knighton tells how, in July 1335, Edward III. made a plundering raid upon Scotland from Carlisle, crossing the vadum Sulwath on entry and returning b}’ the vadum AnandicB- In the 14th century, in the days of Caeriaverock's great siege, the firth was known as the ” Irish Sea,” and it is still so named by Bishop Leslie in the late i6th centuryFrom the latter half of the 12th century the Esk had been the recognised boundary of Scotland. At some date in the first quarter of the 14th century the men of Cumberland and Westmorland, about whose services on the Border there had been dispute,^ represent to Edward III. that ” the service due in war to his ancestors ” on their part was “ that, on his march to Scotland, the}^ should meet him at the Rerecross on Stanemoor and go in his vanguard as far as ‘ la Marche de Solewathe,’ ” taking the rearguard on the return.^ But the lower Esk was not suffered to remain much longer as the frontier line. So accessible was the land on its northern side as a mere prolongation of the level to the south, and so intermingled and homogeneous the population in consequence, that the district between the Esk and the little river Sark became a ” debateable land ” between the two countries, and its inhabitants even were familiarly referred to in the i6th century as the ” Baitablers.” This feature had a profound influence on the histor\* of the MTst March. The fact that there was no clear definition of juris¬ dictions made it an ideal resort for the more lawless spirits of the Border, who, while the wardens jealously disputed, went their own way. ‘‘ For neither I will suffer the warden of Scotland to answer for it,” Lord Dacre informs the English Privy Council in 1550, ” because I will not affirm it to be Scotland, nor will they, on the contrary, consent that it shall be England.” ® The usual provision for the Debateable Land in truces between England and Scotland was that it should not be occupied on behalf of either kingdom, “ ‘ neither with stub, stake, nor otherwise, but with bit of mouth for pasturing of cattle ’ from sunrise to sunset, according to old custom.” ^ The Prior of Canonbie, however, was allowed to enclose and build upon his section, about four square miles in area. The trading relations of the district with Carlisle formed the basis of the English claim to Canonbie as being really part of England. But in the lengthy diplomatic corre¬ spondence over the question (c/. p. xxxvi.) the Scots would always furnish a counter¬ plea to each argument, and in logic no decision was possible. The unsatisfactory condition of the district, however, forced on the question of its delimitation, and in 1552 Commissioners from the two countries apportioned the doubtful territory. After all, it appears the local borderers had preferences in the matter, which preferences were taken as a guide to division. The inhabitants of the eastern part had their inclinations set towards Scotland, those of the western towards England ; and it was settled so. Thus the Canonbie part became Scottish, while England had the barony of Kirkandrews, once property of the Rossedals of the ^ Fordun refers to fluvium Esk, quod dicitur Scotiswath sive Sulwath {Chronica Gentis Scotonmi, lib. ii., cap. ii.). Cf. p. xxx. ^ Chronicon (R.S.), i. p. 472. ® See also Neilson’s Annals of the Solway, passim ; Trans. Diimf. and Gall. Antiq. Soc., 1895-6, p. 156. ^ Bain’s Calendar, ii. No. 1134. ^ Ibid., iii. No. 716; la Rerecroiz sus Estaynmor. ^ Letter in Nicolson and Burn’s History of Westmorland, etc., i. p. Ixxv. ' Treaty of Berwick, Dec. 1528, Letters and Papers For. and Dom., Henry VIIL, vol. iv. part ii. ^'0. 5030 ; Leges Marchiariim (1549), Nicolson, pp. 80-81. xviii INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. lower Esk. The line of division was drawn in the form of a “ditch or furrow’’ (fossa vel sulco) from a bend of the Sark to another bend on the Esk, the particular spots being marked by squared, pointed stones bearing on the face looking east the Scottish royal arms, and on that looking west those of England. In case of accident, the actual position of these stones was also topographically dehned in the deed of division. Henceforward the boundary between the countries is the Sark and the “ Scots Dike ” or ditch.^ It is the rivers that are the determining geographical feature of Dumfriesshire. It is their basins, Nithsdale, Annandale, and Eskdale, that are its historic units. The higher land to the east also gives the chief tributaries of the Esk a dehnite importance, as Ewesdale and Wauchopedale. The slope of the country is southwards, and the hills send their spurs southwards. Between the vales of Nith and Annan lie the Torthorwald Hills and the Lochar Moss; Annandale, in its upper division, is separated from Eskdale by a high-lying plateau. Of the dales Annandale was easily, from its central position and its extent, pre-eminent in a geographical sense. Its roots spread wide. The original grant to the family of Bruce extends the area from the border of Cumberland on the one side to that of Nithsdale on the other, and in this sense it was generally understood : “ The Stewartrie of Annandale from Erickstone (or rather Tweeds Cross which is a mile farther north and the boundary of Tweddale) to xVlisonbank the southmost part and outmost limits of the Kingdome will be 27 miles in length from North to South ; and from Mortoun town alias Tower of Sark on the east to the Castle of Cockpool alias Cumlongan ^ on the west will be about 14 myles in breadth.” ^ For this reason it seems sometimes to have been used as equivalent to the present county.^ An early description includes Annandale in the region of Galloway, without specifying the other dales.® Of much importance, too, were the many patches of morass that anciently distinguished the more level parts of the country, and still characterise these to a conspicuous extent. Between Lochar Moss east of the Nith and “ Sollom ” Moss by the Esk stretched a chain of such obstructions, of which Hightae Moss and Nutberry Moss are considerable survivals, and which seriously limited the approaches westwards. Lockerbie was surrounded by mosses. But boggy land was not confined to the south. Lochwood Tower stands on the margin of what is still a considerable morass. The Cairn valley in the parish of Glencairn must once have had extensive bogland. But the rivers, as usual in Scotland, in contrast with England, favour advance north and south. The manner in which one railway follows the line of the Nith and the other that of the Annan graphically records this determinant. It is indicated even in the shape of the parishes, which tend to have their longer axis in these directions. The great historic families of Dumfriesshire are apportioned to the dales, and all Dumfriesshire history—economic, administrative, and military- moves along their furrows. ^ The frontier was defined thus : ut in ipso utrius partis discrimine trames linearis rectus transversim ab Esk ad Sark fluviuni ducatur, fossa vel sulco vestigium ipsius denotante (Rymer’s Fcedera). The “ Scots’ Dike ” now, however, appeal's as a low mound, with the trace of a shallow ditch on each side, running in a straight line through the middle of the plantation on the boundary, e.g. on each side of the road going south from Glenzierfoot. What remains is in danger of being obliterated. 2 The castles of Cockpool (No. 542) and Comlongon (No. 537) are really different places. ® Macfarlane’s Geographical Collections (Scot. Hist. Soc.), i. pp. 365-6. Cf. Calendar of Border Papers, i. pp. 393-4. ® Description of Scotland, 1292-6, in Chronicles of the Piets and Scots, p. 215. XIX HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. Tlie same cliaracteristic also determined the great historical jurisdictions. Dumfries is of the type of sliires which take their name from the principal town, that town having achieved its original importance as a military centre. The town of Dumfries had a royal castle and a sheriff about the end of the twelfth century. Later the sheriffdom of Dumfries becomes synonymous with Nithsdale, but in its fullest extent included Gallowa}’ east of the river Cree. Annandale, when it became a Crown holding, ranked as a stewartry (see p. xxvi.) having its courts at Loch- maben. The lordship of Eskdale was erected into a regality for the Douglases. These jurisdictions became hereditary, and compensation had to be paid to their owners on their abolition in 1747. II. Early History. The Romans found in Dumfriesshire a people whom they call SelgovcE, a word which may contain the Celtic root selg, “ hunting," and so mean' “ the huntsmen.” In Ptolemy’s map the Selgovse are given four towns or fortified sites, of which two are east of Novii ostia or mouth of the Nith : Uxellum meaning “ the height,” ^ which has been allotted to the enclosure with ramparts and ditch on Wardlaw Hill (No. 35) and Trimontium “ triple hill,” identified with the fortified summit of Birrenswark or Burnswork ^ (No. 272). But the calculations underlying the map are not likely to be even approximately accurate, and ‘ ‘ Trimontium ’ ’ is generally placed, with all possible plausibility, at the “ triple peak ” of Eildon. Birrenswark is the best-defined and farthest-seen ” height ” of southern Dumfriesshire, but identification with Uxellum or of the more inland Corda with Sanquhar is little better than guesswork. That a Roman route went northwards through Dumfriesshire to the limes between the Forth and Clyde is very probable, though to lay it out is another matter.® Various indications go to suggest that the station at Birrens was about the last place in Scotland to be held in the clutches of the imperial eagle. There is evidence of an early occupation, and abundant evidence of an occupation in the second half of the 2nd century. The Roman camps at Gilnockie (No. 45) and Raeburnfoot (No. 172) suggest operations in the Esk valley, either as a short route to the Tweed valley at Peebles, where there is another camp, or as the scene of an expedition against the tribes who then occupied this tract of ground and have left so many impressive traces of their presence in the hill forts of Eskdalemuir, particularly in Castle O’er (No. 177) and Bailiehill (No. 640). The first historic figure to be associated with Dumfriesshire was Kentigern or St Mungo. The county was then part of the “ Cambrian region,”^ which, in its fullest extent, extended from the neighbourhood of the Clyde to the English Channel, and explains the saint’s personal connection also with Cumberland and North Wales. Later this continuous strip of land, held by the resisting Britons, was broken up by Saxon intrusion. ^ Cf. “ Ochil ” Hills ; the phonetic change to ch is Brittonic, not Gaelic. 2 Skene, Celtic Scotland, i. p. 72. Skene suggests that Trimontium represents a native word with the Welsh form Tre or Tref, “ town,” and so Trefmynydd or ” the town on the mountain.” ® On " Roman ” and other early roads in Dumfriesshire see articles by Dr. James Macdonald, Proc. Soc. Ant. Scot., vol. xxviii. (1893-4), pp. 43-4, 298-320; and Hislop’s Langholm As It Was, pp. 113-7. ^ " regionis Cambrensis,” Vita St. Kentig., cap. xi. But this use of the name is late {cf. p. xxiii.) ; Jocelyn wrote the life in the twelfth century. XX INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. Kentigern flourished in the 6th century, when one result of the troubles through which the island had passed was a plain set-back to Christian teaching. Columba was a younger contemporary. It was but a Christian remnant among the northern Britons, though it included the local king, who selected Kentigern, as a young man, for their bishop. His see he fixed in what was to be Glasgow. There arose in time a king of a different persuasion, whose kin finally made it so hard for Kentigern that he had to take refuge in Wales, where he remained till after the battle of Ardderyd in 573. Ardderyd is clearly to be identified with a site on the south side of the River Liddell, a plain “ between Lidel and Carwanolow,” ^ the latter a small southern tributary of the Esk. A victory for the Christianising party in British politics, it raised Rederch or Rydderch “ Hael ” or “ Roderec the Liberal ” to the throne. His Irish designation—his mother was Irish—was King of Alcluyd, “ the rock by the Clyde ” or Dumbarton.^ So he and his successors are styled in the Irish annals. Dumfriesshire preserves the name in Carruthers, “ the caer (Brittonic) or fort of Rydderch.” Rydderch secured the return of the discreet and tactful Kentigern to his kingdom, wherein the Christian religion had well-nigh perished. What thus amounted to a fresh missionary effort had its beginning on the haugh of Hoddom,^ where King Rydderch and a multitude of the people met the returning apostle, who forthwith addressed the gathering, assuring them that their idols were the wt)rk of men’s hands, that the elements that they deified were but instruments of their Maker, and that Woden whom they, and especially the Angles, worshipped was probably once merely a mortal king of the Saxons. This adoption of Woden by a Celtic people indicates a change of religion which was a tribute to Saxon success in conquest. An edifying miracle occurred when the flat where Kentigern was placed rose into a not incon¬ siderable little hill,^ and as such it remained in the days of the narrator, the 12th century, and presumably ever since. Possibly, therefore, Trailtrow Hill overlooking Hoddom, where a graveyard still exists in a not very suitable situation because a chapel once stood there, may mark the place of the preaching of Kentigern ; old Hoddom Church was by the river bank (No. 271). For a time, too, ” Holdelm ” was further honoured by being made the bishop’s see, where churches were constituted and clergy ordained. Hoddom was thus an ecclesiastical centre of much importance during a brief period, till circumstances secured the re-establishment of the see in Kentigern’s own city of Glasgow.” ^ We are told by his biographer that it was the custom of Kentigern to erect a cross—of stone presumably in his opinion, since the two specifically mentioned are of stone—in any place where he had made converts or had lived for some time.^* Certainly the crosses and fragments of crosses in Dumfriesshire make a remarkable group, and those formerly standing at Hoddom are, no doubt, due to the saint’s special connection with that neighbourhood, though their date is much later in time. Kentigern’s crosses, like that of King Oswald in Bernicia as late as 635, were in all probability of wood." And before Oswald’s cross there was no outward sign of the Christian faith in that province. ^ Scotichronicon, bk. iii. cap. x.xxi. 2 “ Petra Cloithe ” in Adamnan’s Vita Columbce. 2 “ in planitie campi, vocabulo Holdelm,” Vita Kentig., cap. xxxii. ^ ‘‘ monticulum altum.” ibid. s cap. xxxiii. ® Ibid., cap. xli. ’’ Bede, Eccl. Hist., iii. cap. 2. But in the 12th-century biography he is credited with a great cross of stone at Glasgow and another miraculously made from sea-sand {de sola arena mar is) at ” Lothever- werd ” [Vita Kentig., cap. xli.). XXI c HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. Kentigern and Rydderch both died in the same year, 6oi, after which the records of "the northern Britons again become scrappy and rare. In 613 the battle of Chester marks the piercing of the British line by the Northumbrian Angles, and the northern Britons are definitely dissociated from those in Wales. The farthest limit of the northern section was the River Derwent in northern Lancashire, which down to the 19th centnr}" formed the boundary between the bishoprics of Carlisle and Chester. But in the histor}" of this province boundaries are uncertain. Rydderch seems to have carved out a kingdom based on Alcluyd, and his control certainly extended to the Solway, and perhaps beyond. References to the Rerecross on Stanemoor, now on the boundary between York and Westmorland, suggest that here once ran the line between Cumbria and Northumbria.^ But we do not hear of a particular territorial name till we reach a reference to Strathclyde [Stratha- Cluaidhe) in the Annals of Ulster under 873, and another to the people, two years later, in the Anglo-Saxon Chronicle. Thereafter Strathclyde is generally used for this variable kingdom. The fortunes of the portion later known as Dumfries cannot, of course, be separated from those of the kingdom as a whole, but certain occurrences have more definite bearing upon that quarter. From the reign of Oswald of Northumbria (634-642), through that of Oswy, and down to the defeat and death of Ecgfrith at the hands of the revolting Piets in the battle of Dunnichen, 685, the Britons were subject to the Northumbrian kingdom. As kings of Alcluyd are nevertheless also mentioned during this time,^ it may be taken that their status was that of vassals, and that the subjection of the people amounted to the payment of a yearly tribute, as in the case of the Piets and Scots.^ Civil supremacy, however, again as in the Pictish case, would bring with it ecclesiastical control, and this epoch of Northumbrian lordship, extending to about half a century, has been fixed upon as one of the possible occasions on which a cross inscribed with a Northumbrian poem in runes might be erected at Ruth well (see Appendix). The Northumbrian disaster at Dunnichen restored to “ some part ” of the Britons their liberty.^ As we find that the Anglian hold continues on the west side, since in 696 Cunningham (Ayrshire) is reckoned Northumbrian,^ while in 731 the Anglian bishopric of Candida Casa, “ White House,” or Whithorn in Galloway, being in the province of the Bernicians,® has just been constituted,'^ we infer that the base for this and later advances in the west must have been the British lands south of the Solway, that the Britons freed by Dunnichen were therefore north of that firth, that Dumfriesshire, and particularly Nithsdale, afforded the approach to Cunningham, to Edbert’s acquisition of Kyle (Cyil) in 750,® and the attack by that monarch and the Piets on “ Alcwith ” or Alcluyd in 756, when the Britons were reduced to terms and Edbert’s army perished [interiit) on the return,® that there¬ fore the free Britons were those of the Clyde valley, and that Northumbrian dominion on the west and south thus particularised them as the “ Strathclyde Welsh,” which name appears on record in the following century. ^ Cf. reference to boundaries of old Scottish kingdom in Fordun, Chronica Gentis Scotonim, bk. iii. cap. ii., de mora lapidea, “ from stone-moor.” Modem forms : Rey Cross, Stainmore. ^ e.g. the Annals of Ulster give in 642 ” Hoan ” (Ewen), King of the Britons, as the slayer of Donald Brec, and in 657 the death of Gureit, King of Alcluyd. ® Bede, ii. cap. 5. ^ Ibid., iv. 26. ® Ibid., v. 12. ® Ibid., iii. 4. Ibid., v. 23. ® Chronicon, Bede, ed. Stevenson, Eng. Hist. Soc. ® Symeon of Durham, R.S., ii. pp. 40-41. xxii INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. Early in the 9th century the line of Anglian bishops in Whithorn ceased, and there is a blank but probably anarchic time in the south-west,^ till the striped sails of the Northmen rose against the sunset, bringing no peace but a sharper sword. In 870 Alcluyd fell after a siege of four months by Danes and Norse from Ireland. Five years later Half dan with his Danes traversed the country into Northumbria, and is recorded to have wasted the “ Strathclyde Welsh," ^ or “ Strathclydians,’’ ^ or, as in a third place and for the first time, the “ Cumbri." This was an enterprise of Danes from Ireland, and does not seem to have been more than a foray of ex¬ ceptional destructiveness. Permanent settlement was the work of the Norse bear¬ ing elements of Irish culture. Tinwald {Thing-vollr), “ field of the meeting," a short way north-east of Dumfries, seems to have been, as its name suggests, the centre of local Norse control. The first serious settlement would be about the year 880, which date would apply also to the settlements in Cumberland. For the Norse as for the Britons the land north and south of the Solway was all one. So much is made clear by the place-names. Certain of these are identical on both sides of the firth, e.g. Eskdale (or Askdale=Ashdale), Dalton, Brydekirk, Ousby (Oseby), Canonbie, etc. ; others have identical elements, as in the various compounds with thwaite {“ clearing " or “ sloping pasture ") and by {“ settlement ") and in Smailholm [smali, small cattle : cf. Cumberland “ Smallthwaite "), Closeburn (Kil-Osbjorn), the old Butterthuate or Butterquhat [cf. Buttermere), Langholm etc. Many names belong, however, not to the early settlement but to later times when the latter part of the compound was an established local form, e.g. Lockerby (1198 Locardebi) registers the personal name Locard introduced to Scotland in the 12th century [cf. p. xxiv.). Fell, beck, and gill names, which also occur in Cumberland and Westmorland, occur here as far north as Moffat. Applegarth may be compared with Appleby in Westmorland and Calgarth [i.e. “ calf-enclosure ") at Windermere. The Dumfriesshire Norse, equally with their British neighbours, might be expected to regard unfavourably the imperial activities of the expanding West Saxon kingdom, and Owen, King of the Cumbrians, is one of the kings allied against Athelstan on the occasion of his great victor}/ at Brunanburgh in 937.^ This alliance was in contempt of the arrangement of thirteen years earlier, when the Strathclyde Welsh accepted Edward, the elder brother of Athelstan, as their lord.® Moreover, it was to continue a troublesome district from the English point of view. Its mixture of British and Norse blood did not render it more amenable to outside guidance. Athelstan’s brother Edmund in 946 subjected Strathclyde to another wasting, and finally handed over the kingdom to Malcolm I. of Scotland on condi¬ tion of co-operation by land and sea. Yet in 1018 we again hear of the death of a king of Strathclyde, the last, as it happened, for Malcolm II. of Scotland now placed on the vacant seat his son Duncan, who was to be his own successor. So Strathclyde merges at last into the vider realm of Scotland. Since the beginning of the loth century, when the royal line of Alba had already supplied an occupant for the throne, Scottish permeation and influence had been growing. Its final outcome was the obliteration not only of the independence of ^ Cf. p. xxiv. 2 “ Straecled-Walas,” yl.S. Chronicle. ^ “ Stratcluttenses,” Asser (late gth century), De Rebus Gestis jElfridi. ^ Ethelwerd (late loth century). Chronica, bk. iv., in Monumenta Historica Brit., p. 515, “ A plea has been made for the identification of Brunanburgh with the hill of Birrenswark. See Scot. Hist. Rev., vol. vii. (1909), pp. 37-55. ® Anglo-Saxon Chronicle. Florence of Worcester makes it 921. xxiii HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. the kingdom but also of its language, with probably the infusion of a fresh Gaelic element in the topography, where appear names such as Dunscore (sgdr^a sharp rock), Duncow (ro//=hazel). Lag (=hollow), the old Dunberton {i.e. Briton) in Lochmaben parish. Glencairn and the other “ glen " compounds. Gaelic ousted Welsh, and was still spoken in Carrick in the i6th century, before it in turn was ousted by Scots. The period of the invasions subsequent to the time of Kentigern and his suc¬ cessors is painted in vert’ sombre colours in the Inquisition of David concerning the lands of old possessed bt^ the bishopric of Glasgow, the date of which is the first quarter of the 12th century. Insurrections arose, we are told, the Church was destroved, lands were wasted, and good men driven into exile. Then into the desolate country poured diverse tribes of different nations, unlike in race, language, and customs, among whom paganism prevailed over the Christian faith. Probably in all this there is not only compression of facts but some exaggeration in the interests of the reforms of Prince David.^ III. Territorial Families. It was David I. who left the deepest mark on Dumfriesshire. Before coming to the throne he held a position in southern Scotland, as “ Prince of Cumbria,” which cannot be defined in relation to either power or territory. It seems, however, to have been an actual division of the kingdom, though only his brother Alexander was known as king. In 1124 David himself succeeded to the throne, and so reunited the realm. Virtualh^ by training and preference, David belonged to that international race whom we know as Normans. He did two big things in the south-west. He re¬ constituted the bishopric of Glasgow, which included the old Strathclyde kingdom, and he settled the Yorkshire family of Bruce in Annandale (c. 1124.) The Bruce domain included ” Estrahanent ” (Annandale) and all the land from the boundary of Dunegal of Nithsdale to the boundary of Randolph “Meschin” ( = “the younger”), who possessed Cumberland; that is, as far at least as Gretna. The principal seat of the family at first seems to have been Annan, which was a con¬ venient centre for communication with their Yorkshire lands, but Lochmaben and probably Moffat were also residences. In the original charter Robert de Brus is, b}’ implication, licensed to erect a ” castellum,” ^ but at what place is not indicated. Round the greater light of Bruce gathered the lesser lights of Annandale, some, like their overlord, Norman immigrants, others apparently of the earlier local stock. Names of witnesses attached to Bruce charters of the end of the 12th and the begin¬ ning of the 13th century are these, being names which were to be long familiar in the histoiA’ of the shire ; Robert de Hodalmis or Hodelm (Hoddom), Humphrey del Gardine (Jardine), William de Herez (Herries), Edward de Hodalmia (Hoddom), Hugh de Corri, Robert de Crossbi, Roger de Kirkpatrick, Malcolm Loceard, Sir Gilbert de Jonston, David de Torthorwald. Of these the Johnstones rose to greatest importance, but their early family records perished in a burning of Loch- wood by Maxwells in 1585 (see No. 315). “ Dunegal of Stranith,” who in the ^ Registrum Epis. Glasg., p. 5. Cf. also Scots Lore (1895), p. 36 ff., and Lawrie’s Early Scottish Charters, No. 4. 2 ggg Neilson in Trans. Dumf. and Gall. Antiq. Soc., 1914-15, p. 58. XXIV INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. I2th century was lord of the larger part of Nithsdale, clearly represents a survival among the great Celtic landowners then in process of being replaced by Normans. From his eldest son the family took the surname of Randolph, and its best known representative was the Thomas Randolph of the War of Independence, nephew of Robert the Bruce and first Earl of Moray. A junior branch similarly adopted the surname of Edgar, and in the early 14th century was possessed of the castle of Sanquhar and half the baroity.^ At the southern end of the Nith the family of Maxwell, from the neighbourhood of Kelso in Roxburghshire, probably acquired the barony of Caerlaverock early in the 13th century.^ Above Dumfries are Dalswinton and Duncow, in which there were Corny ns. In the immediate neighbourhood was the barony of Tinwald, which appears to have been in possession of the family of Mande- ville since the time a Mandeville married an illegitimate daughter of William the Lion. In part Tinwald Mote (No. 582), original messuage of the barony, still survives. About the middle of the 15th century the line ended in heiresses, of whom one married Edward Maxwell.^ Maxwell in course of time acquired the other portions of the barony. Eskdale in the 12th and 13th centuries was wholly Norman in lordship. In the upper valley and part of the lower were the Avenels, conspicuous patrons of the Abbey of Melrose. This line ended in heiresses and the Dumfriesshire lands passed to the husband of the elder, a Graeme of Dalkeith. In the middle Esk the barony of Westerkirk {Wathstirker, Watiskirker) was in the possession of the great Liddesdale family of De Soulis. In Ewesdale were Levels ; in Wauchopedale, after 1285, Lindsays, who, with a break extending substantially over the i6th century, continued there till 1707.^ Lower Eskdale was largely owned by the Rossedals (Norse hross-dalr, “ horse-dale ”), another family almost entirely known for its gener¬ osity to Jedburgh Abbey (as the Avenels for their connection with Melrose) and its foundation of the Priory at Canonbie for Augustinian canons as a cell of Jedburgh. The Rossedals make a silent and unexplained exit from history. The War of Independence, and the long struggle against England, brought about a partial redistribution of Dumfriesshire lordships. A temporary imposition of some English owners may be neglected. A Bohun or a Percy in Annandale was but a bird of passage. It was the Scottish kings proper who had the final word. As a result, the main territorial feature in the district during more than a century is the stead}' expansion of the wealth and power of the Douglas family by grant and acquisition. The first of the family to own lands in Dumfriesshire was the “ good Lord James,” Bruce’s friend, when in 1321 he had a grant of lands in the barony of Westerkirk.® That followed on the elimination of De Soulis, who had been forfeited and executed on a treason charge a year before. About twenty years later, the Levels, as supporters of England, disappeared from Ewesdale, and their lands too were added to the Douglas holdings, being transferred to William, nephew of Lord James, first Earl of Douglas.® The most important of the Douglas vassals in this quarter was the knightly family of Glendinning from Roxburghshire in Eskdale- muir." In Ewesdale again were a branch of the Teviotdale Frasers, till, on their resignation, the property in 1426 was granted to Simon Lytil or Little,® with which family it remained for quite two centuries. ^ Reg. Mag. Sig. (new edition), i. p. 8, No. 27. 2 Orig. Paroch., i. p. 446 ; Book of Caerlaverock, i. p. 40. ^ Exchequer Rolls, vi. p. r68. •* Armstrong’s Liddesdale, p. 168. ® Reg. Mag. Sig., i. pp. 522, 544. ® Ibid., p. 565. Armstrong’s Liddesdale, p. 160. ** Reg. Mag. Sig., ii. No. 48. XXV HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. By the elevation of the Bruces to the throne Annandale became a Crown hold¬ ing. and was conferred by Robert I. on his nephew Thomas Randolph, lord of Xtthsdale and Earl of ]\Iora3^l The Moray line ended in a daughter who married the Earl of ;March, and when Earl George was forfeited in 1409 Annandale was acquired b\- the Earl of Douglas,'^ from which family it reverted to the King after the tragedy of 1440. James II. conferred the lordship upon his younger son, the Duke of Albany, who forfeited it b}' rebellion ; and in 1487 it was finally annexed to the Crown. Having been granted to Randolph as a free regality— i.e. with a jurisdiction regal in scope, as far as was possible to a vassal,—on reversion to the Crown it ranked as a stewarfr\’, while Lochmaben remained a royal castle under a constable. In later dsLYS the two offices of Steward of Annandale and Constable of Lochmaben were usuallv held by one person (see p. xli.), and from 1410—first, in the sense of regalitv-depute, under the Earls of Douglas and then from 1455 under the Crown- till the 17th centur}’ the stewardship was hereditary in the Maxwells.® The changes in Nithsdale during the 14th century also worked towards an expansion of the Douglas family. The forfeited Comyns went out, and a marriage brought in the Douglases in the person of William the first Earl, who married a daughter of the Earl of Mar, to whom had come the lordship of Nithsdale. The ancient i\lar line thus failing, the Earls of Douglas enjoyed both that title and the lordship."^ The second Earl, the James Douglas who fell at Otterburn in 1388, left two sons, both illegitimate, of whom William the elder was provided with the barony of Drumlanrig.® Archibald, lord of Galloway and third Earl, conferred the lord- ship of Nithsdale upon his second son, who left one daughter as issue. She married the Earl of Orkney, and so brought Nithsdale to the Sinclairs ; but in 1455 James II. secured a surrender of the lordship by the Earl, as well as of his hereditary office of sheriff of Dumfries, for compensation elsewhere. The Comyn barony of Dal- svinton fell to Walter Stewart of the Galloway family,® and Duncow to a Boyd.'^ Dalswinton remained with the Stewarts till the 17th century, when it passed from the Earl of Galloway to the Earl of Queensberry. After the forfeiture of the Boyds in the middle of the 15th century Duncow is found in possession of a Maxwell. Another family now came into prominence in upper Nithsdale. William de Crichton, of a Midlothian stock, had married the heiress of the Roos or Ross line who held half the barony of Sanquhar ; the other half he acquired by purchase. The seat of these Rosses was probably at Ryehill, by which they were sometimes dis¬ tinguished,® and where there is a mote-hill (No. 556). Later there are Crichtons in RyehiU. William de Crichton’s great-grandson was in 1485 created Lord Crichton of Sanquhar. The Crichtons, too, benefited by some of the Douglas property when that family came to grief, and continued in Sanquhar Castle and barony till the early part of the 17th century. In 1617 William Crichton entertained James VI. lavishly in the Castle, and in 1633 was created Earl of Dumfries. These succes¬ sive honours proved too much for the estate, and in 1639 it was sold to the first Earl of Queensberry. The adjacent barony of Morton was in 1440 granted by James II. to James Douglas of Dalkeith, afterwards Earl of Morton, though deriving his title from ^ Reg. Mag. Sig., i., App. i., No. 34. ^ Ibid., No. 920. 3 Johnstone MSS., p. 10 (Hist. MSS. Comm., xv., App. ix.). ^ Reg. Mag. Sig., i. p. 647. 5 Buccleuch and Queensberry MSS., p. 8 (Hist. MSS. Comm., xv., App. viii.). ® Reg. Mag. Sig., i., App. ii.. No. 323. Ibid., Nos. 306, 315. ® New Stat. Acct., iv. p. 306 n. XXVI INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. “ Mortoune ” in Midlothian.^ Another family of upper Nithsdale was that of Menzies in the baronies of Durisdeer and Enoch. About 1322 there is a charter by Robert I. to Alexander Menyers or Menzies of the lands of Durisdeer.^ Later, Durisdeer and the barony of “ Enache ” (No. 167), resigned by Alexander Menzies, are conferred on James Steward, brother of the High Steward,^ and Durisdeer remained with the Stewarts till near the close of the 17th century. It was otherwise with Enoch. In 1376 we have a grant of the barony to Robert, son of John de “ Meigners,” it having been held and resigned by the said John.^ Thereafter Enoch is possessed by a Menzies till the beginning of the i8th century, when it was sold to James Duke of Queens- berry.^ In the 15th century there was a Menzies in Dalveen,® and another in Castle- hill of Durisdeer.^ Dalveen was in time also to go to the Douglases. Of the minor families between Annan and Nith, that of Torthorwald suffers eclipse as a result of holding to the losing side. Like its neighbours, including the Bruces, it had attached itself to the English interest in the War of Independence ; unlike these, it had remained falsely true. Sir James de Torthorwald had fallen at Bannockburn a “ willing adherent ” of Edward IT, and John de Torthorwald, apparently his eldest son, became a pensioner of Edward III. in 1331.® Thomas de Torthorwald, however, the other son, who also had served the English interest, fought and died for David IT at Durham (1346), and his daughter and heiress, married to Robert de Corrie and personally enfeoffed by that king in the lands of Collin and Roucan, adjoining Torthorwald, died without issue in 1369.^ Meantime, King Robert 1 . had passed on the Torthorwald barony to Sir John Soulis,i° and, after his forfeiture in 1320, to Humphrey de Kirkpatrick.And with the Kirkpatricks of Closeburn the barony remained till some time at the close of the century, when it appears in the possession of Carlyles. The name of Carlyle—de Karliolo—has place among those familiar as witnesses to early 13th-century charters of the Bruces, such designations as Corrie, Herries, Jardine, Charteris (de Carnoto), Kirkpatrick ; all later to become surnames. The original settlement of the Carlyles was at Lockerbie (which they exchanged) and Kinmount (Kynemund).^^ William of Carlyle is styled “ Laird of Los ” by Thomas Randolph as lord of Annandale.^^ This William had married Margaret, sister of Robert Bruce. In 1432 we suddenly have record of William of Carlyle of Torthor¬ wald in a marriage contract with Sir Thomas of Kirkpatrick, lord of “ Killosbern.” By wiiat bridge the Carlyles entered upon the Torthorwald barony is not condescended upon. Sir John Carlyle w^as created Lord Carlyle of Torthorw^ald about 1475, but the direct male line ended in an heiress wiio brought the estate into the family of her Douglas husband. Sir James Douglas of Parkhead ; her eldest son was Lord Carlyle of Torthorw’ald in 1609. Finall}^ the property passed into the hands of the Queens- berry family in 1621-1622.1® No name is more common in the train of the Bruce lords of Annandale than that of Herries or de Heriz, and the title of Lord Herries, as a distinction acquired by a ^ Hist. MSS. Comm., xv., App., part viii. p. 36. » Reg. Mag. Sig., i. p. 517. 3 Ibid., p. 530. * Ibid., No. 585, p. 213. ® Dmmlanrig Castle and the Douglases, p. 93. ® Reg. Mag. Sig., ii. No. 765. ’ Ibid., No. 3492. ® Bain’s Calendar, iii. Nos. 1020, 958. * Reg. Mag. Sig., i. p. 613. Ibid., i. p. 517. ■ Ibid., p. 457. Buccleuch MSS., p. 39. ^3 Luce was an old parish now merged in Hoddom. “ Buccleuch MSS., p. 42. ^3 Ibid., p. 44. Ibid., pp. 43-44 xxvii HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. branch of the Maxwells, is conspicuous in one disturbed period of later Dumfriesshire history. The name was particularh’ associated with the estate of Hoddom, but the Herries holdings in the 14th century, wiien they were conferred, were principally those of Terregles and Kirkgunzeon, on the Kirkcudbright side of the Nith, though in the grants referred to they are included in the Sheriffdom of Dumfries.^ Hoddom we hrst hear of in 1199, when Robert de “ Hodelme ” is accused of having allied him¬ self at the siege of Carlisle with the King of Scotland (William the Lion, 1173 or 1174) against King Henry II., his lord for lands in England.^ In the plea over this affair we have mention of Robert's tw'o sons Udard and Randulph de “ Hodamme." ^ But heirs male ceased, for in 1257 we find Thomas de Lacelles, the husband of the daughter and heiress of Christiana daughter of “ Odard de Hodeholm," in possession of the English property’ in Cumberland in virtue of his wife’s inheritance from her mother.^ In 1292 we have Robert de Brus (the Competitor) in a successful lawsuit over the same English lands in association with his wife Christiana, heiress to her grandfather Odard of Hoddom,® both married for the second time. Adam de “ Hodolm ” appears on the Ragman’s Roll in 1296,® so that the Scottish property must have gone a different way. As w'e have seen above, Annandale came to the Douglases in 1409, and soon after that date Earl Archibald, in a charter now lost, gifted to Simon of Carruthers the lands of Hoddom among others," and in 1452 King James II. erected all these possessions of the Carruthers family, including Hoddom, into the barony of Carruthers.® The cessation of the Carruthers family after the middle of the i6th century brought about a fresh allotment of the Hoddom property. The next great reconstruction of Dumfriesshire territorial ownership followed on the suppression and forfeiture of the main branch of the Douglases in 1455. That famih’ was then planted to a greater or less extent in every dale of the county. But in a country as yet administered almost wholly on territorial lines, through the principal families, the extent of the possessions and power of the Douglas earldom was a menace to the Crown, and the Earls did not trouble to dissemble the fact. Indeed, this house occupied two other earldoms, brothers of the Earl of Douglas being Earls of Moray and Ormond. When the Earl himself was forced into England, these two opposed the forces of the Crown and were defeated at Arkinholm (Langholm) in 1455. The royal army itself was under the command of a Douglas, the Earl of Angus. The lo}’alists duly had their share of the extensive territorial spoils of the ruined earldom. Angus, among other things, had a gift of the lordship of Eskdale. The IMaxweUs, who had held for the Earl of Douglas the hereditary office of Steward of Annandale, received it now from the Crown, and likewise supplanted the Douglas in Xithsdale. Of the smaller folk the Beatsons profited most. John and Nicholas “ Bat3'SOune,” two brothers, had an hereditary grant of the five-mark lands of Dalbeth in upper Eskdale for their services at Arkinholm, while Robert “ Batysoune,” for the same reason, got Whiteshield.® Apparently anti-royalist sympathies in Dumfriesshire were confined to the Corries, who suffered accordingly. George Corrie of Corrie backed the Albany- Douglas raid upon Lochmaben in 1484, and was stripped of all his lands and possessions, of some, however, it would seem only for a time. The lands of Corrie 1 Reg. Mag. Sig., i. pp. 98, 615. Cf. p. xx. ^ Bain’s Calendar, i. No. 280. ® Ibid., No. 449. Ibid., No. 2101. 5 ii. p. 151. 6 p. 203. "• Bnccleuch MSS., p. 56. ® Ibid., p. 58. ® Exchequer Rolls, vi. p. 557; Reg. Mag. Sig. (1424-1513), Nos. 632, 633. xxviii INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. were given to Thomas Carruthers in that year.^ But early in the next century, we find Corrie in the hands of James Johnstone of Johnstone, who conferred it upon his second son Adam, whence the family of Johnstones of Corrie.^ The “Johnstone grey ” ^ was spreading over Annandale. Wamphray in the north, familiar from the ballad of “The Lads of Wamphray," was acquired by pur¬ chase in 1476 and given by Johnstone to a younger son.^ In 1536 we find that Newbie has been sold by George Corrie to William Johnstone of Gretna,® and Newbie, with other lands, was six years later erected into a free barony in favour of the same Johnstone.® Carruthers is an old place-name, and the family was in Dumfriesshire as early at least as the 13th century, but Carruthers passed out of their hands at some later period. In the 14th century (1315-21), a son of John of Carruthers received the lands of Mouswald (Musfald) and Applegarth (Appiltretwayt) from Robert Bruce.” Archibald Earl of Douglas in 1426 conferred Holmains, Little Dalton, etc., upon a son of the laird of Mouswald. Mickle Dalton and Dormont he had granted to his “ shield-bearer " Gilbert “ Greresoun " some years before. These lands another Gilbert Grierson sold in 1552.® The Murrays were descended from a sister of Thomas Randolph, and were destined to a peerage (Mansfield) in the 17th century, in which the family disappeared. Their hereditary lands were Cockpool, Comlongon, and Ruthwell, and from the Corrie estates they seem to have acquired Redkirk. The lands of Cockpool, “ Ruvale tenement," the tower and fortalice of Comlongon, Rainpatrick, and other estates comprised the barony of Cockpool on its erection in 1508.® Charteris (de Carnoto) of Amisfield goes back to the close of the 12th century. In September 1298 Edward 1 . granted to the Earl of Warwick the castle of “ Amesfeld " and land of Drungrey belonging to Andrew de Chartres.^® The family had lands also in the south of England, which were apparently restored to Andrew de Chartres on submission in 1304.^^ By June 1314 Andrew was dead, and the lands in “ Aldredestone in Wilts " had been forfeited by the rebellion of Robert de Chartres, his son and heir.^^ In the 15th century we have the emergence of the Border clans or “ surnames." In the Act of 1587 the clans of the West March are listed as Scotts of Ewesdale, Batesons or Beatsons (Eskdalemuir and Westerkirk), Littles (lower part of Upper Eskdale), Thomsons (Upper Eskdale), Glendinnings (Upper Eskdale, Wauchopedale), Irvings (Lower Annandale to Lower Eskdale), Bells (Kirtle Water), Carruthers [cf. p. xxviii.), Grahams (c/. p. xxxv.), Johnstones [cf. p. xxiv.), Jardines (Lower Annandale), Moffats (Black Esk), and Latimers or Lorimers (Upper Nithsdale). These conditions prevailed generally throughout the 15th and i6th centuries, with such modifications as befell in the natural course of things : as the failure of the line of Carruthers of Mouswald and the acquisition of the propert}^ by Douglas of Drumlanrig. In the 15th century, too, a branch of the latter family appears in Dalveen. The 17th century saw many of the old baronies, such as Torthorwald, Closeburn, Enoch, etc., acquired by the Douglases, while the barony of Langholm was sold by the second Earl of Nithsdale to the Duke of Buccleuch. These transactions, however, are out with our special interests. ^ Reg. Mag. Sig., ii. No. 1590. ^ Annandale Family Book, i., xxx. ^ The family livery. See “ Katherine Janfare ’’ in Minstrelsy of the Scottish Border. ^ History of the Johnstones, p. 9 ; Annandale Family Book, i., xxiii. “ Reg. Mag. Sig., hi. No. 1598. ® Ibid., No. 2570. ^ Ibid., i. No. 92. ® Hist. MSS. Comm., vi. pp. 710, 712. ^ Reg. Mag. Sig., ii. No. 3194. Bain’s Calendar, ii. No. 1009. Ibid., No. 1481. Ibid., hi. No. 366. xxix HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. IV. Relations with England : The Debateable Land. That the relics of the primitive folk should be traceable along the line of the watercourses, and that, to avoid the objectionable features of the forest and bogland on the levels, these pioneers should have occupied the higher and drier flanks, is what might be expected. This applies to peaceful penetration, which is necessarily a leisurely and scattered process. Hostile invasion follows a more beaten track. Viiether the Romans made their first entry into Caledonia by the west side is not certain, though usually affirmed. Birrenswark has given us the glandes (or acorn¬ shaped sling bolts of lead) which are peculiarly associated with Agricola's time. The Antonine Itinerary starts on this side at the station of Blatobulgium or Birrens, though this road-map is not necessarily complete. A prolongation of the road to the western extremity of the Vallum of Forth and Clyde seems inevitable. The camps at Gilnockie and Raeburnfoot by the White Esk raise another problem (see p. XX.). In aiw case, we are moving along the rivers. Even to-day travelling across countr}’ in Dumfriesshire is inconvenient ; the very railways reflect the north and south trend of the forces which have moulded the district. The mediaeval routes are scarcely in doubt. The main one at least ran from Annan to Lochmaben, thence westwards Tinwald way, and so by the side of the Nith to where the road forked, as it still does, between Tibbers and Morton, one fork going up by Durisdeer to the passes through the hills into Clydesdale, the other by Sanquhar into Ayrshire. The upper Annandale route by Moffat was also much used as the most direct way from the capital to the West March. It led to the head waters of the Tweed, and so to Peebles and the way to Edinburgh.^ From the succession of fortified sites along both sides we may infer that it was also a well-trodden prehistoric route. In number the sites exceed those in the upper part of the Nith valley. But in later military history, on any scale greater than a parochial feud, the Annan-Lochmaben-Nithsdale road was the main strategical feature of Dumfriesshire. Relatively to England there was also this fact, that on Dumfriesshire opened the western door past the mountain partition of the Cheviots. The Solway Firth on the one side and the hills on the other, with only the Esk as an ineffectual barrier, canalised all advances by land on this side from one country to the other. Thus, from the very outset of hostilities in the War of Independence, the cardinal position of Dumfriesshire became apparent. King John’s offensive opened, two days before King Edward crossed the Tweed, with a stroke as far as Carlisle. The Scots issued from Annandale and crossed the “ water of Sulewath ” at three places. They did a lot of mischief, but had to relinquish the siege of Carlisle and retire to Annandale.^ From the other side Annandale became a favourite raiding ground. Twice in the year of Stirling Bridge (1297) it suffered a foray from Carlisle ; the second occasion was a little before Christmas, and an improvised resistance of the natives brought about what is piously remembered as the battle of Annan, a local defeat. More than ten hamlets were burned within the range of a few parishes. Next spring Annan itself was spoiled and burned, church and all.^ ^ Cf. Letters and Papers, Henry VIII., vol. xviii. part ii. No. 237, for various routes from Carlisle to Edinburgh and Glasgow. 2 Hemingburgh, Chronicon, London, 1849, vol. ii. pp. 95-96. ^ Ibid., pp. 146-7. XXX INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. So far Annandale and Annan town had suffered for their proximity to the enemy base at Carlisle, and Annan was the only place worth spoiling till one came to Dumfries. But when the military occupation of the district began, a different use was made of the town. Edward I. returned from his victory at Falkirk through Annandale, coming over from Tibbers, and received the surrender of Lochmaben Castle. From this stage Lochmaben is recognised as the strategic centre of Dumfriesshire, and its accommodation is immediately extended and strengthened. Edward added a peel or palisaded bailey and erected a tower,i which enabled the place to hold out among the very last in the War of Independence. The later stone fortress was accounted, even in the i6th century, to be strong enough to withstand any assailants short of “ the hole armye of Scotlande.” ^ Possession of Lochmaben indeed was of cardinal importance to an enemy. It could be supplied in a short land journey from Annan, which in turn could draw upon the Cumberland ports, particularly Skimburness. It was a nodal point or junction of roads, of that up Annandale and the more prac¬ ticable route into Nithsdale. For between the lower Annan and the Nith dangerous mosses straitened the ways, particularly the Lochar Moss, the black heart of which still stretches northwards to beyond the town of Dumfries, and which must have been an even more formidable obstacle in olden days than it would prove now. It then effectively covered the approach to either Caerlaverock or the royal castle of Dumfries. To both there were but two possible roads. One was at the southern end of the moss from Annan by Cockpool and Bankend, but between the latter places it was carried over the moss on a narrow artificial bank, which could be cut and rendered impractic¬ able.^ The other was by Lochmaben and Locharbriggs,'* and this was clearly the safer and more usual way. The castle of Dumfries was also accessible from the sea by the River Nith.® Probably the mote of Castledykes on the Dumfries side and that of Troqueer in the Stewartry on the other, mark an ancient ferry,® as did the twin castles of York. But south and west of Caerlaverock even the sea was held at arm’s length by half-drowned and water-logged land.'^ Intrinsicall}^ the castle thus owed its importance to its strong defensive position, and its consequent capacity for annoy¬ ance to hostile neighbours. It was within easy striking distance of Lochmaben, and to have Lochmaben garrison in comfort Caerlaverock must be reduced.® Also a hostile force operating across the Nith in Galloway might be liable to its attentions. It is as an incident in such a campaign that Caerlaverock fell easily to Edward’s assault in the summer of 1300. At the port of Annan the oldest defensive post was the 12th-century mote-castle of the Bruces ; when a supplement to this was sought, it was found in the steeple of the church, and here, in 1299, Edward 1 . was having victuals stored against a possible attack by Robert Bruce.® In 1547 the steeple, which had but one storey above the basement, was regularly besieged, captured, and razed to the ground by an English force.More elaborate defensive works were undertaken by Lord Merries less than twenty years after. The ^'ear 1565 saw Annan equipped with “ a fair tower, able to ^ See Art. 443, and Bain’s Calendar, ii. No. 1112, p. 535. 2 .\rmstrong’s Liddesdale, App. lx.\. p. cxiii. ® State Papers, Henry VIII., vol. v. part iv. {contd.), p. 554. * Armstrong, App. Ixx. p. cx. ® Bain’s Calendar, iii. pp. 283-4 J Armstrong, App. Ixx. p. cx. ® Cf. Shirley’s Growth of a Scottish Burgh, p. 13. " State Papers, etc., v. part iv. {contd.), p. 554. ® Bain’s Calendar, ii. p. 535. “ Ibid., ii. p. 284. Calendar Scottish Papers, i. pp. 19-20. Cf. p. Ixiv. xxxi HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. receive above a hundred persons ‘ at ease,’ and forty or fifty horses.” From the town to the sea a ditch was dug, with but three places of passage, and another ditch landward to a moss, similarly restricted in its approaches. There were also, within two miles of the town, another tower to accommodate twent3^-fonr horsemen, and a high watch-tower with beacon and bell for warning. Such works put effective fetters upon both freebooting exploits from the west and sudden raids from across the Border —at least so far as the western districts were concerned.^ The strategic elements of Dumfriesshire, then, are obvious. They attach them¬ selves primarih' to the line of the lower Annan, where the cardinal point was the town of Annan. This is further made clear by the Act of 1481 for the garrisoning on that line of Lochmaben, Castlemilk, Bell’s Tower (? Kirkconnel Tower on the Kirtle), and Annan ; also b^" Lord Herries’s recommendation in 1579 W- P- “ strenthin the keipar dyke that environ- ettis the town of Annan” and “cast and strenthin the fuirds ” of the river as had been the “ancient ordour ” [cf. also Art. 89). The second but more vital line is that of the Nithsdale fortresses—Caer- laverock, Dumfries, Dalswinton, Tibbers, Durisdeer, and, it may be added, Morton. Connection with England could be main¬ tained through Dumfries and the estuary of the Nith, or more regularly, through Lochmaben and Annan to Skimburness. Lochmaben was thus the strategic nucleus of the defensive system, a fact abundantly illustrated by its history. The Niths¬ dale line seems curiously interdependent: it goes down either way as a whole. After his assassination of Comyn in 1306, Bruce seized Dumfries, Caerlaverock, Tibbers, and Durisdeer. The English king promptly set about their recapture. In 1309, out of about twenty-seven castles in English occupation, there are here Caerlaverock, Lochmaben, Dumfries, Dalswinton, Tibbers. By the close of 1313 probably all, and certainly the last three, had fallen to Bruce. Lochmaben was among the very last strongholds in Scotland to hold out for Edward 11 . Apart from Lochmaben, the most important positions, to judge from garrison figures, were Dumfries and Tibbers ; but the numbers in the former, as a base, fluctuate considerably from time to time. Local names crop up in the English accounts as in service on that side : in 1299 Sir Humphrey de Jardine and Sir William de Herez ; ^ in the garrisons of 1306 Thomas de Torthorwald, Hugh de Dalswinton, Thomas Bell in command at Tibbers and Robert Bell at Durisdeer. The same general principles characterise the fourth phase of the War of Independ¬ ence, namely, that of the resistance to Edward III., which covers the reign of David IT, and the results of which are prolonged down to the reign of the second James. It includes, however, a definite handing over to Edward III. by that transient king- figure, Edward Balliol, of a huge sector of Lowland Scotland, including the town, castle, and county of Dumfries. As in the earlier stages, too, the action of the local 1 Calendar Scottish Papers, ii. p. 155. ^ ggg Armstrong, App. Ixx. p. cxii. ® Bain, ii. No. 1115. xxxii INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. magnates displays some instability. Eustace Maxwell of Caerlaverock had been shifty in his loyalties while Robert Bruce was in the field. Now in 1332 he was a consenting witness to the coronation of Edward Balliol, and, when Dumfries became English, Maxwell as Sheriff attended to its royal revenues. But when Robert the Steward initiated the final and successful effort to throw off the English yoke. Maxwell, after receiving English munitions for Caerlaverock, reverted to his own country for a twelvemonth or so, and then returned to activity on the other side. His nephew and successor, Sir Herbert, carried on the tradition, and Caerlaverock, which he surrendered to England in 1347,^ remained with that country till it was captured by Sir Roger de Kirkpatrick in 1356. But the times were a hard test for territorial lords in a district so near the enemy. In March 1333 a successful English raiding party was opposed by Scots from Loch- maben garrison on its return “near Dornock at the Sandyford.”^ Thus befell the battle of Dornock, which resulted in the capture of the “ flower of the knighthood of the whole vale of Annan.’’ ^ Individual flowers were Sir Humphrey Jardine and William Carlyle. Yet was there a gallant remnant which no misfortune could bend to submission, namely, the brothers and other relatives of William de Carruthers, who, scattered and in great straits, lurking and wandering “ like wild men ’’ {tanquam silvestres), held out till the Steward in 1338 revived the national cause, and then, gathering like a swarm of bees [quasi examen apum congregantes), attached them¬ selves to that leader.^ Three years later the Earl of Moray, appointed by the Scottish Guardian to the custody of the West March, was able to make himself master of the open countr}^ and hold hostile movements in check.^ The defeat at Durham in 1346 and capture of David II. brought to Scotland another hard ten years. But, in 1356, while William Douglas recovered Galloway,. Roger Kirkpatrick did a similar service as regards Nithsdale, possessing himself of Caerlaverock and Dalswinton. John, son of the Steward, afterwards Robert HI., took the field in Annandale, and there remained till he had brought the whole district back to Scottish Allegiance.® All this meant the recovery of the castles, and particularly those on the Nithsdale line, which were credited with doing serious mischief to the English. Eor this reason David, as a condition of his release in 1357, had to promise their destruction, and so threw down Dalswinton, Dumfries, Morton, and Durisdeer, with nine others in Nithsdale." Lochmaben, however, as usual, remained last in English hands, and did not fall till 1384 (see p. 152). In all these activities the eastern dales make no special figure. They were not, however, indifferent. In the first year of the reign of King Edward Balliol (1332), “ Sir John de Lyndesey of Walghope knight ’’ forfeited his lands by “ rebellion,’’ that is, by supporting the regency.® Sir John de Orreton thus occupies Wauchopedale for a term, having his charter confirmed by the English King as late as 1340.® And Lindsay’s example had imitators. In the spring of 1337 Edward HI. was ordering investigation by juries of Roxburgh and Dumfries shires for discovery of the persons in “ Eskedale, Ledelesdale, Ewithesdale, Walughopdale, and Bretallaughe ’’ [i.e. Canonbie) who assisted “ the enemy,’’ that is, the nationalist Scots. ^ Bain, iii. No. 1507. 2 “ juxta villam de Drunnok apud Sandywathe,” Chronicon de Lanercost. ^ Scotichronicon, lib. xiii. cap. 27. “ The floure ... off the West Marche men ” is Wyntoun’s phrase. According to Bower and Wjmtoun, they were captured at Lochmaben. ■* Scotichronicon, xiii. 32. ^ Ibid., xiii. 48. ® Ibid., xiv. 15. ^ Ibid., xiv. 18. ® Bain, iii. No. 1354. ® Ibid., No. 1328. Ibid., No. 1226. xxxiii HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. In tnith, however, war was never to be long a stranger to Border life ; if not national war, then the scrimmage of local feud or the foray of needy or robbed neigh¬ bours. Constant reminders of the possibilities of the situation were the lines of beacon stations for warning the inner countr}’ of the approach of invaders, one fol¬ lowing a succession of heights up Annandale, the other up Nithsdale.^ Such a warning in the late autumn of 1448 - ma}’ have brought out the force that stopped a Percy raiding column from Northumberland at Gretna between the Sark and Esk, when Douglas and his brothers the Earls of Moray and Ormond were leaders, and the English were thrust back into the rising tide of the Solway. This failure was an endeavour to inflict reprisals for a Douglas raid as far as Alnwick the year before. Another example of what the Border country was specially exposed to suffer is in the raiding activity of English columns for many months after Flodden. In May 1514 Lord Dacre completes his report of destruction thus ; “ And upon the West Marchies of Scotland, I half burnt and distroyed the town- shipps of Annand, Dronok, Dronokwod, Tordoff, Eyshegewghe, Stokes, Estrige, Rye- lande, Blawetwood, Eoulsyke, Westhill, Berghe, Rigge, Stapilton, Wodhall, Rayn- patrike, Woddishill, Overbrotts, Nethirbrotts, Elistrige, Caluertsholme, Beltemmount, Hole, Kirkpatrike, Hyrdhill, Mossesyde, Stakehughe, Bromeholme, Walghopp, Walghopdale, Baggra3'e, Murtholme, Langhane, Grymesley, and the Watter of Esk, fro Stabulgorton downe to Cannonby, beyng vi myle in lienth. Where as there was, in all t3'mes passed, cccc‘'' pleughes, and above; whiche er now Merely waisted, and noo man duelling in any of them, at this daye ; saue oonly in the Towrys of Annand, Stepill, and Walghopp.” ^ These episodes were connected with national policy as a whole, but Dumfriesshire had a standing source of trouble of its own in the Debateable Land, to which reference has alread3^ been made (see p. xviii.). The understanding as to this piece of territory was that there should be nothing that could be interpreted as permanent occupation b3" subjects of either country. With the special exception of the priory of Canonby and its dependencies, it was to be a Border waste. Settlements upon it might be raided with impunity, but clearly there was here an opening for difference of judgment. Thus, on 23rd June 1517, some leading Dumfriesshire gentry, including the brother of Lord Maxwell the Warden, the lairds of “ Hempesfielde ” (Charteris of Amisfield), Twnwald (Tinwald, Maxwell), Ross and Holmeendes (Carruthers), and John Irwen ^ Ac(s Pari. Scot., i. 716, a.d. 1448 : “ Item it is fundin statut and vsitintymeof werfar anentisbailis birning and keping for earning of ane Inglis oist in Scotland, ther sal ane baill be brynt on Trailtrow hill; and ane uther on the Panchnat (Panteth) hill; ane on the Bailze (Bailie) hill, abone the Holmendis; ane on the Coldanis (Cowdens), abone Castelmylk; ane on Quhitwewin (Whitwollin), in Drivisdaill; ane on the Burane Skentoun {see Art. 18), in Apilgarth perochin ; ane on the Browane (Brown) hill; and ane on the Bleise, in the tenement of Wamfray ; ane on Kyndilknok (Kinnelknock), in the tenement of Johnestoune ; ane on the Gallowhill, in Moffet perochin : and syne in Nyddisdaill, ane on the Wardlaw ; ane on Rahothtoun (PTrohoughton) ; -ane on Barlouch (Beacon Hill) ; ane on the Pantwa hill (same as above) ; ane on the Malowhill (Art. 339) ; ane on Corswyntoun (Corsincon) ; ane on Crwfell (Crufell, Sanquhar) ; ane onrthe fell abone the Dowlwerk (? Dowlarg) ; and ane on the Watchfell. And to ger thir balls be kepit and maid the shiref of Nyddisdaill and the Stewart of Ananderdaill, and the Stewart of Kirkcudbricht, in Gallowai, salbe dettouris, and quhasa kepis nocht the balls ordinance and statut beand maid in tym of werfar sal pay for ilk defalt a merk.” Most of these sites can still be identified on the map, as given above from a paper by Dr George Neilson in Trans. Glasgow Arch. Soc. (1889-95), P- 356 - The watch tower on Panteth hill (Mouswald) was still identifiable as late as 1845 {New Stat. Acet., iv. p. 445). ^ On the date, see Scot. Hist. Rev., vol. ix. p. 197. 2 Pinkerton’s History of Scotland, vol. ii. p. 462 ; cf. Letters and Papers, Henry VHI., vol. i. p. 806. xxxiv INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. (Irving) with his clan, entered the Debateable Land at ii a.m. and carried off seven hundred English cattle. The retort of Lord Maxwell to an English remonstrance was that the English had “ sett stob and staik ” in the ground, and so begun an effective occupation for their own country in violation of the above conditions. The case came up again in 1522 for consideration at a meeting of Commissioners representing both countries, when it was denied on behalf of the English that any settlement had been made. Moreover, even if dwellings had been erected, that did not justify the taking of the cattle in the daytime, it being permissible only to burn the houses and take any men or goods within. Everything outside was inviolable except during the night hours.^ The case seems to have dragged on with new pleas and no definite result save a suggested division of the masterless territory. More serious was the migration of Grahams and Armstrongs into the Debateable Land. A family of Grahams, banished from Scotland about 1516, settled on the English side of the Esk {cf. p. xxxviii.). Henry VIIL, for service done, gave the eldest son “ good lands,” and the head of the family or “ chief ” was established at Netherb}^ They were allied with and favoured by neighbouring great men, associated them¬ selves with other Grahams on the Sark and Leven and intermarried freely with their neighbours, the Armstrongs. Certain of them in time received lands in Scotland from Lord Maxwell, while others had pensions from noblemen in Scotland “ for service done and to, be done ” ; just as certain of the Armstrongs became pensioners of Henry VIIL and got lands in Cumberland.^ The Armstrongs were a Liddesdale clan, with the head of the family at Mangertoun, but apparently had grown too numerous for their share of Liddesdale. Parties of them, during the i6th century, migrated into the neighbouring river valleys of Ewes, Esk, and Wauchope, and one company settled in the northern part of the untenanted Debateable Land, thus becoming neighbours to the Grahams. These two clans were “ well matched for a pair of quiet ones ” ; for both, in view of what is said above, it was a necessity that they should hold their position by force ; and their opportunist politics and irregular habits within a disputed territory were a main source of local and international trouble on the West March {cf. p. xxxviii.). Now in 1518 we hear that the Armstrongs “ ar in the Debateable landis and agreit with Ingland, and kepis there markat daily in Ingland.” ^ Ten years later three Armstrongs—John, Simon, and Thomas—each called ” the laird,” and two others have erected their towers in the district. Lord Dacre, the English Warden, was accused of conniving at the Scots settlements.'^ This encroachment, however, was apparently not agreeable to higher quarters, and Dacre undertook an expedi¬ tion against them. The Armstrongs were warned, and Dacre suffered a humiliating repulse, though he succeeded in burning “ ane place called the Holehouse,” which was apparently the tower by the river Esk, now called the “ Hollows ” (No. 43). On the same day the Armstrongs made a counterstroke to Netherb}'' and worked their will there. When Dacre demanded compensation for the Netherb}" raid. Lord Maxwell presented a contra account in the burning of the Hollows. Dacre pointed out that the “ Holl house ” was an illegal residence, since it was in the Debateable Land : ® Maxwell insisted that it was within the lordship of Eskdale. A second descent by Dacre was more successful. He completed the burning of the houses, ^ Armstrong’s Liddesdale, p. 215. - Border Papers, ii. Appendix. 3 Act. Doni. Con., cited in Armstrong, p. 211. * Letters and Papers, iv. part ii. No. 4420. ® Ibid., No. 4014. XXXV HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. but had to have axes used to cut down Ill Will Armstrong’s “ strong peel.” ^ Once more Canonbie was left isolated in the waste. These proceedings were followed throughout the year by a series of forays upon the \\'est March of England, in the course of which the land between Esk and Leven was cleared of its inhabitants and made to resemble the Debateable Land. The Armstrongs certainly thought they had grievances. Their appearance in the Debateable Land was not resented by the English Warden, and he allowed them to make use of Carlisle market. Then, under instructions from headquarters,^ he reversed his attitude, and the Armstrongs retaliated in kind. Border politics indeed had special local complexities. It was in the interests of both nations that the rieving practices, the seizure of cattle, the burning of houses and barns, and the sla3dng of men—though this last normally a regrettable necessity ^ —should cease, and "from time to time both Governments addressed themselves rigorously to this cause. No small obstacle was the fact that Lord Maxwell, the W’arden on the west, had taken the Armstrongs of Eskdale under his patronage, and had John Armstrong ” the laird ” as a tenant. Similarly, Lord Dacre on his side had been accused of being too complacent to the same clan. When James V. got rid of the Douglas control, one of his first tasks was to deal with the evil condi¬ tions on his frontier, urged thereto both by the defiance of all authority and by the complaints of Henry VIII. In 1529 James took a straight course to the seat of trouble ; he began b}^ committing to ward the principal Border lords, including IMaxwell, Johnstone, and Drumlanrig (Douglas) from Dumfriesshire, and then summarily hanged a company of the leading Armstrongs, who came to meet him, on the trees at Caerlanrig between Hawick and Langholm. Among them was John Armstrong of Gilnockie, the ballad hero of the Debateable Land. Of the Liddesdale Armstrongs it was reported to King James the year before that they had boasted “ thay woolde not be ordoured, naither by the King of Scottes thair Soveraine Lorde, nor by the King of Einglande, but after suche maner as thair faders have used before thayme,” likewise that they had been the destruction of fifty-two parish churches in Scotland, besides what they had done in England.^ Such sharp lessons, however, proved to be only a pruning of the mischief, not an uprooting. The problem of the Debateable Land as a refuge for “male¬ factors and trespassers ” remained.® Two years after James’s “ Jethart justice ” on the Armstrongs, Charteris, the Laird of Amisfield, an important figure of the time both locally and about the Court, was approaching Lord Dacre with the proposal that the English Warden should join with Lord Maxwell in “ th’ distroying of th’ inhabitantes of the Debateable ground.” ® Further, too, there was difference on the question whether Canonbie was debateable or not, which was argued at length between the two kings.® These and other Border difficulties were, however, incidental to the main forces of estrangement developing between James V. and his uncle of England. But when war did come in the autumn of 1542, it was mainly a Border affair of forays great and small. November saw several provocative raids in Dumfriesshire as far as Hutton, some miles beyond Lochmaben, one way, and up to Staplegordon on the other.® The principal military object in these operations was burning: mere ^ Letters and Papers, iv. part ii. p. 1828. ® See Bishop Leslie, De Origine, etc., Scotorum. ^ Ibid., V. 4, p. 107. ■ Canaby, Canabe, as so pronounced. ® Hamilton Papers, i. p. Ixviii. 2 Armstrong, App. xxii. p. 251. State Papers, iv. 4, p. 555. ® Ibid., iv. 4, p. 608 n. ® State Papers, iv. 4, pp. 579 ff. xxxvi INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. miscellaneous looting was not worth specifying in detail.^ On 23rd November Lord Wharton, the Warden, burnt Middlebie “ standing nere a strenghe of wood, and on his return “ turved houses ” and corn along the Kirtle. It was a misty morning and he was unable to carry out his full programme. But he was aware that the very same night a Scottish army lay in two divisions at Langholm and Morton Kirk.‘^ James himself went on to Lochmaben, and next morning, from the top of Birrens- wark, marked the progress of his army by the burning houses of the Grahams in the Debateable Land. But a force of horse and archers under Wharton kept in touch with the returning Scots until they came to “ Artureth myln dam, where a strate ford is which is called Sandyforde, having a grete mosse, a grete standyng water and the rever of Heske was afore theym and the mosse upon there left hande.” A fresh onset at this place broke the Scots : twenty were slain, many drowned, and over twelve hundred taken prisoners. Such was the miserable affair of “ Solium,” or, as it is adapted, ‘‘ Solway ” Moss. Lord Maxwell was among the prisoners.^ The success at Solway Moss opened a wider door for Henry’s schemes for the reduction of Scotland to the status of a vassal kingdom. But all that belongs to general history ; here only the effects on Dumfriesshire are to be considered. These took largely the form of Warden’s raids, ” which is to goo and cum in a day ande a night.” ^ The character of these exploits, with burning and the driving of cattle as their chief features, is sufficiently understood. They composed a policy of frightfulness. Annan was accounted in summer a Warden’s raid, so that its ashes were rarely allowed to cool for long. In February 1544 it was more “ surely burnt ” than ever, being, as was said, the ” cheif town in Anendaill unto Dumfries.” ^ Torduff, Dornock, and some other places were embers in the local conflagration. Dumfries suffered about the same time. But conquest was not to be made by burning towns and hamlets. To that end it was necessary to secure the principal castles of the district. At Langholm there was now a tower, and as this stood at the junction of Esk, Ewes, and Wauchope, it was of much strategical importance. It was betrayed to the English towards the close of 1544. Even more effective, however, would be the possession of Caer- laverock and Lochmaben; and Lord Maxwell and his eldest son, being captives, were worked on to hand over these places. Lochmaben, which the Maxwells held as constables of the castle, does not seem to have been secured, but in 1545 iMaxwell struck his bargain for Caerlaverock, and a small English garrison was thrown into it. The garrison was at once blockaded by the Laird of Johnstone and some Borderers ; ® no assistance could be given, and soon after it was again in Scottish hands. Whether realh’ so or not, it was urged by iVIaxwell that Caerlaverock was a stronger place than Lochmaben, and more suitable for a garrison. Such an opinion was probably intended as a blind. Certainly Caerlaverock had great natural ad¬ vantages : it took in “a great strenght of crikes and moss and but one way to come to it.” ' To take it by force or relieve it was no light matter. The narrow, direct road over Lochar Moss offered too many risks, and could be cut (seep. xxxi.). The only alternative was to go round b}’ Dumfries. The English administration, ^ Hamilton Papers, i. p. Ixi.x. - Ibid., p. l.xxx. ^ ibid., p. Ixxxv. No. 240. ^ Calendar Scottish Papers, i. No. 44. s Hamilton Papers, ii. p. 2S1. ® State Papers, v. part iv. p. 552. ’ Ibid., pp. 543, 557. XXXVll d HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. looking at the map, thought the place could be reached by sea. Wharton had to explain that save in exceptional circumstances this was impossible. But the English method at this stage was not that of the Edwards. It was not to make Dumfriesshire a safe approach for a regular military subjection of Scot¬ land, but to get astride the district and appropriate it piecemeal. Hence the anxiety for the Maxwell holds, the acquisition of Langholm Tower, whose significance was purely local, and in 1547 the seizure of Lochwood Tower, the seat of the Laird of Johnstone (see No. 315). This year, 1547, indeed saw the culmination of their local success. Dumfries¬ shire, on paper, seems wholly English. The chief local lairds and the local clans are in sworn allegiance to Edward VI. But Caerlaverock and Lochmaben were still Scottish, though likely to suffer from lack of victuals.^ In July Langholm Tower had fallen to the guns of the Scottish regent. Castlemilk, however, had been surrendered to the English by James Stewart, its captain, and a Graham was put in as commander (September). Lochwood, as we have seen, had an English garrison. But there was no permanency in this transformation. In 1550 peace was made, and two years later the constant friction over the Debateable Land was ended by its division. At that time there were still accounted to be “ bounde and sworne to serve the Kt’nges majeste” in Dumfriesshire Beatsons, Thomsons, Glendinnings, and Littles of Eskdale, Ewesdale, and Wauchopedale, “ and surnames under them,” to the number of 304 ; Johnstones of Gretna, 6 ; Bells of Middlebie, 104 ; Jardines and Moffats, 55.^ But soon history, as it affected the two countries, was to' have its course violently deflected, and other aims and activities were to come to the surface. V. Foray and Feud : The Wardenship.^ By the middle of the i6th centur}^ to the more settled and law-abiding section of the Scottish population, the Borderers in general were simply thieves. No literary glamour had as yet been thrown over their high-handed and irresponsible life. In the nature of the case, too, as time went on and efforts to exercise control over the state of things on the Border continued to take the form merely of spasmodic outbursts of legalised violence, things could only grow worse. Lord Herries in his Discourse and Advise on the Evil Estate present of the West Marches, presented to James VI. in 1579, traces all the trouble to the intrusion of the Grahams on the waste ground of the frontier {cf. p. xxxv.), their support by England, the failure of James V. to suppress them, their consequent increase in wealth and numbers, and their alliance and intermarriages with neighbours of similar character on the Esk, Leven, and Sark. At the time of the death of James V. they were not more than twenty or thirty at most; now with their “ assisters ” they numbered between three and four hundred ready to take the field on horseback at an hour’s warning, and ^ Calendar Scottish Papers, i. p. 20. 2 p. igi. ® The Warden had a deputy, and there was a Captain of Langholm with a company in the castle there (No. 429), who was also known as the Keeper of Annandale, like the Keeper of Liddesdale— another specially troublesome district. There was also a Sheriff at Dumfries with control over certain royal tenants and subject to the Warden’s orders; but he was rarely employed. The deputy and the Captain of Langholm w'ere the principal officers. {Border Papers, i. pp. 393-5.) xxxviii INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. leading lives of idleness and plundering. These thieves, with the assistance of Englisli thieves, had slain the principal Scottish barons nearest the frontier—Lord Carlyle, and the lairds of Mouswald, Kirkmichael, Kirkconnel, and Logan (Annan- dale), with “ many uther sober landit men ”—entering upon and occupying the greater parts of their lands, and so reducing the law-abiding part of the population on whom the Warden could call for support. Twenty years later a report by Sir William Bowes views the situation from another angle. And both memoranda, of course, are applicable to a long-standing state of affairs. According to Bowes, that state, along the whole Border, was one of “ winter war ” by opposite garrisons, the garrisons being the “ riding surnames ” or clans who lived on other men's goods. Thus “ contignuall intercourse of winninge and losing of goods do ebb and flowe like the sea.” The losers had either to steal in turn or fall into poverty. “ Wherefore may be gathered this strange conclusion that, where suche an opposite neighbour is founde, nothing is more pernicious to a frontier then is, in the commander, peaceable justice, and, in the obeyer, patient innocencie.” ^ In other words, the whole weight of advantage was upon violence and brigandage. Nor was the course of affairs in the country as a whole during the second half of the century favourable to peaceable occupations. The Reformation was not consummated without blood. The tension did not relax under Queen Mary : intrigue, rebellion, assassination, and hnally civil war maintained the atmosphere of disorderliness. Outside the town of Dumfries the ecclesiastical element in the Reformation was scarcely likely to make appeal ; on the lay or political side landed leadership determined allegiance. Here the most powerful personality was that of Sir John Maxwell, who in 1566 was created the first Lord Herries, and for the present outshone his nephew, the eighth Lord Maxwell, as yet a minor. At first Sir John showed active favour to the Protestant party ; after 1565 he is a Queen’s man, and her ablest. In thus identifying himself with the Queen’s cause Sir John was bring¬ ing himself into line with his neighbours, for among those who, at this crisis, had banded themselves for the Queen’s support were Lord Sanquhar, the Laird of John¬ stone and James Johnstone of Corrie, the Kirkpatricks of Closeburn and Kirkmichael, and Jardine of Applegarth.'^ Hereafter Dumfriesshire, town and country, is almost wholly on the Queen’s side. In the band at Hamilton, after her escape from Loch- leven in 1568, are the names of Lords Herries, Maxwell, and Sanquhar, the barons Johnstone, Closeburn, and Jardine, the Abbot of Holyrood, lairds Johnstone of Torry, Johnstone of Lochmaben, Crichton of Ryehill, and Murray of Cockpool ; ^ and the part which Herries played in her flight to England after the defeat at Lang- side is well known. The Regent ^Murray followed up his success by traversing Dumfriesshire, where so many of the Queen’s supporters were to be found. On i8th June he was at Dumfries, and received the surrender of divers Maxwells, Johnstones, Irvings, Grahams, and Bells, besides the strong-house of the Maxwells in the town. He followed a thousand fugitives to Hoddom, where only a show of resistance was made ; and he returned by way of Lochwood and Lochhouse, of both of which he took possession. This was known as the ” Hoddom Road.” ■* The Regent could count on only two supporters of standing in the whole district, Douglas of Drumlanrig, ^ Border Papers, ii. No. 508. ^ Keith's History of Affairs, etc., iii. (App., bk ii.), p. 2-19. * Calendar Scottish Papers, ii. No. 650 ; Keith, ii. p. 809. ^ Calderwood’s History of the Kirk, ii. p. 417. XXXIX HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. who was made \\’arden in place of Herries, and Jardine of Applegarth. In company with the Earl of iMorton, he therefore paid a visit in the following year, in order, b\' taking hostages, to secure a check upon the Border opposition. The critical period ensuing on the Regent Murray’s assassination in 1570 brought about the armed intervention of England, and the spring and early autumn of that \'ear each saw one of those unw’elcome visitations. On both occasions the objectives were the lands of the iMaxw'ells and the Murrays, as Mary’s conspicuous champions. In the first case. Lord Scrope, having burned Ecclefechan and the hamlets about Hoddom and Ruthwell, seems to have made for Dumfries by the Cockpool route, where he was attacked, unsuccessfully, by the young Lord Maxwell, but later held up at Locharwood a stage farther on. The Earl of Sussex, Lieutenant of the North, himself conducted the August expedition, in which he reported he had “ thrown down ” the castles of Annan and Hoddom, belonging to Herries, of Dumfries and Caerlaverock, which were Lord Maxwell’s, andTinwald (Tynhill) andCowhill (Coohill), which also were Maxwell houses.^ Sussex considered that, in the circumstances, he had acted vith great restraint, having refrained from indiscriminate plunder and burning ; but a complaint from the Scottish side to Queen Elizabeth accused him of just the contrar}^, as well as of the destruction of ten of the principal castles, two of which, Annan and Hoddom, “ were most strong.” ^ But the cause of the exiled Queen was foregone ; and, when the temporary union of the magnates in that cause dissolved, the local issues and problems reappeared in even fiercer guise. In the Memorial of 1579 cited above. Lord Herries sketches out roughl}" the character of the West March of Scotland in comparison with that of England. In the latter country, he says, the West March was planted with strong¬ holds even to the very frontier, strongholds including stone houses of every sort. Moreover, the soil there was fertile, the corn crops good, and the laws well obeyed ; in which last quality Herries is certainly exaggerating. As against this, “ Scotland upoun that Marche is ane pastour ground, verray barrane quhill (till) it cum far within the realme, and unproffitabill in a maner to the greit part, bot for bestiale ; as it is knawin ane man, to be sustenit honestlie upoun his stoir in lyk maner as his nychbour salbe sustenit with cornis, sail occupy mair ground nor ten tymes he that levis be the cornis dois in boundis ; swa that the West Marchis of England is meikle mair populus, and may, be the fertilitie of the ground, sustene mony ma men adjacent to the fronteiris upon that Marche nor Scotland may.” ^ That Dumfries was thus so largely a cattle country, made the industry of thieving more feasible ; grain is not mobile. It also, as Herries indicates, made particularly difficult the problem of a growing population. In such an atmosphere lordly jealousies and clan feuds flourished handsomely. It has been noted above how under these conditions the freebooting companies on the eastern side of the county had pushed their activities and forcible settlements be^^ond Annan, and had thus, too, weakened the power of the Government. For the Government here meant the Warden of the West March, an office practically mono¬ polised by the Maxwells. During the regency, however, they had been in opposition, and so from 1568 to 1573 Douglas of Drumlanrig was Warden. In 1574 the young eighth Lord Maxwell received the office, but the Earl of Angus intervened with a lieutenancy over all the Marches for about a year, 1577-78, when Maxwell was re- ^ Calendar Scottish Papers, iii. No. 436. ^ Ibid., No. 441. ® Register Privy Council, iii. p. 79. xl INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. appointed. But it was increasingly clear that things could not go on in this makeshift way, that some special effort should be made to root out Border lawlessness. It was in this interest that Merries tendered his proposals of 1579. Maxwell had protested to the Privy Council that he could not retain the office of Warden except on terms which would make him a petty sovereign in the district; “he would needs be absolute in these parts,” ^ and with this presented a roll of five hundred names of men deemed disobedient within his bounds. Merries proposed that the Warden should have his official residence at Lochmaben, that local lairds should, in times of disobedience, remain in the country with their households, that there should be a police force of twenty-four horsemen at Annan, and a captain with another force at Langholm, that Lochmaben and Annan should be adequately repaired. Repentance Tower properly equipped, the fortifications of Annan strengthened, the fords on the river provided with defences as formerly, and courts held in the Debateable Land.^ All this Lord Maxwell took very ill. Much of it presented itself to him as an encroachment upon his private rights. The custody of Lochmaben Castle, he pointed out, was a separate office having appropriate fees and duties, to which office he had a preferable claim, both because it had been long in his family and because it was the most convenient place for performing the duties as Steward of Annandale. To impose the keeping of a larger household as garrison at Langholm was to impose a special burden upon himself, who had, like other freeholders of the wardenry, pledged himself to the King for his lands and servants. Trailtrow Tower was a small matter, but, inasmuch as it was his private property, there seemed no reason why it should be put to a public use any more than the houses of others in the neighbourhood. A sug¬ gestion by Merries as to associating representatives of the Johnstones in certain matters of administration could scarcely be expected to meet with Maxwell’s approval ; but the mere suggestion shows in what quarter trouble was brewing. The proposals of Lord Merries certainly impressed the Council, and later wardens seem to have been appointed under “ the conditionis mentionat in the Lord Mereis buke.” ^ Meantime Lord Maxwell refused compromise, and Lord Merries himself resumed the wardenry of the West March, which he had just described as having “ bene evir the maist trubilsum part of the realm.” In line with his own suggestion, we now find the stewardship of Annandale associated with the office, and thus Lochmaben Castle was used to strengthen the Warden’s position. Merries held office only till the end of the summer, when Johnstone took his place. Two years later (1581) Johnstone was removed on the ground that he showed favour and gave protection to persons whom he ought to have punished.*^ Maxwell filled his place, but only for about a year, being discharged on the same ground as his predecessor, that of “ slothfulness ” in punishing offenders. The charge in each case was probably quite true ; neither Johnstone nor Maxwell was sufficiently devoted to the common weal to disregard the interests of friends, clansmen, and supporters. But such a charge was possible at any time. It was a Scottish rule in all departments of administration to show favour towards one’s kin and friends. Such occasions as these, when the charge was made a reason for the Warden’s remo\"al, were no doubt due to political factors arising out of the mau}^ changes of Government that kept in turmoil the minority of James VI. Maxwell was certainly unfortunate ^ Spottiswoode’s History of the Church of Scotland, vol. ii. p. 263. 2 Register Privy Council, iii. pp. 79-82. ^ p. 4 Ihid., p. 374. xli HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. in his relations with some of the regents. With the Earl of Morton he quarrelled over matters relating to that earldom, and Morton, who had made him Warden, now unmade him. But the Douglas earl ended his career on the scaffold in 1581, and a few months later the eighth Lord Maxwell became first (and only) Earl of Morton of his line as well as MMrden. Another change of Government led to his displacement from office, and the appointment of Johnstone in his room (1583). Maxwell’s re¬ sponse was to prohibit “ all his adherentes, tenantes and dependers to make him [i.e. Johnstone) answere or service as wardein.” ^ For a time this meant mainly trouble with England, since offenders against that country could rely upon Maxwell protection. Division between the two great groups, the lairds and clans in Nithsdale and Eskdale (Armstrongs) who attached themselves to Maxwell and those in Annan- dale who stood by Johnstone, was still racking the West March in the spring of 1585. Then more active hostilities began, which are typical of what took place under similar conditions at other times, and, as usual, involved almost every surname on the West Border. In April Robert Maxwell, brother of the Earl, with four hundred men struck at the Johnstone heart in Lochwood, slew some Johnstones, took others prisoners, and burnt the house of Lochwood.^ Towards the end of the month the same Robert with his Armstrongs burnt about eighty houses of Johnstone’s tenants and friends, after plundering them of cattle and furnishing.^ Contemporaneously the Johnstones had got to work. They burnt Duncow, but were driven off by Maxwells ; whereupon the Earl of IMorton (Maxwell) himself did some burning and spoiling as reprisals on the Johnstone bounds. As a separate adventure, on the same day, Robert Maxwell, with Armstrongs, Beatsons, Littles, and Carruthers, harried Dryfesdale and burnt part of Lockerbie, meeting with no resistance.^ Early in May the Earl tried to recover from the Irvings the “ stone house ” of Kirkconnel which had once been his own, and failing, with the loss of two men, next attacked the two “ stone houses ” of the Johnstones in Lockerbie, captured them, and hanged four of their defenders.® Sir John Johnstone (he had been knighted in 1584) was now on his way to Court to seek assistance against the rebel of his wardenry.® Later in the month Maxwell, with seventeen hundred men, horse and foot, marched rapidly from Dumfries to Moffat, where his horsemen made a sixteen-mile circuit, in which they burnt three hundred houses and carried off one thousand cattle, two thousand sheep, a hundred horses, and a store of household stuff; ' thus sacking the whole barony of Johnstone, where the tenants, we note, were “ baith Englessmen and Scottis.” ® In June Lochmaben Castle and Bonshaw Tower were being besieged by the triumphant Earl,^ while in July Johnstone fell upon the “ sheyles ” or shielings of some of the Maxwells and brought away two hundred head of cattle and sixty nags, killing but one man ; Maxwell’s people retorting with a lifting of eighty cattle from Johnstones.Things continued to go badly with the Johnstones. By August all the stone houses of strength on that border, with one exception, were in Maxwell’s possession, the Earl now maintaining in pay “ 200 horse and 300 ‘ shotte,’ besides the whole force of the country at his devotion,” while Johnstone was the “late warden” and “straitlye warded” by Maxwell in Caerlaverock.^^ In September the ” whole surname of the Johnstons” had yielded themselves to Maxwell, and Sir John was allowed to go free to meditate revenge : by November Maxwell was the new warden. ^ Calendar Border Papers, i. No. 153. 2 Border Papers, i. Nos. 303, 304. 2 Ibid., No. 308. ^ Ibid., No. 311. ^ Ibid., No. 312. ® Ibid., No. 316. ’ Ibid., No. 317. 8 Bist. MSS. Comm., Rep. XV. App. ix., Johnstone MSS., p. 32. 2 Border Papers, i. No. 321. Ibid., No. 327. Ibid., Nos. 340, 349. xlii INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. The Earl was now at the height of his power in the west, but at the same time entangling himself in the net of Catholic intrigue. He was a mark for political as well as personal enemies ; and these latter were not given to forgetting. In the spring of 1586 Sir John Johnstone was in the held and threatening the towns of Dumfries and Annan, from which he was beaten off by the weather.^ But he raided the powerful and “ well-beloved ” Sir Alexander Jardine of Applegarth, and subjected his houses and property to the usual outrage.'^ This was followed up by burning and spoiling of about a dozen hamlets of Maxwell tenantry, these “ poor commons ” having “ to paye for the sins of others.” ^ By May both Maxwell and Johnstone were under detention, but their friends were left to carry on. Thus in the hrst week of that same month. Merries, Drumlanrig, Amisheld (Empsheld), Apple- garth, and brother Robert Maxwell, with other allies, again harried the Johnstone lands in Annandale, burning Bonshawside and the Johnstone lands along the Nith, the Dryfe, and the upper Annan, killing two tenants only but bearing away a great booty.^ The Maxwells went even greater lengths, for in June Maxwells and Douglases were over the Border in England in order to get at the Grahams, who favoured their rival.® In June of the following year Johnstone® died, and Merries replaced Maxwell in the wardenry.'^ But Maxwell had bigger things to occupy his energies. The plotting of the powerful Catholic group in Scotland with Spain was growing more definite. In order to strengthen his position on the Border for eventualities. Maxwell was even earnest to let bygones be bygones and reconcile himself with the young chief of Johnstone.® He was preparing Jo facilitate a Spanish invasion of England through Scotland, and in the course of 1587 was in Spain on this business. The royal dealings with him were tender, because the King himself was not above suspicion of trafficking with the Catholic interest. But his bargain was finally made with England, and when Maxwell returned in the spring of 1588 to complete preparations. King James in person led a force into Dumfriesshire, secured the Maxwell castles, burning those of Langholm, Castlemilk, and Morton, and capturing that Lord himself. Only Lochmaben held out against Sir William Stewart for two days, when the garrison surrendered on promise of their lives. But James had the commander David Maxwell and hve of the leaders hanged on the plea that he had made no promise. Sir William having ” counterfoote his hand writt.” ® After this outburst a pause. King James made it a worthy hobby to reconcile family feuds, and his hand perhaps was behind the friendly approach of Maxwells and Johnstones in 1590.1® In 1592 Maxwell was once more Warden, though his activities on behalf of Spain were still proceeding, and on this account he was an object of suspicion to England, where his “ unaccustomed kyndnes ” to the Laird of Johnstone in 1593 was remarked, also his having two hundred men employed daily in fortifying Caerlaverock.^i But it was this year which was to change all. About twelve months before, the Wamphray Johnstones had raided the lands of Crichton and Drumlanrig, and there had been a tough struggle, with some loss in the retreat. It took all that time to bring the Crown and the Warden to see the desirability of visiting this offence upon the responsible chief, who was caution for the beha\’iour of his clan. To encourage Maxwell, the injured lairds, Drumlanrig, Sanquhar, ^ Border Papers, i. No. 418. ^ Ibid., No. 425. ' Reg. P.C., IV. p. 188. See Johnstone MSS., No. 68. “ Ibid., No. 419. ^ Ibid., No. 432. ® Border Papers, i. No. 462. Border Papers, i. No. 845. xliii ^ Ibid., No. 420. ® Annandale Family Book, i., ci. ® Caldenvood, iv. p. 679. HISTORICAL :\IOXUMEXTS (SCOTLAND) COMMISSION. and Lag. made a personal band with him of mutual support, and Maxwell with two thousand men proceeded towards Lockerbie. Johnstone, however, was aware of wiiat was coming, and had added to his immediate supporters his easterly friends, the Scotts of Ewesdale, the Grahams and Elliots of Esk. Johnstone, having an inferior force, pla3’ed a common Border ruse. A few horsemen pricked forward and drew a considerable body of Maxwell’s men in pursuit. These were received and driven back b^' a larger Johnstone bod^y in their retirement throwing their friends into confusion. Immediate!}- Johnstone flung his whole weight on his disordered foes and scattered them in flight with little loss. Such was the battle of Dryfesands, 7th December 1593, the last of the clan battles on the Border. Maxwell, “ a tall man and heavy in armour,” was killed.^ This, of course, w'as an outrageous deflance of public authority, and no Government could do less than put the offending Johnstone under ban for rebellion ; no Scottish Government in the circumstances could do any more. Again, therefore, the old feud blazed out more strongl}^ than ever. There is no need to dwell upon its incidents ; a sample of such has been afforded above, and there was little chance of variety. The onh’ notable point is that in these plundering raids the number slain was remarkably small, a fact wEich bears out Leslie’s comment upon such Border affairs ; aggressors were out for plunder rather than blood (see p. xxxvi.). Necessarily all the dales were implicated in this civil warfare. Maxw-^ell was supreme in Nithsdale and Eskdale, thus carr^’ing with him the upper Nithsdale lairds, notably Drumlanrig, while his “ friends ” the Carlyles and Bells and the town of Annan carried his interests into the lower Annan, and in Eskdale the Armstrongs were clients from of old. The John- stones covered Annandale from Lockerbie northwards, and their principal allies were the Irvings along the Sark, while they could draw upon the Moffats and Scotts of upper Eskdale, above the Armstrongs, and the Grahams in the lower portion of the Debateable Land. Meantime the Privy Council postponed decisive action in the Johnstone case, and accordingly in October 1595 that clan added another item to its calendar, when the Warden, Lord Herries, going “ to seke some of the John- stones at Lockerbie,” was driven off with loss.^ Obviously a Maxwell Warden could not hold his own on the West March, so the whole problem was characteristically solved b}^ the appointment of Johnstone as Warden in 1596. For such a course there was already a Maxwell precedent (see p. xliii.), but the step was little likely to mollify the Maxwells. And Johnstone, though not yet thirty years of age, had not less than twenty murders to his credit, both Scots and English.® None was so active on the Maxwell side as Drumlanrig, between whom and Johnstone a settlement was struck in 1597, only to be speedily broken, each accusing the other of perjury. A feature of the complaints here illustrates that procedure of forcible settlement referred to above (p. xxxix.). Carlyles and Bells entered upon some ^ Spottiswoode’s History, ii. p. 446. “ Adieu ! Drumlanrig, false wert aye. And Closeburn in a band ! The Laird of Lag, frae my father that fled, When the Johnston struck aff his hand.” (” Lord Maxwell’s Good-night,” in Minstrelsy of the Scottish Border.) As to the allegation that Maxwell’s right hand was struck off when raised for quarter, the historian says : “ I can affirm nothing.” In the Border Papers, i. (No. 918), Scrope explains that the fray was due to Maxwell’s attempting to cast down Mungo Johnstone’s house in Lockerbie. 2 Calderwood, v. p. 385. ^ Border Papers, ii. No. 485. xliv INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. of Johnstone’s lands and tilled and sowed them ; Bells had “ beaten the servants of the goodman of Bonshaw, taken their ploughs and forcibly tilled their land.” John¬ stone admits retaliatory proceedings, in which some of his “ puir folkishes coft pairt of thair awin geir bak agane.” ^ In a list of Johnstone outrages on Lord Sanquhar’s tenants in 1599 is the complaint of Janet M'Millan that they had burnt her house of ” twa hous {i.e. rooms) hicht with a laich hall,” etc.^ Already, however, despite the King’s weakness for him,^ the Privy Council had in 1598 denounced Sir James Johnstone of Dunskillie (as he is now generally entitled) for detestable slaughters and bloodshed, his slaying of Maxwells and “ honest men of the Sanquhar,” and his per¬ sistence in a ” maist wyld and bludie cours,” for all which he was put under sentence of outlawry, none to hold any communication with him.^ But by August 1600 the wild and bloody Johnstone was Warden again.^ Such being the conditions of life on the West March in the second half of the i6th century, one grasps the significance of its great equipment of all degrees of defensive dwellings from the castle proper to the mere stone house. Every man who could afford it found it incumbent to have some sort of dwelling not easily forced or inflammable from outside.® So are explained, too, the various reconstructions and rebuilding still in many cases dated upon such of the structures as survive. The next century saw the beginning of the great change which followed on the accession of King James to the throne of England. That a new spirit and a fuller power affected the Government was shown in the last act of the Johnstone-Maxwell feud. In 1607, at a prearranged meeting of the two heads at “ Achnanhill,” " which was to prepare a reconciliation, Johnstone was assassinated. But Lord Maxwell could not now outface the consequences. To avoid arrest he had to say “good-night” to Scotland: " Adieu ! Dumfries, my proper place. But and Caerlaverock fair ! ” In 1613 he ventured back to Scotland, was arrested, and suffered the death penalty at Edinburgh. Estates and honours were forfeited, but five years later restoration was initiated in favour of the heir, a younger brother, who, further, in virtue of the loss of the earldom of Morton, was in 1620 created Earl of Nithsdale. The Union of the kingdoms in 1603 obliterated the Borders in a political and administrative sense. They were now “ the verie hart of the cuntry.” ® Still, much in the way of special measures had to be taken, and it was a long time ere the peculiar features of Border life were completely eliminated. Some years before the Union the Government had come to the conclusion that one origin of Border malpractices was neglect of religion,® and initiated a movement towards the rebuilding of churches, which seem generally to have been in a ruinous condition. Then it was found that in certain quarters “ the povertie of the inhabitantis ” was so great that neither could kirks be repaired nor ministers supported unless adjacent parishes were united. Thus came about the uniting in 1609 of groups of parishes, served by a new church in a central position : Little Dalton, serving Meikle and Little Dalton and Mouswald ; Cummertrees and Trailtrow having a common church at Cummertrees ; Redkirk and Gretna at Gretna ; Kirkpatrick and Kirkconnell at Kirkconnell ; Middlebie, “ Tundersachs,” and Carruthers at the first named ; St. Mungo and Tundergarth at ^ Johnstone MSS., p. 37. Reg. P.C., vi. p. 115. ^ Border Papers, ii. No. 546. ^ Acta Pari., iv. p. 166 ; Birrel’s Diary, p. 46. ^ Border Papers, ii. No. 1231. ® See map on p. Ixiii. ; and cf. note on p. Ixii. " Spottiswoode, ii. p. 191. ® Reg. P.C., vi. p. 560. * Johnstone MSS., No. 87, p. 40. xlv HISTORIC AT. MONUMENTS (SCOTLAND) COMMISSION. Tundergarth ; Sibblebie and Applegarth at Applegarth ; Hoddom, Ecclefechan, and Luce at Hoddom “ near the towne thairof " ; Corrie and Hutton at Hutton—that parish still bears the double name.i The town of Annan was so “ miserablie impover- ischeit ” that it could not build a kirk of any kind as yet, and therefore was granted for that purpose “ the hous callit the castell of Annand, the hall and the towre thairof to serve for ane kirk.” - Too often the procedure on the Marches had been the reverse of this : a house of prayer turned into a castle. And not only had churches been neglected, but, in common with a tendency of the time, the more considerable lairds had given up the practice of living at their country-seats, and this too was accounted an encouragement of disorder ; wherefore in 1600 they had been instructed to repair and dwell in their residences in order to police the districts more effectually, Herries either at Hoddom or Lockerbie, Charterhouse of Amisfield in the house of Bent, Grierson of Lag either in Rockhall (Rochell) or Mouswald, Jardine of Applegarth in ” the hous of Speldingis,” etc.^ More direct measures followed the Union. The office of Warden disappeared, and a Commission of Scots and English Border gentlemen took in hand to compose the unsettled district. They had two companies of horsemen at their service as a police, and one of these was stationed at Hollows, the old Debateable Land and the Grahams there still retaining their character as the heart of the mischief. The Grahams were broken up by deportations. Complementary to the new churches w^as the novelty of specific gaols. By 1608 it was reported to the King, with reference to the services of the Earl of Dunbar, that he had purged the Borders, as Hercules did ‘‘ Augeas his escuries, by the cutting off by the sword of justice and your majesty’s authority and laws, the Laird of Tynwell, Maxwell, sindry Douglases, Johnstones, Jardines, Armstrongs, Betisons and such other.” ^ But this jubilation proved premature, and the Commission, modified from time to time, continued till 1625. The methods of “ justice ” were very similar in kind to those of the lawlessness against which they operated, being fire-raising, destruction of houses, eviction, and summary execution, M^ith the use of the feud feeling and a partiality to friends ; and the agents of the Commission found it necessary to secure the protection of frequent indemnities. One main factor in the change was the dissolution of the clan groups, which came as a consequence of the changed conditions. The gap which had opened between chief and clan fully showed itself in the Covenanting troubles, where the people in general, sooner or later, adhered to the cause of revolt; the lairds were royalist. Among the latter the Earl of Nithsdale was leader, and in 1640 Caer- laverock underwent its last siege and dismantling. Erom persecuting days, or rather later in time, the “ tombs of the martyrs ” in Nithsdale and Eskdale remain as memorials of a new enthusiasm ; among the most prominent of the Council’s agents were such familiar March names as Sir James Johnstone of Westerhall, Lieut.-Col. Douglas, brother of Drumlanrig, and Sir Robert Grierson of Lag. Meantime the Drumlanrig family was growing throughout this century in extent of lands and in dignity. Its outward symbol of territorial and political success was the great pile of Drumlanrig Castle, 1689, whose builder was the second Duke of Queensberry, the adroit statesman of the Union of 1707. Its purely domestic character signified that the days of fortified residences had for ever gone by. 2 Ibid., p. 441. ^ Cited, Hill Burton’s History, vol. vi. p. 19. xlvi ^ Acta Pari., iv. p. 441. 3 Reg. P.C., vi. pp. 154-5. INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. GENERAL INTRODUCTION. I. Archeological. Cairns.— The earliest monuments in this country, as has been frequently pointed out in the Introductions to previous Inventories, are the long cairns. These structures, containing one or more massive cists or chambers, were erected by the early inhabitants for the disposal of their illustrious dead, for it cannot be supposed that such were the burial-places for all and sundry. In treating of the archaeology of the Stewartry of Kirkcudbright, it was pointed out that the people who reared these impressive monuments had evidently, to judge by their distribution, approached the parts of the county to which such evidence of their presence is con¬ fined, from two directions : either from the Solway, passing up the valley of the Cree, or from the north, crossing what is now the Ayrshire border in the neighbour¬ hood of Carsphairn. The southern region is remote from Dumfriesshire, and the builders of the long cairns do not seem to have spread, by the evidence of the monuments, farther west than the lower reaches of the Cree. The northern district, however, in which these long cairns occur, lies much nearer to Dumfriesshire, and from it the earliest inhabitants appear to have penetrated into that county by way of Stroanfreggan, thence westward by Moniaive and Thornhill to the moorland region between Queensberry Hill and Annandale. Taking a breadth of a few miles to either side of this line, one will include probably all the cairns, whether long or round, in the county which appear to belong to the Neolithic period.^ Of long cairns there are at most four: the “White Cairn” (No. 249) at Fleuchlarg in Glencairn Parish, a cairn on Capenoch Moor (No. 329) in Keir Parish, the wholly reduced remains near Clonfeckle in Kirkmahoe (No. 351), and a cairn of smaller dimensions than any of the others on the moor near Stiddrig (No. 415) in the parish of Kirkpatrick-Juxta. As none of these cairns has been excavated, it is not possible to say what the form of the chamber may be. In this same region are several other cairns which, though not of the long type but circular, are yet of sufficient magnitude to render it probable that they too contain either chambers or megalithic cists of the transitional period between the Stone and Bronze Ages, such as the cairn at Stroanfreggan contained.^ When we turn to consider the cairns which indicate a purely Bronze Age origin we find that they have a wider distribution throughout the county, though they cannot be reckoned numerous in any district. They are sparsely distributed along the south from Mouswald to Robgill and Mossknow ; they are more scarce in the central region, and only practically in the tract of countr}^ in which are found the few long cairns are these cairns of the Bronze Age comparatively numerous. Towards the east side of the count}’, in the Eskdale and Ewesdale district, which being largely pastoral has probably suffered less by the dilapidation of its early monuments than the more highly cultivated districts, one is struck by the absence of such remains. There is a cist marking the site of a cairn on Bankhead Hill (No. 648) in the parish of Westerkirk in Eskdale, and the remains of a cairn still exist in a plantation at Sorbie Bridge (No. 222) in Ewes ; but beyond these, in ^ The later inclusion in the Inventory of what proved on investigation to be long cairns in the parish of Canonbie (No. 47) does not substantially modify these generalisations. It increases the total to six. ^ Inventory of Monuments in Stewartry of Kirkcudbright, No. 160. xlvii HISTORICAL :\IOXUMENTS (SCOTLAND) COMMISSION. the whole region north of Langholm no other significant remains are found. This points to the fact that, in the earlier prehistoric times, these two dales were but sparsely populated. Little recorded exploration has been made on the cairns in the county. Some years ago a cairn uas excavated near Auchencairn (No. 75) in Closeburn Parish, and remains of a drinking-cup urn of rather exceptional size, measuring when restored 10 inches in height, were found along with a flint implement of the type formerly designated a “ fabricator,” but now recognised as an object used with a piece of p\Tites for the purpose of producing fire. The urn and flint are now preserved in the National Museum of Antiquities. Another cairn was excavated at Mossknow (No. 371), in the parish of Kirkpatrick-Fleming, in 1908, and a cist was exposed the joints of which were plastered with clay and the bottom covered with river gravel, and which contained an unburnt burial but no grave goods. Stone Circles.—The one incontestable fact connected with stone circles is that in numerous instances they have been used as places of interment in the Bronze Age, as is proved by the finding of burials of this period within them. We may thus consider briefly the stone circles of Dumfriesshire in sequence to the cairns. The number remaining recognisable in the county is six. Of these not one is to be found in the regions where the Bronze Age cairns abound, and only those in Eskdale, to be afterwards mentioned, have the now recognisable site of a cairn an3'where near them, and that the remains of an isolated example; if their principal purpose was other than sepulchral, it is strange that in this county at all events the remains of stone circles should be most noticeable in regions where evidence of inhabitation in the Bronze Age is least discernible. The most remark¬ able circle in the county, both from the massive size of the monoliths which compose it, and the dimensions of the space which they enclose, is that known as the “ Twelve Apostles” at Hol^^wood (No. 284). This is situated in an agricultural district, and possibty in the process of clearing the ground and enclosing, existing cairns in the neighbourhood may have been swept away ; but the absence of cairns cannot be so easity accounted for in the case of the five remaining circles situated either on moor¬ land, as the circles at Kirkhill (No. 625) inWamphray Parish, on Whiteholm Rig (No. 603) in Tundergarth, and Whitcastles (No. 307) in Hutton and Corrie, or on meadowland in a purely pastoral region, as the “Girdle Stanes ” (No. 198) and “ Loupin Stanes ” in Eskdalemuir (No. 199). Another feature in regard to the four last-named circles which is worth consideration is the occurrence of two of them along the line of approach from lower Annandale to the upper waters of Eskdale, followed at the present da}^ by a main turnpike road ; while the other pair are situated close together and not far up Eskdale beyond the point where a branch from this road penetrates into the valley. The significance of this statement as indicating the line of approach of these Bronze Age people into Eskdale is increased by the fact that the remains of the only cairn observed in Eskdale, “ King Schaw’s Grave ” (No. 648), lie adjacent to where this road strikes the dale, on the opposite side of the Esk on Bankhead Hill. Though the fact of the lack of association in localities of cairns and stone circles in these cases is deemed worth drawing attention to, it should be stated that in the extreme south-east of the county, adjacent to the upper end of the Solway, there is a site of a stone circle (No. 5), some miles to the east of Annan, while the “ Lochmaben Stane” (boulders) (No. 263), near where the Kirtle Water joins the Esk, are probably the remains of another. The latter would be within 4 miles of the remains of a group of cairns at Mossknow. xlviii INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. Small Cairns and Hut Circles.— The recent discovery of the remains of Bronze Age vessels, seemingly of domestic type, in the centre of two hut circles in Ayr¬ shire,^ is irrefutable evidence that such sites were occupied during that period. But it does not follow that such inhabitation was confined to that period, and there can be little doubt that, if none of our hut circles has a Stone Age origin, some of them certainly are the foundations of dwellings of the Iron Age. The association of small cairns, beyond placing the remains in pre-Christian times, does not actually help us, for Iron Age sepulture occasionally took place in cairns ; but as more relics of the Bronze Age have been recovered from such erections, there is at least a presumption that the hut circles with associated small cairns belong to the latter epoch. There have been located in the county some thirty-one groups either of small cairns, such as are usually found in association with hut circles, by themselves, or of smaU cairns with the accompanying hut circles ; and the fact that there is not a single one of these groups in Eskdale or Ewesdale supports the inference which the distribution of the larger cairns leads up to, that in early prehistoric times the eastern districts of the county were very sparsely peopled. The groups lie in ten parishes; Closeburn, Dunscore, Glencairn, Keir, Kirkmahoe, Kirkpatrick-Juxta, Kirkmichael, Middlebie, Sanquhar, and Tynron. Eight of these parishes are in Nithsdale. Kirkpatrick-Juxta and Kirkmichael are in xVnnandale, but the groups in the former parish lie all on the west side of the Annan, and are on the moorland reaching back to Queensberry Hill, a region, as shown above, in which early cairns also occur in considerable numbers, while the two groups in Kirkmichael Parish lie on the high ground between the two dales, and are only a few miles distant from a group near Glenmaid in the Nithsdale parish of Kirkmahoe. There is another point to observe about these groups for which a satisfactory explanation has yet to be discovered, and that is the remarkable uniformity of elevation at which they are found. In the Inventory the approximate height above sea-level is given of twenty-nine out of the thirty-one groups, and an analysis of these statements yields the following results. No less than twenty-three of the groups lie at an altitude of between 800 and 900 feet, four between 700 and 800 feet, and only two below the 700-feet elevation. Many groups formerly existing at lower levels have doubtless been eliminated by the action of the plough, but, if the extension of agriculture in comparatively recent times afforded an explanation, we should expect to find those constructions which still remain situated at the edge of the moorland, which is by no means the case. This may be noticed particularly with regard to the small cairns at Knockespen in Kirkmichael Parish, from the position of which, high up on a long ridge, there stretches below a wide reach of moorland wiiich has never been broken in to the plough. The best-preserved examples of hut circles are those on Whitestanes Moor, Kirkmahoe Parish. They seem for the most part to be oval, and present features which did not occur in the hut circles of Galloway or of the northern counties, in that they have been dug out to such an extent that the present floor- level in the interior is sunk in one case as much as 11 to 2 feet below the natural surface on the outside. Similar pit dw'ellings were met with in Lauderdale, but not in association with small cairns.^ Rock Sculptures.— The limitation eastwards of rock sculptures in the Stewartry w^as remarked on in the Inventory of that county, and it is not surprising therefore, as all the evidence points to the populating of Galloway and western ^ Proc. Soc. Ant. Scot., xlviii. p. 376. ^ Berwickshire Inventory, p. 122, No. 231. xlix HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. Dumfriesshire having been done from the west, to find not a single example of cup- and-ring markings in this county. One stone only was met with on which occur markings probably belonging to this class, and that was a detached slab (fig. 2) form¬ ing the sill of the doorway into the vaulted basement of Hollows Tower (No. 43) in Canonbie Parish. In place of cup and rings, spirals are traced on it, and it bears a considerable resemblance to a stone from the Island of Eday, Orkney, found in what appears to have been a chambered cairn and now in the National Museum of Anti¬ quities, Edinburgh.! 7 Pe provenance of the stone at Hollows Tower is unknown. Defensive Constructions. —The remains in this county which fall to be considered under this heading number 220. These may be separated into two distinct classes, viz. those whose main purpose has been, by choice of situation and construction of defences, the prevention or repulsion of attack, and consequently are “ forts ” ; and others which, though possessing certain features of defence, combine with these elements of conceal¬ ment such as would in troublous times be applic¬ able in pastoral districts to shelters for sheep and cattle. These latter not being actu¬ ally forts have been desig¬ nated merely “ enclosures.” To the first class belong 143 constructions; to the second, 77. Taking a sur¬ vey from the west across the county, and commenc¬ ing with Nithsdale, we find in that region forts only, numbering 25; in Annan- dale we find 94 forts, but also 37 enclosures ; while in Eskdale and Ewesdale we meet with only 24 forts, but as many as 40 enclosures. In one or two cases all over classification may be doubtful. The lack of definite knowledge regarding the period of erection of the forts in this country, owing to the limited amount of excavation which has thus far been done on them, renders the synthesis of these structures in any manner which may be illuminating a matter of no little difficulty. The usual method of consider¬ ing them mainly according to the physical qualities of the sites they occupy, does not afford much help, for we have no reason to suppose that the people who occupied a promontory, if in their immediate neighbourhood, would not as readily have drawn their lines of defence in a geometrical figure around the crest of some swelling ridge had it been nearer at hand and equally suitable. There are, fortunately, a few out¬ standing facts which act as guide-posts along the ill-defined track which we have to follow in our endeavour to pick out and set in some order the relations of these constructions to the prehistoric periods. The Roman fort or camp is, with rare exceptions, in form a rectangular oblong with the corners rounded ; the camp of ^ Proc. Soc. Ant. Scot., iv. p. 186. , 1 Fig. 2.—Spiral-marked Slab, Hollows Tower. INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. mediaeval times, which also intrudes itself to intensify the confusion, is probably likewise straight-sided. The prehistoric fort is, on the other hand, as an invariable rule, curvilinear, either oval or circular or of some irregularly round hgure such as the eccentricities of the site demanded. The curvilinear forts, further, may be constructed of stone, with or without entrenchment, or they may be pure earthworks with ram¬ parts and ditch, or a combination of stone and earthwork ; and lastly, their ditches may be entirely excavated in soil, or their construction may have necessitated the cutting of rock. As the rectangular oblong fort, by its shape and certain considera¬ tions of situation, ■ may, with some degree of certainty be recognised as Roman, if the mediaeval element be disregarded, so the curvilinear fort, the trench of which is cut through rock, may be presumed to belong to the Iron Age. We are thus left with stoneworks and simple earthworks to assign to their proper periods. This is probably not possible by the aid of superhcial observation, as the builder of an Iron Age fort might be so fortunate as to find no rock to interfere with his entrenchment, or the erection of a stone fortress might be his simplest and most effective contrivance on a particular site which he deemed it necessary to occupy. To lift the veil and indicate the features which may be peculiar to any given period, spade-work is necessary, and until that is forthcoming the chronology of our numerous forts must in large measure remain in doubt. Environment, comparison with excavated examples, and occasional discoveries of associated relics are all factors which may be called in to help, and the cautious use of these may enable us to determine the period to which certain of the Dumfriesshire forts belong. First, as being least open to doubt, let us take the Roman forts into considera¬ tion. Of these there are four which by plan as well as by the positive evidence afforded by excavation are assuredly of this origin : Birrens (No. 462), two at Birrens- wark (No. 272), and Raeburnfoot (No. 172). Another, Gilnockie (No. 45), by the details of its plan and by analogy with ascertained examples elsewhere, there can be little doubt merits a similar attribution. But, after accounting for these five, we have still a number of rectilinear forts in the county concerning which only the ap¬ plication of the spade will suffice to determine whether they are mediaeval or Roman. Certain of their characteristics may be pointed out. One only of them, the fort at Kirkmahoe Manse (No. 340), is large, and it is in a very fragmentary condition. The site, a plateau flanked on one side by low marshy ground and on the other by a steep bank overlooking haugh-land that stretches to the Nith, is such as a Roman general might well have selected. But when we say that the site alone presents no inherent impossibility to such an attribution, it is as much as we are justified in asserting. Proceeding up Nithsdale for some miles to Durisdeer, in the glen of the Kirkburn, at the entrance to the Well-path, a pass which leads through wild hill countrv into Clydesdale, we find another small oblong rectangular earthwork of uncertain origin (No. 163). The trench which surrounds it is boldly cut, and for some distance its • course has been hewn through rock, which does not suggest that we have here a mere temporary encampment. The entrance is in the middle of one end, and some 24 feet in front of it there has been dug an outer ditch or traverse, a feature quite consonant with Roman methods. Here again the spade alone can decide, but, as iu the case of the fort at Kirkmahoe, there is no inherent impossibility of a Roman origin. Passing into Annandale, we find in the parish of ^Middlebie, at no great distance from Birrens, two constructions which deserve some notice. One at Purdomstown (No. 466), adjoining, and parth^ covered by, the Annan Waterworks, is a quasi- li HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. rectangular oblong contained in a loop of the Middlebie Burn, the high banks of which afford a considerable measure of defence. No entrance remains visible. The rampart is low and not ver\" broad, and beyond the rectilinear shape and something in the situation, there is nothing about the construction suggestive of a Roman origin. The second fort of this character referred to is more remarkable. It is situated near Carruthers on Birrens Hill (No. 464). In form it is oblong and approximately rectangular vith rounded angles. The rampart rises boldly from the interior plat, and the covering trench, where it remains uninterfered with by later works, has its scarps smooth and sharply cut. The entrance is through the centre of one end. In front the ground has been much interfered with by quarrying. Here again the features do not suggest a prehistoric origin, but whether this fort is Roman or mediaeval, excavation alone can decide. The probability is, however, from its exposed situation, that it belongs to the latter class. Proceeding up Annandale, at Gotterbie ^loor (No. 451) in Lochmaben Parish, we find another small oblong quasi-rectangular fort. The entrance in this case is not in the centre but towards one side of the south-east end, and, 20 feet in front of it extending divergently past it, is a deep irregular!}^ excavated hollow, more like a quarry-hole than a trench. The situa¬ tion of this construction in a depression of the ground, the slightness of its rampart, and the water-holding character of its ditch, all militate against the theory of a Roman origin. Some miles farther north, in the parish of Kirkpatrick-Juxta, is another small oblong rectangular fort, which merits more attention. This fort, near the farm of Milton, is situated on a ridge known as Tassie’s Height (No. 411), and is seemingl}^ adjacent to the site of a fort noted by Roy. It has been sur¬ rounded by a single rampart of earth of very considerable bulk, though now greatly spread by cultivation, and with a trench to the outside. The entrance has been through the centre of one end and has faced the site of an old road which leads up Annandale, and to which a tentative Roman attribution has in the past been assigned. The situation of this fort, commanding an extensive prospect from a moderate elevation, is such as the Roman engineers greatly affected ; moreover, the placing of the axis at right angles to that of the trend of the ridge, as if to face on to a road passing along it, is an arrangement quite unlike that adopted in native forts. Here again the spade alone can solve the problem. Turning to a consideration of the curvilinear forts, and commencing with Niths- dale, we cannot in this region recognise any arrangement of defences which we can point to as typical of the district or of any particular period. The principal forts have for the most part multiple defences of ditches and ramparts, and as a rule are earthworks. The most important fortress in the dale, and also in the county, is Tynron Boon (No. 609) in Tynron Parish, occupying the summit of an imposing peak. Its bold ramparts of earth and splintered rock, and the abundant evidence of rock-cutting in its lowest trench, indicate for it a late origin, presumably in the Iron Age. A similar characteristic marks the fine fort on Barr’s Hill (No. 581) in that part of Tinwald * Parish which may be reckoned for our purpose in Nithsdale. An earthwork which shows no resemblance to any other fort in the county is that crowning the Castle Hill (No. 236) above the Dal what Glen, some 3I miles westward of Moniaive. With its defending terraces, however, it distinctly recalls the fort overlooking the Laggan Loch in the parish of Glasserton, Wigtownshire.^ Another which displays unusual features is that on Morton Mains Hill (No. 511), Morton Parish. This fort has all the ^ Wigtownshire Inventory, No. 5. lii INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. appearance of an unfinished work. It is a pure earthwork and consists of various unconnected segments of ditch and rampart around a hill top. Stoneworks are not common among the forts of Nithsdale, but there are two, both in the parish of Kirkmahoe, which call for comment. The one at the Belt, High Townhead (No. 342), on a promontory overlooking the valley, is remarkable for the extent of its defences. It is comparatively small, measuring in the interior 163 feet by 109 feet, and its situation is a strong one. On its more assailable front it has seemingly been protected by three outer walls, and the entrance has been carried through these walls by a passage some 95 feet in length. There is an absence of trench work in its defences which may indicate an early period for its origin ; and the groups of small cairns and hut circles on Glenmaid Moor (No. 343), Whitestanes Moor (No. 344), and Shaws Moor (No. 345), at no great distance to the north, are evidence of the early occupation of the neigh¬ bourhood. The other stone fort, the Mullach (No. 339), occupies the summit of a pro¬ minent hill about i| miles to the north-west. It is the only vitrihed fort observed in the county. The two walls which enclose the enceinte are at a considerable dis¬ tance apart, and here also there is no entrenchment. The vitrifaction appears in both walls, and it is noteworthy, as bearing on the question of the production of that condition, that there is no trace of anything of the kind on the rocky summit which forms the centre of the fort, where it might have been expected, had signal hres been the accidental cause of vitrifaction in forts. It lies at a distance of 10 miles from the sea, which is somewhat unusual in the case of a fort of this class. In the valley of the Cairn, a tributary of the Nith, there lies, some seven miles below Moniaive at Snade, an earthwork of unusual character, which is probably late, but which does not fall into any other class of earthwork in the county. This is a circular plat known as “ The Orchard,” measuring some 116 feet by T03 feet, lying on low ground near the river defended by ditches and ramparts, the former of which are capable of being artificially flooded from the Cairn. The ditches are broad and deep, the ramparts massive, and the situation with its wet ditches seems to indicate a possible mediseval origin. Though this construction is much more imposing, it recalls the so-called Trowdale “ Mote ” in the parish of Crossmichael in the Stewartry of Kirkcudbright,^ situated in low-lying swampy ground and surrounded by two concentric ditches which appear to have held water. Of the ninety-four - forts of Annandale, we have alread}^ dealt with those of rectilinear plan, which number only seven. The remainder, following the principles which have been adopted in this survey, we may divide in the first case into stoneworks and earthworks. The former, it will be observed on reference to the Inventory, are almost entirely conhned to the parish of Kirkpatrick-Juxta, in the stretch of country bounded by the River Annan for a few miles southward from Beattock on the east, and by Oueensberry Hill on the west. And as in Nithsdale we found the only two stone-built forts adjacent to the region of the small cairns and hut circles, so here also this class of fort is situated in that portion of the county where large cairns are least scarce, and where in places the small cairns abound. The earthworks are not confined to any particular locality, nor do they exhibit any peculiarities of structure or of plan which distinguish them specially from forts found elsewhere. Considered according to the factors noted in reference to the Nithsdale forts, a considerable number show rock-cutting in their trenches, such, for example, as the forts of Range Castle (No. 98), Dalton Parish, Carthur Hill (No. 291), Hutton and Corrie Parish, ^ Kirkcud. Inv., No. 140. ^ A segment of another is reported at Greenhill plantation, Cummertrees. liii e HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. and Haggy Hill (No. 596) in Tundergartli Parish. Moreover, on the ramparts of those which show in their trenches what is presumed to be the application of an iron tool, there may also be observed along the erests of some of them the foundations of a stone parapet. Such is the case on the forts of Craighousesteads Hill (No. 600) and Haggy Hill (No. 596) in Tundergartli. Certain of the earthworks with bold ramparts and trenches, such as Woody Castle (No. 450) overlooking Lochmaben, or the im¬ pressive remains of the tripl3’-ramparted fort at Gallaberry, Dr^deholm (No. 115) in Drvfesdale, show no peculiarities of construction that enable the observer to hazard an opinion as to the period to which the\’ are referable ; the same may be said of the segmental earthworks at the edge of the banks of the Mollin Burn (No. 320), Johnstone Parish, and Auchencat Burn (No. 485), Moffat Parish. In the Stewartry of Kirkcud¬ bright a circular fort at Drumcoltran ^delded from the bottom of its surrounding trench some \’ears ago a hoard of Bronze Age rapier-blades, and, as far as the characteristics of that fort ma}" be regarded as typical, it affords a definite index for the identifica¬ tion of other forts of that period. Two forts in Annandale certainly present a superficial resemblance to it : these are the fort on Castlehill, Pilmuir Common (No. 113), in Dr5Tesdale, and that at Millbank (No. 14) in Applegarth Parish. Both are approximately circular, are surrounded by single ditches with earthen mounds on scarji and counterscarp, and are pure earthworks, features all possessed by the Drum¬ coltran fort. A peculiarity noticeable in a number of the forts, and almost universal in the enclosures, is the opening of the entrance into an excavated hollow on the interior, so as to be commanded by higher ground all round. This may be seen in the fort in Corncockle Plantation (No. 449), Lochmaben Parish, and in the fort on Newland Hill (No. 599) in Tundergartli Parish, which has previously been quoted as an example with a stone parapet above the rampart. A fort which seems to be unique in this district is that near Crawthat Cottage (No. 595) near the road from Lockerbie to Langholm, and also in Tundergarth Parish, its peculiarity being its division by a cross trench into two separately defensible areas. The Eskdale and Ewesdale region contains only some twenty-four forts, and of these twent}^ are situated above the junction of the Esk and Ewes. Eor the most part they lie in the Esk valley, clustering to the north of a point where the Black Esk coming from the west mingles with the White Esk from the north, both streams thereafter flowing on in an easterly direction. The most remarkable of the group is the fine fort of Castle O’er (No. 177). It occupies the crest of a long ridge, also a considerable area of ground below the eminence, and its defences, which are formid¬ able, combine wall, trenches, and ramparts. Hut circles are evident in the interior, and there is abundant evidence of rock-cutting. In various aspects it recalls the fort on the summit of Bonchester Hill in Roxburghshire,^ which shows a similar employment in its defensive system of walls, trenches, and ramparts. A further remarkable arrangement intensifies the analogy, that is the enclosing of an area of ground at the base of the eminence crowned by the fort. A slight excavation on the Bonchester Hill fort produced an iron shouldered pin which, along with the type of querns found, all of the saddle variety, suggested an early Iron Age date for the construction. It seems likely, therefore, without straining the analogy, that the Castle O’er fort originated in the same period. There are three features generally noticeable in these forts which link them to others noted in Annandale, and these are: rock-cutting in the formation of the trenches, ^ See Proc. Soc. Ant. Scot., xliv., 1909-10, p. 225. liv INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. remains of a stone parapet surmounting the ramparts, and the lowering of the level of the interior by excavation, usually at the entrance. As affording some sort of analogy to the fort at Crawthat Cottage in Tundergarth Parish, attention may be directed to the fort at Over Cassock, Eskdalemuir, which is unlike the generality of the forts in the dale, being formed on a promontory and separated into two divisions by a cross rampart and trench ; but in this case the upper enceinte into which the main entrance opens appears to be the more important enclosure, and not, as in the Crawthat Cottage fort, an outer bailey. A fort on the Loch Hill (No. 211) in Ewes Parish, though now much worn away, appears to be distinct from the general type in the neighbourhood. A remarkable construction (No. 175), which, though an earth¬ work, can hardly be deemed either a fort or an enclosure, is situated on the right bank of the Esk some | mile above the mansion-house of Castle O’er. It is a plat at the top of a steep bank rising from the river but at the bottom of a semicircular hollow in the hillside, that towers above it, and from which it is overlooked at all points. It has around it certain lines that are of a defensive character, but its purpose is inexplicable. A Roman fort at Raeburnfoot in Eskdale has already been noted, and its presence there may hnd a likely explanation in the group of forts referred to as indicating a considerable population, which the Romans may have found it necessary thus to overawe. Of stone-walled forts in the region of upper Eskdale or of Ewesdale there is not one. A single specimen, however, crowns the Craig Hill in Westerkirk Parish (No. 637), some 3 miles above Langholm. It now remains to consider those defensive constructions which we have classified under the name of “ Enclosures.” In form they are as a general rule circular, or oval, protected by a single rampart with a ditch in front of it, and having the entrance giving into an excavated hollow in the interior. But one remarkable feature distinguishes the whole class, that is the hollowing or lowering by excavation of almost the whole interior surface, so that in some cases the floor actualh' lies at a depth of from 4 to 5 feet below the surface of the surrounding ground. Their close resemblance to certain of the forts, especially to those earthworks which carry the remains of a stone parapet on their respective ramparts, and on the sides of whose trenches rock-cutting is visible, renders it a matter of no small difficulty to distinguish between the two kinds of constructions and also indicates that they are of late date. Not a single example is recorded in Nithsdale, thirty-seven appear in Annandale, being in the proportion of somewhat more than one-third to the number of forts, whereas in the Eskdale and Ewesdale districts the^^ number forty, exceeding the forts by nearly two to one. The situations which many of these enclosures occupy are not in them¬ selves highly defensible, but, set back from the edge of some high bank which margins the river valle\% they are such as would easilv escape the notice of marauders on the roadway through the haugh-land below, while the depression of the interior would further tend to the concealment of stock herded within. A small enclosure showing an excavated interior is one of the group of construc¬ tions which lie on the flanks of Birrenswark Hill. This particular entrenchment is situated at the west end, and was examined by the Society of Antiquaries of Scotland when they made a partial exploration of the Roman camps there in 1898. W'ithin it were found a broken quern and a piece of bracelet of opaque glass,^ the latter an object whose probable date is in the ist or 2nd century of our era. Though some of ^ Proc. Soc. Ant. Scot., vol. xxxiii. p. 235. Iv HISTORICAL :\rONUMENTS (SCOTLAND) COMMISSION. these enclosures undoubtedly are, from the uniformity of plan, contemporaneous with certain of the forts, \'et others, there can be little doubt, belong to mediaeval times, and conceivably some are even later. The name “ Birren ” applied to them as a class seems to have as its root meaning that of “ shelter,” A.S. beorgan, found also in ‘‘ burgh.” Its range of application is thus wide, including both stone and earthen constructions, since the “ shelter” thought of was apparently with reference to cattle and sheep. In not a few the foundations of small rectangular huts, seemingly contemporan¬ eous, are apparent, and in one at Mosspeeble, Ewes Parish (No. 215), there still stands the shepherd’s cottage. It is remarkable that in these eastern dales of the county a ” birren ” is almost invariably to be found in the neighbourhood of any place with a name which appears in history in the rieving and raiding days, and the proximity of these places to the English border rendered some ” corral,” in which the cattle could be concealed, an indispensable adjunct. In the foregoing review of the defensive constructions of the county, an attempt has been made to indicate lines of inquiry, through minute observation of detail and comparison, by which it may be possible to reach some sort of conclusion as to the periods to which the numerous classes of forts belong ; but while such methods may be interesting, and instructive to some extent, the only sure source of information is scientifically conducted exploration with the spade and the consequent recovery of relics. Years of study have now familiarised archaeologists with the art of the potter or of the craftsman in metalwork of the various periods of our prehistory, and the potsherd or the fibula which may be obtained from an excavation is evidence of folk or chronology almost as incontestable as the written word of the historian. The “ DeiTs Dike” or “Dyke.” —In the north-western district of the county are to be found detached examples of the low stony mound and shallow ditch, no doubt much reduced from their original condition, known as the “ Deil’s ” ^ or “ Piets ” (Sanquhar) or “ Celtic ” Dike, and in the north-east, in the parish of Eskdalemuir, are portions of a similar feature, which is there called the ” Diel’s Jingle ” (No. 176). The longest section, running from near the Nith to the boundary of Ayrshire, is in the parish of Sanquhar (No. 566), and stretches, though not continuously, for about ten miles. Other portions, each extending for rather less than a third of this distance, are in the parishes of Durisdeer (No. 163) and Closeburn (No. 80). In every case the “ Dike ” is sinuous rather than straight, generally following a contour line ; the 600-feet level south of Sanquhar, where the land as a whole lies high, and up to the 700 to 950 feet line on still higher ground. The portions along the east and west sides of the Carron Water in the parish of Durisdeer are significant. Starting at the Enterkin Water on the 500 contour, 50 feet above the stream, the “ Dike ” runs due east as far as the line of an old drove road. About a quarter of a mile due north on the road it again appears and passes northwards, first on the 900 and then on the 800-feet level, to end in an eastwards curl at Nether Dalveen. It is next found on the eastern hillslope on the opposite side of the Carron, running south at an elevation of 750 to 800 feet, till above Durisdeer it turns with the salient of the hill and follows the 700-feet contour north-eastwards parallel with the Kirk Burn. An 1 Cf. also Report and Inventory of Wigtown, passim, and of Kirkcudbright, p. xxii., and Art. No. 368. The line of broad mounds west of Hightae and Heck in Lochmaben parish, which is marked on the O.S. 6-inch maps as " Murthat or Deil’s Dike,” is really a natural formation of stratified sand and gravel {kames) utterly different in size and character from the Dike proper. Ivi INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. isolated portion occurs on the opposite slope of the burn behind Durisdeer, and a much longer section to the south of the Hapland Burn follows the above levels south-west¬ wards. What is noticeable along this whole stretch is that it almost entirely still marks the boundary between the cultivated land and the moorland. Nor does the nature of the structure anywhere suggest that it ever had any other purpose than that of a boundary or march or head-dike/ though, on the assumption that it was con¬ tinuous, a strategically defensive purpose has also been claimed for it. The bank is everywhere low and the trench slight, seemingly only what was left when the earth was heaped up to form the bank. In one case at least what was probably a core of unusually large stones has been exposed (No. 8o). The service of a typical mediaeval march of this sort no doubt varied; south of Sanquhar it strikes across the moor still roughly in line with the river ; but a march to all appearance it was. As a defence it could be penetrated anywhere, unless well defended ; the population could never be sufficient to defend its whole length, and any local defence could be turned. “ Celtic ” is obviously a comparatively modern term ; and the “ Deil” is a favourite engineer all over the country, as also are the Piets in their semi-mythological stage. Another name to which reference may be made is “ Kemp’s Castle ” (No. 557) for a hill fort. In various parts of southern and north-eastern Scotland, from Wigtown to Forfar, the name appears either in this form or as Kemp’s Graves, Kemp’s Cairn, etc. A gloss accompanying a 13th-century charter of lands in the Registrum Moraviense, App. No. 4, gives for one name the meaning “ of the Grett or Kempis men callit Fenis.” “ Kemp,” indeed, occurs sporadically in the literature of both England and Scotland from a very early period in the sense of a “ great warrior ” or “ champion,” latterly with a suggestion of something monstrous either in size or acti¬ vities. Thus, as above, the Gaelic or Fingalians were ” Kemps.” The applica¬ tion to imposing prehistoric structures of unknown origin is obvious. “ Bogle Walls ” (No. 638) is simply a more eerie version of the same idea. Crannogs. —Crannogs, which are islands in whole or part formed artificially for residence, are not, as a rule, conspicuous structures. Of the four so far located in Dumfriesshire the crannog at the Black Loch Sanquhar is noted in Art. No. 568, and described in the Transactions Dumfries and Galloway Soc., 1864-5, pp. 4 -5. At the time of its exposure, it was found to rest, apparently directly, upon the subsoil, having only a ring of boulders to strengthen the base. The upright piles were of oak ” dressed and sharpened by a metal tool ” and ” some of them morticed at the head ” for the transverse beams, which were ‘‘ chiefly of birch wood.” On this wooden platform was a layer of broken stone from 12 to 18 inches deep, on which had accumulated the vegetable mould covered with vegetation, the surface being 6 to 8 feet above the bottom of the loch. A narrow, curving causeway connected the island with the shore, and in the mud was found a canoe, which was formed from a single oak tree 16 feet in length and tapered from 3 feet at its widest to i foot 10 inches at the prow. No other relic was discovered. Here ma\' also be mentioned the relics of a stockade found about 1877 on the farm of Kelloside, Kirkconnel. The stakes, about three feet in length and six inches in diameter, enclosed an area ^ In the parish of Kells, Kirkcudbright, it is known as " the Auld Head Dyke of Scotland ” (Chalmers, Caledonia, v. p. 237). Munro’s Ancient Scottish Lake Dwellings, p. 245. Of the others included in the list that at Loch Urr is the subject of Art. No. 144, while the references in the list to Lochwood, Closeburn, and Morton apply to the position of the castellated structures at these places described in the Inventory. Ivii HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. of about half an acre and “ in appearance and morticing ”—the “ mortice holes about one foot from the bottom of each stake”—“were exactly similar to those in the stockade ” of the Sanquhar lake dwelling d There was no vestige of the wood on the hard ground. On the south-west side of the Castle Loch, Lochmaben, is a small artificial island, now several feet under water, from which oak mortised beams have been recovered.- In 1863 there was exposed in a peat bank in Corncockle, in the parish of Applegarth, a stratum of parallel oak logs with from 6 to 7 feet of peat above and below. The platform of logs was covered with birch twigs on which was a layer of bracken, the latter two together giving a thickness of 10 inches. At one spot was a circular paving of flatfish whinstones 6 to 7 feet in diameter, on which were many fragments of burnt wood. Beside it were seven large oak bowls 10 to 12 inches in diameter and an oak mallet. There were no piles, but the ends of the logs, of which the largest was 14 inches in diameter, had obviously been cut, and two had square mortise holes. The portion of the platform uncovered was from 20 to 30 feet wide, while the ends of the logs could be followed on the face of the bank for 150 feet.^ Near Friar’s Carse, in the parish of Dunscore, a small loch, on being partially drained, revealed the presence of an artificial island already noted by Grose.^ The island was slightly oval and was surrounded by piles, while the plat was com¬ posed of oak beams, the ends of which overlapped or were mortised. Within the piles the space measured some 80 by 70 feet. Near the centre was a circular paving of small stones, and there were also some remains of clay flooring. In the same quarter was a heap of debris 2 to 3 feet thick, which contained ashes, charcoal, some bones, and fragments of pottery. Two of these fragments were “ handles of jars with a yellowish glaze, inclining in some parts to a green and in others to a reddish- brown colour ”—obviously mediaeval. About sixty yards from the island a canoe was found, 22 feet long and 2 feet 10 inches broad, with a flat stern-piece fitted into a groove. From the west side of the loch came a paddle 3 feet 10 inches long and a hammer-head of whinstone 10 inches by 5, which was perforated for a handle.^ A canoe “ cut out of one solid piece of wood ” was found also, about the beginning of the i8th century, in a moss not far from Morton Castle.® II. Castellated and Domestic Structures. The defensive constructions described above are those of communities large or small. The private castle as the residence of a lord and his retainers, for whose defence it was primarily intended, was introduced into this country by the Normans, and the earliest form of such castle was of the mote-and-bailey type. The mote was a hillock of earth with steep sides surrounded by a deep trench ; the bailey was an attached enclosure at a lower level likewise entrenched (fig. 3). On both the superficial defences were of wood. On the hillock stood a wooden castle or hretasche within a palisade. An earthen rampart, which was also crowned by a palisade, rose above the scarp of the bailey, while the counterscarp generally bore some form of thorn entanglement or hedge {heriqon). From this general type there were several deviations, some of which are illustrated in Dumfriesshire. ^ Trans. Dumf. and Gall., 1897-8, pp. 32-3. 2 Munro as cited, p. 32 ; Proc. Soc. Ant. Scot., vi. p. 60 ; Archceol. Scot., iii. p. 77, n. 2 Proc. Soc. Ant. Scot., vi. pp. 163-5. ^ Antiquities of Scotland, i. p. 146. ® Munro as cited, pp. 152-8 ; Proc. Soc. Ant. Scot., xvi. pp. 73-8. ® New Stat. Acct., iv. p. 96. Iviii Ancient and Historical Monnments — Dumfries. Tofi,, p. Iviii. -MOTES, AND BRUCE STONE EROM ANNAN. INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. In many cases the mound alone survives, and there is no trace of a court or bailey, which may indeed have disappeared in certain instances through agricultural opera¬ tions ; in others we are not justified in believing it ever existed. At Lochmaben (No. 445 (i)) there is only the great mound with its encircling ditch and evidence of other ditches on one side. The dimensions of this mound are unusually large for a mote, but may be compared with those of Troqueer in Kirkcudbright, and of similar mounds in England. What was the mote-hill at Tibbers (No. 157), too, was also of exceptional size. While at Coats Hill (No. 395) cultivation may well have encroached upon the further defences of the mound, there is more doubt with regard to the Mote of Hutton (No. 296), where cultivation seems unlikely and where there is no sign of a bailey. Good standard examples of the complete mote and bailey are Auldton Mote at Moffat (No. 483), the Mote of Rockhall (No. 448), and the Mote of Ingleston (No. 238), while Dinning Mote (No. 65), too, is normal in plan, save that there is no ditch intervening between the mote and bailey. In each of these cases, and indeed in almost every case in the county, defences have been carved out of natural ridges, though no doubt work was done in heightening the mound and sloping the scarps. Hutton Mote appears to be entirely artificial, though placed upon a naturally lofty site, where the upcast earth has been used to give additional height to the mound. Thus, in most cases— Dinning, Rockhall, Maxwelton (No. 241) are good examples—the slope of the hillside augments the defences of both bailey and mound. A notable feature in some of the motes is the presence of terracing on the scarp, of which examples also occur in the neighbouring county of Kirkcudbright. Here may be noted the two terraces on the mound at Lochwood (No. 316) ; at Garpol Water (No. 397) the terrace round the mound is a prolongation of the bailey court, but the ditch also is continued below. In both examples there are traces of drystone parapet walls on the terrace, and at Garpol even round the bailey court. Though the standard plan of mote is that of the round hillock, and the bailey is fitted to it with a curvilinear outline, the shape of both is generally determined by the nature of the high ground which has been utilised. Thus at Annan (No. 3) the mote is pear-shaped, and the bailey very long in comparison with its breadth, while at Dinning (No. 65) the bailey is rectangular. As might be expected, no signs of the wooden defences are now discernible on the surface, and no proper excavation of these sites has been made, but this type of wooden defence persisted in Scotland generally till a comparatively late period. Similar in principle were the characteristic “ peels ” or palisaded enclosures erected during the War of Independence by Edward I., of which Lochmaben Peel (see No. 445) was an example ; while the fact that Edward Bruce in 1313 could capture thirteen castles in Galloway in one year suggests that these were still the small castella of the mote-and-bailey t\'pe, of which so many traces survive. The peel, indeed, as the simplest and cheapest form of defensive structure, persisted right through Border history (c/. p. Ixii.). The position of these Dumfriesshire motes was apparently dependent upon different considerations. That at Castledykes, Dumfries (No. 128), was of course a royal construction ; with Troqueer Mote on the opposite side of the river it probably covered a ferry crossing, as in the parallel case at York. Others, such as Annan, were manorial residences or the head places of baronies. Annan, Lochmaben, and Moffat motes were the work of the Bruces, the dominant family in Annandale. Of the more outlying examples nothing very definite can be said. Many are in the neighbourhood of fords ; Garpol iMote (No. 397) is a conspicuous example of this position. Hutton Mote is on a retired but lofty site with a wide view of the surrounding country ; it was lix HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. the head of a barony. In size, too, these structures vary greatly, and in most cases the number of occupants must have been smaU, as the motes represent, at least in the main, the incoming of Norman settlers and their planting a fixed footing in the country. Necessarily their quarters, however strong, cannot have been of great dimensions. Some examples may well be due to local lords in imitation of the master¬ ful incomers. Chronologicalh^ the t5'’pe at least must be assigned to the 12th century as the time of its introduction. Noteworthy in their relation to these places is the occurrence of such names as “ Boreland ” and “ Ingleston ” [i.e. Engiish-town) in the vicinit}'. The former stands for the “ bordland,” that is the land provisioning the “ board ” or “ table ” of the lord, while the latter represents a settlement of retainers or followers. ^ Of the earlier type of Norman stone castle, the tower and courtyard which re¬ peated in stone the features of the mote-and-bailey wooden castle, no contemporary example exists or is likely ever to have existed in Dumfriesshire, though this type of residence never died out, and examples of a late date are plentiful in Scotland and in England. The form of stone castle which immediately succeeded the timber type in Scotland is that of the wall of enceinte with flanking towers exampled at Tibbers (No. 157), where it crowns the original mote, Auchencass (No. 384), and the first Caerlaverock (No. 33 (i)). We have perhaps evidence of the planting of such castles of a new t\'pe towards the end of the 13th century in references to Clifford’s “ house ” at Tibbers “ just begun ” in 1298,^ and the new place of “ Seneware ” (Sanquhar) in 1296.^ If not the main work, these are at least buildings within it. The examples cited follow the contemporary form of English late 13th-century castle. The earlier Caerlaverock, however, though apparently belonging to the same class, must have been older than the others, if it was the place besieged and captured by Edward I. in 1300. These late 13th-century structures are not upon lofty sites, with the exception of Tibbers, where the original mound was strong enough to bear the heavier structure, but depend for strength of situation upon surrounding marshy land. Within this structure of wall and towers, and apart from the residential facilities afforded by the towers, all buildings of a domestic or service character would probably, in the first instance at least, be substantially of wood, for which, however, in general, at a later stage, buildings of stone less massive in character than the defensive walls and towers were substituted. It will be observed that in none of these 13th-century examples is there any dominating tower or donjon as the main defence and final refuge of the garrison. This feature had for the time gone out of military fashion in France and England, owing partly to the development of siege craft, partly to the desire of occupants to have more space and more comfortable quarters in which to stand a siege. By the rounded towers at corners of the enclosure a flanking fire was secured for the curtain walls, while a marshy site or broad wet ditch on the level, or a spit of land in a lake, inter¬ posed an effective obstacle to the mining operations of a besieging force. Caerlaverock (No. 33 (2)), Auchencass (No. 384), Morton (No. 510), and Lochmaben (No. 445 (2)) are all in their degree examples of this class. Special attention is given to the gateway as the vulnerable place in the structure. Its outer opening is flanked with round towers and is continued in a vaulted passage. At Auchencass, however, there is the simpler feature of an entrance past the corner tower turning upon itself at right angles, a donjon type. At Caerlaverock additions and elaborations from time to ^ But cj. reference to “ Engless men ” on p. xlii. ^ Bain’s Calendar, ii. No. 1005. ^ Ibid., ii. p. 206. lx An. iei:: and Historic al ^fonuments — Dtimfrits. To face /. Ixi. Fig. 4.— towers. INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. time made the original entrance more formidable. Morton is probably the latest of these enceinte castles ; its defences are concentrated on the one way of approach, where, too, the lower part of the curtain wall shows a pronounced batter, which forced into fuller exposure assailants attempting to undermine the foundations. Here, as also at Lochmaben, three sides of the building are flanked by the water of the loch. In the case of Lochmaben the entrance between half-round towers and over the canal¬ like ditch—the last of three ditches—comes immediately upon the mass of the main building, behind which extend successive rectangular wards. Several circumstances, however, brought the square tower back to favour in England, and it was at this stage that the tower residence properly took root in Scot¬ land. On the borders of both countries it figures in isolation as an independent structure (fig. 4). Where it is incorporated in more extensive wards, as at Lochwood (No. 315) and Sanquhar (No. 551), it will be found that these added wards are of much later date, are mainly domestic in character—so far as the distinction can be made— and bear witness to the growing fortunes of the families concerned. These two are probably 15th-century towers. Certain examples, such as Closeburn (No. 59), Tor- thorwald (No. 590), and Spedlin’s (No. 446) are of earlier type, and accordingly more massive. Comlongon (No. 537) alone supplies examples of chambers in the mass of its ii-feet-thick walls. The latest examples of free-standing towers, which are also the most numerous, display an increasing fondness for ornament. Repentance Tower (No. 89) was purely a watch-tower. The fact is that in these towers we see not a fortress in the strict sense of the term, in which everything is subordinated to military purposes, but the residence of a local magnate or laird which was also fortified. “ The houses of the Grames that were,'’ writes a traveller of 1629, “are but one little stone tower garretted and slated or thatched, some of the form of a little tower not garretted ; such be all the leards’ houses in Scotland.’’ ^ They were not expected to withstand a regular siege, but with a small garrison offered quite effective protection against a raid. In their external features the towers thus display a progressive insistence upon residential conveniences, of which Caerlaverock, in a different class, affords an impressive illustra¬ tion in the extension and adornment of the domestic buildings within the curtain walls. In the case of the towers such additions and reconstructions from century to century make it difficult to give them a strict chronological sequence. As a type, however, this form of residence may be described as a square-angled tower varying in dimensions and rising to a height of 40 or 50 feet at the wall head. The walls are generally about 5 feet thick, but, while some are rather less, the lower parts of Lochwood Tower are 9 feet thick, of Hoddom feet and nowhere less than 8 feet, of Closeburn 10 feet, and of Stapleton 12 feet, thinning above to less than 6 feet. The lower stages were, of course, the more vulnerable, but fire seems to have been the enemy most feared. Still, in the i6th century, the “ viii foote ’’ walls of Lochmaben were accounted of “ small thyknes ’’—probably with respect to artillery.^ The walls of Castlemilk Tower were ii feet thick, and of Cockpool 14.^ We might expect a tower to be surrounded by a boundary' wall, within which, or in the line of which, minor enclosures would be found. A late 18th-century illustration of Hoddom Tower (No. 90) shows it standing within such a close. When the enclosing wall was of stone it was known as a barmkin or barnekin {cf. p. Ixv.); in certain cases it might be a palisade ^ Hist. MSS. Comm, xiii., App., part vii. ^ Armstrong’s Liddesdale, App. Ixx. p. cxiii. Ixi 3 Ibid. HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. of timber known as a peel} In da^^s when the palisade no longer existed, the word survived, and the peel tower continued to be known as the peel, a descriptive name looseh- extended to all such towers whether originally they possessed a peel proper or not. Internallv the tower contained at least two floors above the basement, and in most cases more than two. The first floor was the hall or main living-room of the tower. The only feature common to every important example, with one exception, is that the basement should be vaulted in stone. But Elshieshields (No. 447), which is among the latest, has its basement roofed with oak-beams. This basement was always a storeroom, and it might be a stable. The upper part of the vault was usually floored at the springing of the arch to give further accommoda¬ tion. In the logic of the building there should be no direct access from the basement to the upper floors, save perhaps by a hatch, but this cannot be predicated of the Dumfriesshire examples. Nor are there many cases of the correspondmg feature that the main entrance of the tower should be on the first floor. This is true of Lochwood, Sanquhar and Closeburn, was probably true of Spedlin’s, and possibly of many others in their earlier forms. The an¬ nexed illustration shows this characteristic in the case of the old tower of Castlemilk (fig. 5), with the usual mode of approach —a wooden ladder. The upper rooms were floored with wood, and numbered two or more, probably according to the age of the Fig. 5.—Castlemilk from the “ Plat of Milk Castle,” c. 1547 tOWCr. ThuS Torthorwald^ haS tWO Vaulted (Hatfield). From tracing in Armstrong MSS. storeys, which was also Originally the Con¬ dition of Closeburn and Spedlin’s. The later demand for greater comfort increased the accommodation in the provision of upper rooms and by adding to the height of the building, as also by the projection of turrets at the angles, till in the late structure of Amisfield we find the upper part of the square block opening out in such excrescences like a flower. We find, further, the circular stair or vice, which had generally been tucked into a corner of the building, as at Closeburn, Lochhouse (No. 388), Robgill (No. 107), Lag (No. 136), etc., and even in the much later towers of Hollows (No. 43), Stapleton (No. 106), etc., and so, as in the latter examples, had frequently encroached upon the internal space, removed, as at Amisfield, to a corner turret rising from the first-floor level, or, as in Elshieshields and Blacket House (No. 460), wholly confined to a separate wing. At ^ Peel or pele is for Old French pel, from the Latin accusative Aw, a stake. In 1544 we have a note of the burning, among other things, of “ peel houses, corn and steads in Hodholme . . . and all the peels in Myddleby and Middleby Woods ” {Letters and Papers, Henry VIII., Foreign and Domestic, vol. xix. part ii. p. 373). An Act of 1535 ordered every man dwelling in the inland or border having land to the annual value of £100 to build in a convenient place a “ barmekyn ” of stone and lime 60 feet square, with walls an ell thick and 6 ells (Scots ell=34! English inches) high, as a refuge to himself and his tenants in troublous times, with a tower in the same for himself, if thought expedient. Those having a smaller rental w'ere to construct peels or great strengths, as they pleased, for saving themselves, tenants, and goods. Ixii INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. Barjarg (No. 327) the tower containing the stair and basement entrance had been at the re-entering angle formed by the junction of the tower with its original wing. In Frenchland Tower (No. 480), where there was a good deal of reconstruction, we have the common arrangement of a stair to the first floor provided in the new wing, the upper part of which, however, was laid out in rooms, while the old wheel-staircase still served all above the first-floor level. One sign of a late period of construction is the presence of ornament based on military features ; Hollows thus displaying its late 16th-century character in the ornamental corbelling and cable ornament of its parapet, which projects so slightly Fig. 6. —Map showing the situation of Castles and Fortified Houses “ in the Debateable Land,” 1590. From iMap in British Museum (Bib. Reg., 18 D iii.). as to be of little military effectiveness. In Isle Tower (No. 337), too, as in Elshie- shields, we see the total disappearance of the defensive wall-head, the sides passing immediately into the gabled crow-stepped roof. At Spedlin’s, however, where the same construction appears, the upper floors have been imposed on an older as much more massive portion. Ornamental detail is most conspicuous in the case of Amisfield, where Renaissance pilasters on the dormer outface Gothic dog-tooth ornament on other windows and string courses of both early and late design. In the articles on Amisfield and Elshieshields reasons are given for the belief that both places are due to the same designer. A feature of these defensible Border houses was the iron “ yett ” or gate, of which a few examples survive (see fig. 133 and Index), and which was placed just within the wooden door {cf. p. Ixv.). That these iron gates were both formidable and Lxiii HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. numerous is indicated by the decree of the Privy Council in 1606 ordering their destruction in all “ houssis and strenthis ” in the Borders save those of “ answerable baronis.” The reason given was that their presence made it difficult, in case of trouble, “ to wyn and recover the saidis houssis and to apprehend the lymmairis being thairintill.” The “yettis” were accordingly to be removed and “ turnit in plew (plough) irnis or sic other necessar werk.” ^ This measure was part of the general policy for the establishment of peaceful conditions on the Borders, but the “yetts” are known as defensive features that were common throughout the country.^ In every case their manner of construction is similar and apparently peculiar to Scotland: the bars penetrate mutually and alternatively in alternate compartments. As suggested by the map (fig. 6), most of these ancient structures have been swept awa}^, mainly, it would appear, within the last hundred years or so. For example, at the close of the eighteenth century there were still the remains of five towers in IMouswald parish,® where now is but the fragment of one. Even fifty years later, in the parish of Kirkpatrick-Fleming, there were seven similar towers within four miles of the one at Woodhouse,^ the solitary and ruinous survivor. A few in greater or less degree still serve as dwelling-places. Bonshaw (No. i), Stapleton, Lochhouse, and Isle have lost little of their original character and are inhabited. Robgill, Breckonside (No. 475), and Sundaywell (No. 137) have been incorporated in modern structures ; but the chief example in this connection is Hoddom Tower (No. 90), the central feature of the present mansion. Fourmerkland Tower (No. 280) was occupied till comparatively lately ; two storeys of Bogrie Tower (No. 138) make a shepherd’s house. On the Borders, indeed, any stone building might have on occasion to serve as a fortress. The map (fig. 6) of strong places on the West March in the i6th century thus includes all types from the castles proper of Lochmaben and Caerlaverock to the humblest of residential towers. Thus, too, Annan Church steeple could attain the rank of a fortress, and suffer siege (see p. xxxi.) ; a process reversed in the case of the later castle proper, which was adapted as a church (see p. xlvi.). What probably was the ordinary type of town house of the poorer sort is described by the traveller of 1629 already cited. At Langholm he lodged “in a poor thatched house the wall of it being one course of stones, another of sods of earth, it had a door of wicker rods.’’ The story of the successful attack upon the steeple at Annan by an English column in September 1547 illustrates the method of such operations upon the Borders. “ And we having in ordenaunce but a facon, a faconett, and foure quarter faeons, for that ther is no baterie peice at Carlisle, divised that night [i.e. Sunday, September 4) howe we shulde maik warr agaynst the house on the morowe. At viijth of the clok in the mornying, we laid those sex peices to beit the battailling, and appoyntid certane archers and hagbutters to maik warre also untill a paveis {i.e. large shield) of tymbre might be drawn to the sidde of the steplee, under whiche sexe pyoners might work to have undermyened the sam ; and in putting these to effectes, they in the house maid sharpe warre, and slewe foure of our men and hurt divers others. And with grett stones from the steple toppe, brooke the paveis after it was sett, and being in that extrymytie, lakking ordenaunce for that purpose, we caused certane pyoners cutt the walle of the east end of the quere, overthuart abone the earthe, and caused the hooll ende to falle, wherwith the rooff and tymbre falling inward, slewe vij Scotes- 2 C/. Proc. Soc. Ant. Scot., xvii. pp. 98 ff. ^ New Stat. Acet., iv. p. 279. Ixiv ^ Reg. P.C. vii. p. 271. ^ Stat. Acet., vii. p. 298. INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. men. And after that we caused the peices be laid to shoot at the doore of the steplee which was a house hight [i.e. one “room” or “house” above the ground floor), and that house hight rampered with earthe, and caused them further to myen.” And then the captain about 4 p.m. took down his “ pensall (flag) of defyaunce” and he and his men “ cried for marcie.” So they surrendered without conditions, and the captain, “ a tall gentleman,” with his fifty-seven men came out and delivered the “kies.” On Tuesday morning we “ cutt and raiced down the churche wallis and steplee, and brent the towne, not leving any thing therin unbrent; which was the best town in Anderdaill.” ^ The capture of Lochwood (No. 315) in the same year was a more humiliating affair, but the account contains several references of structural interest :—“ We came there about an hour before day ; and the greater part of us lay close without the barnekin : But about a dozen of the men got over the barnekin wall, and stole close into the house within the barnekin, and took the wenches and kept them secure in the house till daylight. And at sun-rising, two men and a woman being in the tower, one of the men rising in his shirt, and going to the tower head, and seeing nothing stir about, he called on the wench that lay in the tower, and bade her rise and open the tower door and call up them that lay beneath. She so doing and opening the iron door, and a wood door without it, our men with the barnekin brake a little too soon to the door ; for the wench perceiving them, leaped back into the tower, and ha'd gotten almost the wood door to, but one got hold of it that she could not get it close to ; so the skirmish rose, and we over the barnekin and broke open the wood door, and she being troubled with the wood door left the iron door open, and so we entred and wan the Loghwood; where we found truly the house well purveyed for beef salted, malt, big {i.e. barley), havermeal {i.e. oatmeal, cf. German hafer, oats), butter and cheese.” Of the furniture of these residences only the structural constituents remain in ornate chimney pieces, such as the early one at Comlongon and the Renaissance examples at Spedlin’s, Amisfield, and Caerlaverock ; the stone-silled recess—buffet or cupboard—at Amisfield, and the late Gothic stone cupboard or buffet, which was shelved, at Comlongon ; as well as various smaller aumbries and lamp-recesses throughout. The cutting off of one end of the hall at Comlongon to form a kitchen is a feature pareilleled in Elphinstone Castle, East Lothian, and the towers of Law and Eairlie on the Eirth of Clyde. At Spedlin’s are indications of the position of a screen and gallery in the hall, and in the second floor at Amisfield are the last crumbling traces of the brightly coloured design which formed a frieze below the patterned corbels, while the room below still bears some of its plaster cornice. It happens, however, that an inventory of the contents of Caerlaverock Castle was made after its surrender in 1640, the main features of which may be briefly described. The bulk of the furnishing is in beds, many of them “ canaby ” or canopy beds, trunks and chests, some having locks, cupboards—apparently of various types, as one was “ lead our with gould lace,”—chairs and stools. Besides those in other rooms there were three beds in the “ hich wardrop ” and three in the new “ wardrope,” four in “ Sanders ” chamber, a canopy bed in a drawing (“ draiug ”) room, and a falling bed as well as a “ burd ” or table in the “ daning (dining) rume before my lady’s camber.” In fact there were taken ^ from the castle five beds richh* equipped ^ Scottish Papers, i. No. 42. “ Cited in the History of Westmorland and Cwnberland, Nicolson and Burn, vol. i. p. liv. ^ ” Intromettit \\ith ” by Lieut.-Col. Home on the plea that the conditions of surrender had been broken. Ixv HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. with curtains edged with hcav}* silk fringes and lace, each having its proper bed¬ clothes “ with chairs and stools ansnerabillie,” and each bed with its furniture and “ bedsteid of timber ” valued at £110 sterling ; 1 ten lesser beds with furnishings each at £15 “ owerheid ” ; and twenty other beds for servants with equipment at £/ each. There were also taken two dozen chairs and stools covered with red velvet fringed with crimson silk and studded with gilt nails, the whole estimated at £60 ; and live dozen of Turkey, i.e. tapestry work, each chair worth fifteen shillings and each stool nine shillings. Added to these was a large spoil of napery of all kinds, tablecloths, napkins, towels, sheets—part of damask, part “ cowrse,”—and eight suits of apparel in a trunk, some of velvet, some of satin and some of cloth. The furniture of “ ane drawing rowme ” was in cloth of silver and included “ ane cutche bed ” [i.e. a bed without canopy or tester), a great chair with a cushion and footstool, six other chairs with backs and six stools, all garnished with silk and silver fringe. In the New Hall were a “ leid ” and “ a maskin fatt,” both being large vats or vessels for brewing ; in the Long Hall six cases of windows, 22 pikes and 13 lances ; and in my Lord’s Hall two “ burds ” or tables and six tapestry stools. Special articles were a painted board “ in the round chamber without my Lords chamber,” ” m3’ lord and my lady’s pictures” in another room with various articles of convenience, a table cloth valued at £20, two red window curtains, a pair of virginals," a long cushion of black and white stuff, some chairs and stools covered with brown cloth embroidered in yellow (passementet ” yealow ”) or red cloth with black embroider}’, five suits of hangings {i.e. for the walls of rooms) of eight pieces each, and each suit worth £60, forty carpets (table-covers or bed-covers) ^ large and small, averaged at forty shillings each, 22 curtain rods, and a library of books which had cost £200. It is evident from the general character of the furnishing—the number of chairs, cupboards, and of beds per room, and the abundance and richness of coverings and draper}’ of all sorts, which were taking the place of the earlier elaborate carving— that it was of comparatively recent origin, probably contemporaneous with the building of the new wing before 1620. The richer stuffs, such as damasks, velvets, etc., must have come from either Italy or England, more probably the former. HI. Ecclesiastical Remains. Dumfries, as part of the old Strathclyde kingdom, was included in the diocese of Glasgow on its re-constitution by King David, while still Prince of Cumbria, in the first quarter of the 12th century. The earlier connection with St Kentigern has been noticed above (p. xxi.). A later ecclesiastical link with Yorkshire was established by the grants to the Augustinian Priory of Gyseburn (Guisbrough) of Annandale churches by the Bruces, who had been founders of the priory {c. 1124). Thus, in the late 12th century, and down to the final breach caused by the War of Independ¬ ence, we find Gyseburn in possession of the churches of Annan, Lochmaben, Kirk¬ patrick with Logan Chapel, Cummertrees, Rainpatrick, and Gretna. Of regular foundations there were three within the county ; Canonbie in Eskdale as a cell of ^ All values are in sterling money. 2 A keyed musical instrument of the pianoforte type. There was only one instrument, “pair” being descriptive, not numeral. ® Some may have been used as " foot-carpets,” but the exclusive use of the word in this sense is not established before the middle of the next century. Ixvi Andoit a/id Historical Monuments — Dumfries. Merkland Cross. Cross-slab, Wauchope Churchyard. Cross-slab, Hoddom Churchyard- Cross-socket, Orchard. Fig. 7 .—crosses. To face p. Ixvii, INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. Canons Regular of Jedburgh, the Abbey of Holywood {de sacro nemore ; Saint Boyse) or Dercongal (“ oak of Congal ”) for Premonstratensians in Nithsdale, and the house of Franciscan Friars in the town of Dumfries. Of these only a site or vague relics remain. What still survived of Holywood was used in 1779 as a quarrj^ for the new parish church.^ The two bells of the abbey also have found refuge in the church. The same general condition applies to the hospitals at Sanquhar (No. 572), How- Spital (Annan), and Spital (Dumfries), the latter two preserving the name,^ and to the chapelry at Trailtrow. But churches on the West March had a hard time like everything else. It has been shown how Annan Church became a fortress, and how the new fortress became a church (p. xlvi.). Also how the vocation of the Armstrongs and their like was inimical even to sacred buildings, so that by the beginning of the 17th century many mediaeval fabrics were in a ruinous condition (pp. xxxvi., xlv.). The thrifty combination of parishes in the course of the same century, with the provision of one new church in place of two or three older ones, further contributed to the disappearance of the original structures. There was thus a fresh building period early in the 17th century, and there is evidence of another about a hundred years later. Garvald Church (No. 355) was a reconstruction of 1617, and is now a ruin. Durisdeer Church (No. 152) is a large composite Renaissance building of the late 17th century, and still in use. The older church of Mickle Dalton (No. 96) is of a few years later, but has been abandoned for the modern edifice. The earliest fragments of mediaeval building are St Cuthbert’s Chapel at Moffat (No. 383), some part, perhaps, of the late church at Glencairn (No. 229), and an arched recess within Canonbie Churchyard (No. 42), all probably of the 13th century. To these must now be added, as the result of excavation in the course of the summer of 1915, the foundations and part of the walls of Old Hoddom Church beside the Annan (No. 271). The Roman streets and buildings at Birrens provided its stones. The chancel is rectangular, inside and out ; the chancel arch is comparatively narrow ; and the dimensions of the building correspond very closely to those of St Helen’s, Cockburnspath.=* One or two minute portions of painted glass were found, of the t3^pe known from Coldingham Priory,"^ which are of a date at the close of the 13th or the beginning of the 14th century. Certain of the cross-slabs found beside the church also appear to be of 13th-century date. Crosses.— The sculptured crosses are few in number, and some are represented only by fragments. All are overshadowed by the magnificent monument of the Ruthwell Cross, which is unique as furnishing also the text of a fragment of an earh- poem now lost in this linguistic form. The whole subject of the Ruthwell Cross, however, occupies a special place in the Appendix. The only other complete examples are the much-worn one at Thornhill beside the Nith (No. 531), and the late mediaeval cross at Merkland (No. 378). With the exception of this last, and that at Thornhill, in which the decoration is wholly zoomorphic, all the examples, whether whole or fragmentary, are of the Northumbrian or Anglian t3"pe, man\" displa^fing the char¬ acteristic decoration of scroll foliage involving birds and other creatures. Among the very numerous cross-slabs of north-eastern Scotland three onl^’ displa\’ this motive —the Hilton and Tarbet cross-slabs at Invergordon, and the one at ]\Iugdrum, Fife — ^ Buccleuch MSS., p. 69, 2 Chalmers, Caledonia, v. p. 154 ; Spittalriddinghill is north-west of Annan. 3 Berwickshire Inventory, No. 46. * Ibid., p. 40, No. 74. Ixvii HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. all of later date ; while it is not found in Scotland outside these limits. In the skill of the relief work, too, these southern crosses are distinguished from similar cases beyond the Forth. The fragments from Knockhill (No. 273), now described in detail for the first time, have suffered most severely. They apparently represent a small group of crosses, and also have their own special features. On the Ruthwell Cross the subjects are scriptural or saintly narrative, or are symbolic in a straightforward wav noticeable also at Knockhill, where, however, the other surviving subjects appear to be allegorical or representative. On the principal specimen in the Grierson Museum, Thornhill (No. 514), the figures are of a S5mrbolism that hitherto has withstood explanation. Bells.—Of the bells in the county the oldest is that still in use in the Parish Church of Lochmaben (No. 452), which may be of the early 14th century. Its companion is much later. A 15th-century bell survives in the Maxwelltown Museum, Dumfries (No. 134). The bells of the vanished Abbey of Holywood (No. 285) are, one certainh’ and probably both, of the early i6th century. The 17th century has left a few examples : one at Closeburn Church (No. 58), another on a tree at Ewes (No. 227), and one in Moffat (No. 496). The bell at Closeburn is a Potterrow (Edinburgh) casting, and the handsome 18th-century bell at Mickle Dalton (No. 96) is also from Edinburgh. Ixviii INVENTORY OF THE ANCIENT AND HISTORICAL MONUMENTS AND CONSTRUCTIONS IN THE COUNTY OF DUMFRIES. ANNAN. Castellated and Domestic Structures. I. Bonshaw Tower. — Bonshaw Tower (fig. 4 of Introduction), dating from the i6th century, lies less than a mile south- south-east of Kirtlebridge Station, on the PLAN or BASEMENT AND EIRST ELOOR. 10 S O 10 20 30 40 SO H fHI III H I I I I-H FEET. Fig. 8.— Bonshaw Tower (No. i). western bank of the Kirtle Water, which washes the eastern base of the declivity on the summit of which the tower is set. To the south is a ravine traversed by a burn ; to the north and west there are no defences visible, a walled courtyard, which extended eastwards from the tower to the cliff and through which the tower was reached, being deemed sufficient. The entrance to the tower is in the east wall; over it is carved in raised characters the motto SOLI • DEO • HONOR • ET • GLORIA. The door opens on a passage admitting to the basement and the wheel-stair in the north-eastern angle. From the stone roof of the vestibule hangs a pendant, on which is carved IHS in monogram, as at Robgill Tower (No. 107). The building measures exteriorly some 36 feet 6 inches by 27 feet i inch ; the walls terminate at a height of 39 feet 9 inches from the ground in a parapet and walk carried on corbels of simple design ; a splayed base¬ ment-course returns along the walls at a height of 2 feet 6 inches from the ground. The basement, measuring 15 feet 9 inches by 25 feet, has a vaulted ceiling fitted with a hatch, and is provided with a gunloop in each wall at the level of the basement-course and a small window in the south wall high up in the vault. A stone bin, possibly for storage of provisions, is built against the east wall. A prison, measuring 8 feet 2 inches by 4 feet 4 inches, is formed in the south-west angle. This apartment has no window, but a flue for ventilation is provided in the vaulted ceiling. The upper floors are three in number. The hall occupies the first floor ; it measures 27 feet 2 inches by 17 feet 8 inches and is lit by a window 2 feet 6 inches wide, with modern mullions, in each wall. Those in the east and west walls have an aumbry set in the jamb, and against the jambs of the south window are stone seats. Projecting some 2 feet from 1 I ANXAX.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [ANNAN. the south gable is a fine stone fireplace measur¬ ing 9 feet over the moulded jambs and some 7 feet high. An aumbry 3 feet 6 inches wide, with an ogival-arched head, is set in the east wall. The second floor resembles the hall in general arrangement, with the addition of a garde-robe in the north-western angle. The third floor consists of a garret within the roof, but the roof itself is modern and less steeply pitched than the original. The parapet, which appears to have been recently restored, has a machicolation over each gunloop. The building is connected with the mansion by a covered passage, and is in excellent repair. Bonshaw estate appears to have been ac¬ quired by the Irvings from the Corries after the suppression of the Douglases (see Introd., p. xxviii.). The tower became one of the principal places of the clan in the latter part of the i6th century. It was burned by MTiarton, the English Warden of the West March, in the raid of Sept. 1544.^ In June 1585, being then in possession of Edward “ Yrwen,” and reported “ one of the strongest howses of that border,” it was besieged by Lord Maxwell.^ In July, Maxwell had again placed his forces round Bonshaw,® which seems to have been successfully defended. Early in the next year the Johnstones fell upon Captain Richard Maxwell and his royal police force and carried him off, wounded, to confinement in ” the Bonshaw,” Edward Irving being their accomplice.* * Hamilton Papers, ii. p. 456; ® Calendar of Border Papers, i. No. 321 ; ® ibid.. No. 327 ; * Register Privy Council, iv. pp. 56-7. Iviii. S.W. 25 July 1912.* 2. St. Bryde’s Tower, Brydekirk Mains.—Of this tower, which lay a mile north of Bryde¬ kirk village, only a fragment of the north wall survives, surrounded by the out-buildings of Brydekirk Mains farm. The wall is 15 feet long, 3 feet broad, and terminates at a height of 25 feet from the ground in a frag¬ mentary corbel course. ♦ The reference throughout is to the Ordnance Sur\'ey maps, 6-inch scale, for Dumfriesshire. The date is that on which the structure was visited. ” Habye Carlile of Brydekirk ” is among the landlords ordered in 1590 to find surety under the Act of 1587. {Reg. Privy Council, iv. p. 790). Ivii. S.E. 28 May 1912. Defensive Construction. 3. Mote of Annan.—This mote (fig. 9) is in the garden of a villa known as “ Moat House ” on the west side of the town of Annan. A low meadow, from which it rises with a steep scarp, intervenes for a distance of 100 yards or thereby between it and the River Annan. The mote proper forms the northern extremity of the construction, rising to an elevation of some 50 feet and measuring 22 feet across its level summit by 50 feet lengthwise. A broad trench separates it from the base-court, which extends southward for a distance of 270 feet in an irregular oblong, expanding from a breadth of 50 feet at its northern end to no feet at the south and rising to an eleva¬ tion of 60 feet above the meadow on the west 2 ANNAN.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [applegarth. and 35 feet above the higher levels on the east and south. The mote has been sur¬ rounded on the sides away from the river by a trench, as also probably was the base- court, but the lines of the whole construction have been seriously interfered with in the formation of the villa garden. (See Introd., pp. xxxi.-ii.) Ixii. N.E. 4 October igi2. Miscellaneous. 4. Inscribed Stone, Annan.—An inscribed stone, said to have come from the ruins of a castle or building at or beside The Moat, was seen and copied by the travellers Pococke and Pennant. In 1760 Pococke described it as “ a stone taken from the old building.” In the Caledonia Chalmers wrote of it as ” built into the wall of a gentleman’s garden.” In the New Statistical Account it is stated to have been “ built into the wall of a small vintage-house in a garden in the town.” Sub¬ sequently acquired by an antiquarian resident in the town, it was taken away by him on his removal to the south of England. The in¬ scription k in well-formed lettering of Lom- bardic capitals, but the Arabic numerals forming the date ” 1300 ” are obviously not original and are not cut with either the depth or breadth of the lettering. The stone is decayed and damaged in parts. The inscrip¬ tion is as follows :— Robert • de • Brvs • COUNTE • DE • CA [rRIK] • ET • SEN^[n]U R] • DU • VAL • D[E • AN]N AND • 1300 At the end of the third line the 3 alone offers difficulty. Pennant, with some justification, read the word as senteur, against which, apart from the sense, there is only to be said that the fourth letter has a straight hori¬ zontal top, while every t in the inscription has a curving top. In all likelihood the word is, as has always been supposed, a corrupt rendering of seigneur, perhaps in some such form as sengnur * or senyour. The date * “ Seingnur ” is found, e.g., in Berne MS. See facsimile in Acts Pari. Scot., vol. i. 1300 was a shrewd enough computation : such an inscription could not correctly date in any case earlier than 1292 or later than 1306. See Pococke’s Tours in Scotland (Scottish History Society, 1887), p. 35, where the bishop’s transcription is reproduced in fac¬ simile ; Pennant’s Tour in Scotland, ii. p. g6 ; Chalmers’ Caledonia, hi. p. 139 ; Neilson in Trans. Dumf. and Gall. Antiq. Soc., 1915- 16, p. 69 ff. 5. Stone Circle and St Marjory’s Cross (re¬ mains of), Woodhead.—On the boundary of the parishes of Dornock and Annan, between two plantations, and about J mile west by south of Woodhead cottage, the O.S. map marks ” Stone Circle and St Marjory’s Cross (re¬ mains of).” These now consist of two granitic boulders about ii feet 6 inches apart, the largest of which is some 3 feet in height above ground. Ixiii. N.W. 6 October 1912. Sites. 6. St Bryde’s Kirk and Well, Brydekirk Mains. Ivii. S.E. 7. Newbie Castle, Newbie Mains. Ixii. S.E. 8. “ Cairn of Creca,” Creca. Iviii. S.W. APPLEGARTH. Defensive Constructions. 9. Fort, Whitecastle Knowe.—This fort occu¬ pies the summit of an oval hillock, known as the Whitecastle Knowe, which crowns the western slope of the watershed between the Dryfe and the Annan, \ mile to the west of the farm of Newbigging. The hillock stands at an altitude of 734 feet above sea-level, and, except for two adjacent heights which obstruct the view to the south-east and north-north-west, commands an extensive panorama. On the north and west it rises abruptly for some 30 to 40 feet, while on the south, and still more on the east, the gradient from the surrounding level is easy. The enceinte is ov^al in form, with its longest axis north and south, along which it measures some 455 feet by 260 feet. It has been surrounded by a rampart of compacted 3 APrLEGAKTH.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [applegarth. clay immixed with stones, rising now at no point more than z feet above the level of the interior. This has been supplemented by a trench around the south extremity and along the eastern flank, which has a width from crest to crest of about 25 feet and lies some S feet down from the crest of the mound. From north to south along the western side the rampart follows the line of the summit, but on the east and more assailable sides it is carried along the flank some 8 to 10 feet below the highest level of the interior, with a slight parallel depression in rear of it. Into this lower level the entrance opens on the east, with a width of some 10 feet, crossing the trench and passing through the rampart, whence a track is observable leading up to the higher level. This construction differs essentially from any in the Langholm district, in that the interior at all points is at a higher level than the land outside, and that, except perhaps in rear of the rampart, where there may have been slight excavation, it shows no hollowing out. xxxiv. S.W. 23 July 1912. 10. Fort, Broomhill Bank Hill.—Situated on the west side of the south end of the summit of Broomhill Bank Hill, at an elevation of more than 700 feet over sea-level, is a fort commanding an extensive view of Annandale. The ground rises very steeply to the level of the fort from the east, but elsewhere from below and towards the actual summit on the north-east it mounts by easy gradients. The enceinte, which is approximately circular, measuring some 230 feet in diameter, is surrounded, except above the east declivity, by two concentric ramparts of earth and stone, the inner 18 feet wide at base and the outer 22 feet, which are separated by a trench some 18 feet wade and 4 feet in depth. Some 70 feet beyond the outer rampart lies a third of low elevation, 16 feet broad at base, which runs concentrically from the north-east, and, as it passes from south to south-west, gradually converges with the intermediate rampart, meeting it 106 feet from its termina¬ tion on the south-west. At the termination of the ramparts on the south-west, a hollow, e\adently the entrance, is observable passing into the enceinte at its lowest part : beyond it the outline of what has been a slighter rampart is discernible for a few feet trending along the east face. The interior surface appears to be at its natural level and un¬ excavated, as also is the space within the outer and middle rampart, except towards the point of contact, where it is hollowed to a trench. 11. Fort, Broomhill Bank Hill.—Some 260 yards to the north-east of the last is another fort on the summit of the hill, at an eleva¬ tion of some 871 feet over sea-level, not visible from its neighbour and commanding a great prospect in all directions. The in¬ clination from the direction of the last fort is slight, but on the west and north the hill falls sharply aw^ay. The enceinte is oval in form, lying with its main axis north and south, measures in diameter igo feet by 170 feet, and is surrounded by a slight parapet mound and a trench partially cut through rock, at most some 6 feet deep and 28 feet wide, with a mound on the counterscarp varying in height as the level beyond rises or falls. There are two entrances, one on the east side of indefinite width overlooking the steep slope to the base of the hill, and the other on the west side, some 10 feet wide, from the direction of the other fort. The interior, which has not been hol¬ lowed, rises at the centre 5 to 6 feet above the level of the entrances. xxxiv. S.W. 23 July 1912. 12. Fort, Blindhillbush Hill.—This fort is situated on the summit of Blindhillbush Hill, at an elevation of 618 feet over sea-level, in an impenetrable fir plantation. It is shown on the O.S. map as oval, with its longest axis north and south, measuring 215 feet by 160 feet. It is surrounded by a rampart of earth and stone rising at most barely 3 feet above the interior, and, in general, having a scarp to the outside some 6 feet in height. At the south-west a bank passes outward from the scarp, with a slight divergence in a south¬ easterly direction, but is soon lost in the cultivated land beyond the wood. xxxiv. S.W. 23 July 1912. 13. Mote, Applegarth.^—The Mote of Apple- garth rises on the termination of a steep bank, APPLEGARTH.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [applegarth. which was no doubt in former days washed at its base on the west by the Annan, though that river now flows through meadow land more than 100 yards away. It is situated to the south of the parish church and within the grounds of the manse, the kitchen garden of which lies on its summit. From the base of the bank on the west the mound rises to a height of 29 feet with a steep scarp, the elevation diminishing as it passes round by the south to the east side to 14 feet, while to the north the height of the summit above a lawn formed on the top of the bank is only some 6 feet. In the latter direction the levels have probably been interfered with, and there is now no trace of the trench which, no doubt, existed here, nor is it possible to say whether a base-court existed on this higher level. Along the east side and round to south, some 6 feet below the summit and 8 feet above the base, is a 6-foot terrace gradually descending to the base level on the north face. This terrace on the east and south appears to be an original feature, but beneath it the mound is faced with a modern retaining wall, and it is possible that the profile has been altered in comparatively recent times. The summit of the mote is circular, measuring in diameter 105 feet from north to south by 116 feet from east to west. xlii. S.E. 2 August 1912. 14. Fort, Millbank.—This fort, which ap¬ pears to be a pure earthwork, is situated on a gentle undulation about J mile west-south¬ west of Millbank Farm, some 2 miles to the north of Lockerbie, and is enclosed and planted with trees. In plan it is circular, with a diameter of some 208 feet, surrounded by a single trench, 35 feet in width and with a depth, where best preserved, of 8 feet below the crests of the scarp and counterscarp. Crowning the scarp is a parapet mound some 18 feet in thickness at base and 3 to 4 feet in height on the interior, while a similar mound surmounts the counterscarp. Near the centre of the north side there is an entrance by a gangway 5 feet wide, crossing the trench at an elevation of 4 feet above the bottom level and carried through the parapet mound by a gap of equal width ; there appears to have been a second entrance from the west, passing inwards at the level of the ground outside into a hollow at the lowest point of the interior. The inner circle of the enceinte has been preserved com¬ plete, but, except towards the north, the trench has passed into land now under cultiva¬ tion and has suffered in consequence. The site, though at an elevation of only 250 feet, commands a fine prospect up Annandale. xliii. S.W. 6 August 1912. 15. Fort, Cumstone Burn.—Some 200 yards west by north of Cumstone farmhouse, on the top of the steep right bank of the Cumstone Burn and some 25 feet above the level of the stream, is an oval enclosure with its longest axis north by west and south by east and measuring interiorly 179 feet by 157 feet. It is surrounded by a rampart of earth and stone some 22 feet broad at base, rising from 3 to 5 feet above the level of the interior, with a concentric trench to the outside carried to the face of the bank of the burn at either end, 22 feet wide and 5 feet below the crest of the ramparts. The situation is at the base of the Bow Hill, and has no great outlook. xliii. N.E. 6 August 1912. 16. Fort, Fir Tree Hill.—This fort is situated on a plateau on the western slope of Fir Tree Hill, at an elevation of 740 feet or thereby above sea-level. It is an oblong enclosure lying with its longest axis north-north-west and south-south-east, measuring interiorly 154 feet by 97 feet, surrounded by a rampart of earth and stone rising some 4 feet above the interior level, with a trench beyond, 26 feet broad and 3 to 4 feet deep below the crest of the scarp, and with a mound on the counterscarp which, on the north-east or higher side, rises 7 feet above the bottom of the trench. The entrance, 5 feet in width, has been from the east, where it passes through the inner mound. It presents a peculiar arrangement. The mound which crowns the counterscarp as it comes round from the nortli is returned across the trench straight towards the opening through the inner rampart, and stops a few feet distant from it, leaving a passage into the trench to the north as well as to the interior. The space left between the return of the mound, where it leaves its regular curve, and the end of the outer 5 APPLEGARTH.] HISTORICAL MONLAIENTS (SCOTLAND) COMMISSION, [arplegarth. mound at its continuance is onh^ some 2 feet. Above the fort, some 50 feet back from the entrance, a broad earthen bank 16 feet wide at base and 3 feet high passes along the hillside and turns away in a south-westerlj^ direction. It is unusuallv massive for a feal dyke, brat it is impossible to say whether it is contem¬ poraneous with the fort, though it is with enclosures, obviously folds, farther to the north. xliii. N.E. 9 August 1912. 17. Fort, Roseburrain.—This enclosure is situated on a plateau somewhat less than J mile to the south-west of the Fir Tree Hill Fort (No. 16). The ground in front of it on the east is level and marshy, while on the south also it is flat. To the north it falls away in a steep gradient for some 60 feet, and to the west, declining gradually for about 40 yards, it drops thereafter sharply to the bed of a burn. The enclosure appears to have been oblong with rounded ends, but the defences to the north and north-west have entirely disap¬ peared, if any permanent rampart ever existed there, while along the west side they are now very slight. Along the south and east there exists a massive mound of earth and stone, with a scarp to the exterior at a very regular height of from 6 to 8 feet and rising from 2 to 4 feet in height on the interior. Where the ground rises towards the enclosure from the outside it is cut through to form a trench. The entrance, which has been wide, is on the east side, somewhat to the north of the centre. At the base of the glacis, leading up to it on the exterior, is an oblong hut foundation, apparently of turf, measuring interiorly 22 feet by II feet ; and in the interior to the right of the entrance is another similar foundation measuring 25 feet by 16 feet. No part of the interior appears to have been hollowed by excavation. xliii. N.W. 9 August 1912. 18. Fort, “ Burrain Skelton,” Cleuchheads Hill.—This fort is placed some 600 feet above sea-level on the top of Cleuchheads Hill and on the west side of the Dryfe valley. It is overgrown with a dense plantation of young fir trees, which makes a survey impossible. The O.S. map shows it on plan to be a long oval with its main axis north and south, measuring some 380 feet by 215 feet. It is scarped apparently all round to a height of from 6 to 8 feet; as far as observable it does not appear to be hollowed by excavation in the interior. This was a beacon hill (see p. xxxiv.). xliii. N.W. 12 August 1912. 19. Fort, near Dalmakethar.—This fort is situated at an elevation of 369 feet above sea- level, about J mile west-north-west of Dalma¬ kethar farm, on the crest of a long grassy round-backed ridge, which lies parallel with the Annan on the east side of the dale and commands an extensive prospect both up and down. On the west the ground declines steeply for some 30 feet, sinking thereafter by an easier gradient to the river ; to the north and south the ridge extends, running level for J mile in the latter direction and dipping to a lower level in the former ; while on the east the surface slopes downwards by an easy gradient. The fort is oval in form, lying with its longest axis north and south, and measures over all some 225 by 175 feet. It has been surrounded by a massive rampart, now greatly reduced and probably much spread, measuring some 40 to 45 feet in width on the south and east. At the north end the mound covers an area 53 feet in breadth, on the top of which is a slight depression ; but whether this is a raised platform within the outer rampart, or a double rampart levelled down, it is not possible to tell without ex¬ cavation. It is unlikely, however, that the defences would be duplicated towards the lower side of the fort and not on the higher. The entrance has been on the east side, con¬ siderably to the north of the centre. The extension of the mound at the north end has reduced the interior to somewhat of a shield form, measuring 132 feet from north to south by 93 feet from east to west. An old road with, locally, a Roman attribu¬ tion, is said to pass near the entrance of this fort and has a place on the O.S. map. The fort itself does not show any features suggestive of Roman castrametation. xxxiii. N.E. 14 August 1912. 20. Fort, Dalmakethar Burn.—About J mile east by north of Dalmakethar farm is another APPLEGARTH.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [applegarth. fort (fig. lo) at an elevation of nearly 400 feet above sea-level and at the edge of a steep bank overhanging the Dalmakethar Burn, which flows by on the north some 50 feet below. East of the fort the ground rises by an easy gradient to the skyline some 300 or 400 yards distant ; to the south it falls away, trending westward ; while on the west it has a rather steep declivity for some 50 feet to a hollow in the cultivated land below. The interior area of the fort is oval, with its longest axis north-north-east and south-south-west, measures 126 feet by 98 Fig. 10.—Fort, Dalmakethar Burn (No. 20). feet, and is entirely surrounded by a rampart of earth and stone. From the edge of the ravine of the burn on the north-east a double trench passes along the east side and the south end, with an intervening rampart broadest and deepest on the south, which diverges from the central enceinte as it passes westward on to the face of the steep slope. Thence it is said to have been continued obliquely to the edge of the ravine. The inner trench on the east has a breadth of 30 feet, and a depth of 4 feet and 5 feet respectively below the crests of scarp and counterscarp ; while on the south it measures 45 feet in breadth, 9 feet in depth below the scarp, and 6 feet below the counter¬ scarp. The outer trench is 23 feet wide on the east and of slight depth, while on the south it has a breadth of 34 feet and depth of 6 feet. The entrances have been from the north- north-east and south-south-east and are 4 to 5 feet wide. The former has been approached over a narrow space flanked by the rampart and the edge of the ravine ; the latter directly through the defences. At both entrances the inner rampart broadens as it approaches the opening from either side. There appears to be a spring in the outer trench at its south¬ west termination. xxxiv. N.W. 14 August 1912. 21. Fort, Dalmakethar.—This fort crosses the neck of a low promontory which projects on the 400-feet contour line on the west of Longerhallis Hill, about | mile south- south-east of Dalmakethar. On the north it overlooks the deep ravine of a burn, and on the south and west it is protected by steep natural slopes. It is now covered by a young plantation, and the only defences traceable are an outer trench 28 feet wide and 5 feet and 3 feet deep below the crests of scarp and counterscarp respectively, running from the edge of the ravine across the neck, with a convex outline to the east ; a rarnpart some 5 feet in height ; a slighter mound 40 feet in rear of it ; and, separated by a shallow trench 17 feet in width, another low mound. The two inner mounds are very slight and noticeable only towards the edge of the ravine. xxxiv. S.W. 14 August 1912. Enclosures. 22. Enclosure, Howthat Burn.—This en¬ closure lies on the lowest slope of the brae, just where it merges into the level ground on the east side of the glen of the Howthat Burn, about I mile east-north-east of Newbigging. It is elliptical in shape, measuring in diameter interiorly 140 feet by no feet,and is surrounded by a mound of earth and stone. There are two entrances 26 feet apart, each about 9 feet wide, which open on the lowest level from the direction of the burn. The interior has been hollowed out to a depth of from 3 to 4 feet on the upper side below the level of the sur¬ rounding ground. At the north end of the 7 APPLEGARTH.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [applegarth. upper side there is a small hut-like recess in the bank, and to the west of it there are low indefinite mounds suggestive of small en¬ closures. The rampart has been broad on the lower slopes, but is much spread, and above the scarp on the upper side is hardly traceable. xxxiv. S.W. 23 July 1912. 23. Enclosure, Ryecastle.—Situated on the crest of a low ridge some 374 feet above sea- level and ^ mile east of Ryecastle is an oval enclosure. It lies on the south side of an old road running north-east from Perchhall, and is partially within and partially withoiit a large wood. With its longest axis north-west and south-east, it measures some 225 feet by 170 feet interiorly, and has been surrounded by a stony bank some 20 feet broad at base, on the sides and crest of which large blocks of stone are in places exposed. The interior has been slightly hollowed by excavation. On the north-east, across an intervening area, is a bank some 15 feet in height overlooking a burn, and from the north face of the enclosure an outer bank covered by a trench 20 feet wide and 3 to 4 feet deep runs to it. Where it impinges on the bank there is a circular de¬ pression measuring 15 feet in diameter, sunk some 3 feet below the natural level and sur¬ rounded by a broad mound, on the face of which are remains of walling. From this hollow a channel, increasing from 3 to 5 feet in width, leads eastward straight down the bank to the edge of the burn, taking a course too steep for a pathway. On the other hand, there is no water channel into the hollow. The periphery is complete at a height of from 3 to 5 feet above the present floor level ; the steep gradient of the channel seems to preclude the idea of the construction having been a lime¬ kiln. It bears a resemblance to the hollow outside the enclosure on the Pyatshaws Rig (No. 304). xliii. N.W. (“ Fort '’). 8 August 1912. 24. Enclosure, Hangingshaw.—On the west slope of a low round-topped ridge some | mile to the east of Hangingshaw is a circular enclosure in an old pasture field, measuring about 100 feet in diameter. It is surrounded by a bank, much spread out by ploughing. some 24 feet broad at base and not above 2 feet in elevation. The entrance is fi-om the west. xxxiii. S.E. (“ Fort ”). 8 August 1912. 25. Enclosure, Dinwoodie.—An oval en¬ closure is situated on the east’ side of the valley, at an elevation of some 370 feet above sea-level and about J mile east-north¬ east of Dinwoodie railway station. It lies with its longest axis north and south, measures from crest to crest 134 feet by 97 feet, and has been surrounded by a stony bank, which is scarcely perceptible on the upper or east side but has an elevation of about 2 feet on the west. The interior has been hollowed by excavation, and lies some 2 feet below the surrounding ground. The entrance, 6 feet wide, has been from the west, and opens on the lowest part of the interior. The site com¬ mands a considerable prospect over Annan- dale. It rests on a deep linn to the south. xxxiii. S.E. (“ Fort ”). 14 August 1912. 26. Enclosure, Mid Hill.—In the hollow which lies between the Mid Hill and Fir Tree Hill are a number of bughts and other ancient enclosures. One of the latter is a circular bowl¬ shaped enclosure hollowed out to a depth of about 4 feet and surrounded by a slight bank. A number of old turf walls run around the base of Fir Tree Hill, connected in some cases with enclosures. xliii. N.W. and N.E. (unnoted). 9 August 1912. Miscellaneous. 27. Construction, Balgray Cleuchheads.— Along the east and south sides of a wooded ravine, 200 yards to the west of Balgray Cleuchheads, runs a bank of earth and stone, making a return northward at its eastern extremity. The construction is noted as a “ fort ” on the O.S. map, but the remains are fragmentary and the designation doubtful. xliii. N.W. (“ Fort ”). 6 August 1912. 28. Heraldic Stone, Dinwoodie Mains.—Built into the front of the porch of Dinwoodie Mains farmhouse is a panel containing in the centre a shield surrounded by strap-work enrichment and bearing in chief two mullets Fig. II. —Caerlaverock Castles (So. 33 (i)). 9 APPLEGARTH.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [caerlaverock. with a human head inverted and suspended by a " woodie ” or rope of withies passed through the mouth. Above are the initials R.M. (Robert Maxwell)/ and, beneath, the date 1631. The stone, according to local infor¬ mation, came from Dinwoodie Castle, which formerly stood near the spot. CAERLAVEROCK. Castellated and Domestic Structures. 33 (i). Old Caerlaverock Castle.—The ruins of the old stronghold of the Maxwells at Caer- ^ Johnstone MSS., p. 42. xxxiii. S.E. 14 August 1912. Sites. The O.S. maps indicate sites as under :— 29. Sibbaldbie Church, Sibbaldbie. xliii. N.W. 30. Monastery, Applegarth Town. xlii. S.E. 31. Market Cross, Applegarth Town. xlii. S.E. 32. Fort, Kirkholm Hill, xxxiii. S.E. laverock and the remains of another castle, apparently its predecessor, are situated on the marshy flats of the Solway, at the mouth of the River Nith, some 9 miles south-south-east of the town of Dumfries (fig. ii). The older site lies within the Castle wood, 200 yards south-south-east of the other and 500 yards north of high-water mark, and is formed in a bed of clay, while the later building stands on an outcrop of rock. The buildings and walls on the older site are demolished, and such foundations as 10 CAERLAVEROCK.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [caerlaverock. remain lie below ground. Many of the stones would be re-used in the later fabric, and even within living memory the site served as a convenient quarry for the neighbourhood. According to Grose (1789) "the site and founda¬ tions ” were in his time " still very conspicu¬ ous.’’ The foundations were again exposed a few years before 1868 {Trans. Dumf. and Gall. Antiq. Soc., 1868-9, P- ^2), but no proper record of the work then done can be traced. M'Dowall, however, reports {History of Dum¬ fries, 1873, p. 74 and note) that they showed the building to have had " a quadrilateral form,’’ and therefore contests the suggestion that this was the site of the castle besieged in 1300, which was triangular {cf. p. 23). He adds that the walls exposed were of " unsub¬ stantial build,’’ and possibly those of " an outwork, erected to defend the dam of the fortress,’’ and so prevent the draining of the moat. But even in an outwork—which, more¬ over, covers an area nearly as great as that of the castle—“ unsubstantial ’’ walls would be unsuitable, whatever may be the precise mean¬ ing here of so indefinite a description. The enceinte is now so overgrown with trees and littered with fallen stone and the debris of the excavation, that little save the general arrange¬ ment of the castle can be traced. On plan (fig. 12) the enceinte is roughly a parallelogram set with the angles to the cardinal points of the compass, and measuring 86 feet from north-east to south-west and 97 feet from north-west to south-east. The scarp was apparently crowned by a curtain wall, slight traces of which are visible on the south¬ west side. At the north angle is a mass of masonry, which slight excavation revealed as the lower stage of a rectangular tower, 17 feet 6 inches broad, projecting ii feet 6 inches from the face of the curtain walls which it Hanked. The masonry is of fine ashlar work, averag¬ ing 12 inches long on face, and is built in courses 8 inches high, which are care¬ fully bedded in clay. A splayed member (possibly the uppermost of a heavy basement course) was found return¬ ing along the face and east side of the tower (fig. 13). The conformation of the debris at Fig. 13. — Splayed Base {No. 33 (i)). the remaining angles of the enceinte indicates towers at these points. Within the eastern angle of the curtain, but clear of these walls, are the remains of a rectangular structure with walls some 3 feet 6 inches thick. An eye-witness of the excavations reported from recollection that the buildings were found to be supported on oak piles driven into the clay solum, and that broken pottery was unearthed and replaced at a point marked X on Plan. Outworks. —The moat has silted up some 2 feet ; under the silt the clay bottom is found to be some 10 feet below the level of the enceinte. The width at base varies from 26 feet on the north to 50 feet on the south. A burn on the east drains the moat and, when dammed, would fill it. The outer rampart, on the summit of which to the north and east is a modern roadway, follows the contour of the enceinte and rises to its level. The outer scarp terminates in a ditch some 15 feet in width. An enclosure an acre and a quarter in extent lies immediately to the north-east and is defined by a continuation of the outer ditch and the burn. The masonry exposed in the north tower and the whole arrangement of the castle in¬ dicate its erection in the early 13th century. 33 (2). Caerlaverock Castle.—The enceinte of the later stronghold, hereafter called by its usual appellation Caerlaverock Castle, is tri¬ angular on plan (fig. 14), with the apex set to the north. The sides and base are enclosed by curtain walls which terminate at the basal angles in a drum tower and at the apex in a gatehouse flanked by drum towers. Between these towers is the entrance. This northern facade gives an impression of great strength and is one of the finest examples of early I5tli- century military architecture in Scotland (fig- 1 . 5 )- The entrance—the outer portion of which is a mid 15th-century addition, as will be ex¬ plained later—is arched and sheltered within an arch-headed recess so formed that the drawbridge when raised would become an extra barrier to the portal. Within this recess, and above the entrance, is a badly weathered oblong panel of 17th-century date II FIRST FLOOR PLAN DRAW MURDOCH 5 TOWER EARLY 15- CENTURY MID 15- CENTURY . LATE 15- CENTURY EARLY CENTURY EARLY ICENTURY, MODERN GROUND FLOOR PLAN lO 0 lO 30 50 4-0 50 60 70 8 0 50 100 110 FEET ^^—i—^^—I—I- ^—I—^—f Fig. 14.—Caerlaverock Castle (No. 33 (2)). 12 Ana't’nl aud Hislorical Monitmeuls—Diitnfi les. To face p. 12. I’lc. 15.—Caerlaverock Castle : North Front and Calehouse (No. 33 (2)). Aili iciit and Historka/ Moiiumcitls — Dumfries. To face p, 13 . Fig. 16.—Caerlaverock Castle : East Wall (No. 33 (2)). CAERLAVEROCK.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [caerlaverock. enclosed by a heavy border. The panel has an escutcheon at each angle, and these are connected by a floral enrichment of oak leaves wreathed. The escutcheons contain armorial bearings as under :— Dexter Chief :—The shield ensigned with a crown bears a lion rampant within a royal tressure for Scotland. Sinister Chief ;—A double-headed eagle displayed beneath an imperial crown—a charge first used by John, 8th Lord Maxwell, Earl of Morton. Dexter Base :—A saltire for Maxwell, impaling a bend between six cross-crosslets fitchy for Mar. Sinister Base :—A fess chequy surmounted of a bend engrailed, for Stewart of Dal- swinton. On the panel is a stag, couchant before a holly bush and supporting between its forelegs a shield charged with a saltire, for Maxwell. At the top of the panel are the initials R.M. —Robert Maxwell, first Earl of Nithsdale (1620)—and at the foot, on an escroll within the border, the motto, I BID YE FAIR. Above the panel is an aperture, slanting upwards as it penetrates the wall, through which passed the chain or rope which raised the drawbridge. It has been much worn by the friction. Over the entrance is a forework of two storeys, beneath which a chamber has been added from which to work the drawbridge and a pair of portcullises. The outer wall of this chamber and the piers supporting it are angled as they near the towers. The piers have two splayed offsets, the lower returning along the face and angles, the upper on the face alone. The beam hole on either side of the entrance immediately over this latter offset, and the raggle on the east jamb of the recess, were formed at a later period. The approach could be enfiladed from tiers of gunloops in the massy flanking towers which rise vertically from a batter at base to a ponderous corbel course—52 feet above the level of the moat—surmounting the gatehouse and bearing the fragments of a machicolated parapet walk. From it was entered a cap house, now ruinous, with turrets corbelled out over the four angles, which command an extensive prospect. The chimney-flues of the tower apartments are conducted into high stalks set on the inner portion of the wall within the parapet walk (fig. 16). The corbelling is carried along the east wall at the same level as on the north to a point 4 feet 6 inches south of the gatehouse. The wall is evidently thus prolonged to protect and cover the south-east angle of the main building. On the first and second floors of the easter flanking tower are windows of considerable size with a south-east aspect. In the other directions there are gunloops. The east wall of the gatehouse is pierced on each of the four floors by a window close to the junction of this building with the tower. South of the gatehouse a range of 17th- century buildings (fig. 17), embodying the curtain and possibly an older building in the same position, runs southward to the extremity of the site where the east wall terminated in a drum tower, now demolished. These 17th-century buildings, of which the northern portion exists in entirety, contain three storeys beneath the wall-head, and ter¬ minate at this level in a cornice of pseudo- corbelling 9 feet below the corbelling of the gatehouse (frontispiece). The greater part of the south curtain wall has been demolished to within 3 feet of the ground. One portion, however, east of the centre, still stands to a height of 25 feet. There has been a postern in this wall at its junction with the west tower. The west tower of the base is contempor¬ aneous with those of the gatehouse. It is entire to the wall head, and is surmounted at a height of 40 feet above the moat by a corbel course similar to that of the gatehouse. Be¬ neath this there are four storeys, illumined by small windows ; those on the first floor, set to the north and east aligning the curtains, have drains in their sills, consisting of open chases splayed at the bottom, to cast sewage into the moat. On the second floor a window is corbelled out over the angle formed by the junstion of the tower and west curtain (fig. 18). The west curtain shows indications of alterations. The lower part of the wall to a height of 21 feet is excellent ashlar work, at 13 CAERLAVEROCK.] HISTORICAL MONLAIENTS (SCOTLAND) COMMISSION, [caerlaverock SECTION THRO' WEST BASAL TOWER SECTION LOOKING NORTH WEST BA 5 ALtower DOMESTIC BUILDINGS GATEHOUSE 15^ CENTURY LATE 15 If CENTURY 15^1 CENTURY LONGITUDINAL SECTION LOOKING WEST 10 20 30 40 30 60 70 60 90 100 llO FEET Fig. 17.—Caerlaverock Castle (No. 33 (2)). 14 Aneitnt and Historical Monuments — Dumfries. To face p. 15. Fig. 18. —Caerlaverock Castle : West Curtain and Base Tower (No. 33 (2)). CAERLAVEROCK.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [caerlaverock. least contemporaneous with the towers, and bears indications of openings filled in at some later period. The upper part, hve feet higher, is inferior work of rubble, similar to that employed in the east wall of the late 15th- century range of buildings built against and incorporating this curtain. The curtain was originally finished with a parapet walk borne on corbels, as the third floor of the west basal tower can only be entered from a doorway set on the line of the curtain with a sill 30 feet above the moat. There is a projecting flue, borne on corbels 6 feet north of the west tower, which conducted sewage to the moat from a garderobe on the parapet. The south-west angle of the gatehouse is pro¬ tected, as on the south-east, by an extension of the lateral wall southward, surmounted by corbelling similar to that on the north and east. A drain from the garderobes situated on the upper floors of the gatehouse is corbelled out over the angle at the junction of the west flanking tower and the main building. The west flanking tower is slightly greater in diameter than that on the east. The batter at base is almost imperceptible and I foot 7 inches lower. Interior. —The ward is entered by a stone- roofed pend or trance which passes under the gatehouse, giving access en route, through a doorway on either side, to guardrooms com¬ municating with the basement of the flanking towers (fig. 19). The pend was originally defended at the outer end by a portcullis and door, and terminated in an archway in the south wall of the gatehouse. When the out¬ most portcullis room was inserted {cf. p. 13), c. 1450, additional defences, consisting of—in order from the north—an iron gate, a port¬ cullis, an inner door, and a second portcullis, were provided in front of the original entrance. At a later period, c. 1500, the archway at the inner end of the trance was contracted and a rear room or gallery with portcullis erected over it at the level of the parapet. In this way the gatehouse could be isolated—there being no internal communication between the base¬ ment and upper floors—in the event of a besieg¬ ing party gaining the ward through the postern or a breach in the curtain. The guardrooms are irregularly shaped and are lit by slits in the south wall of the gate¬ house. The east chamber has an additional and larger window in the east wall. Both are provided with fireplaces in the lateral w^alls. These chambers on the basement floor have stone vaulted ceilings ; in the tower rooms the vaidts have fallen in, but were apparently shaped like a bee-hive and groined, wEere necessary, at the doors and windows. The first, and the three upper floors (fig. 20) of the main body of the gatehouse contained originally one large apartment on each floor, but were divided in the i6th century by a partition wall in which additional fireplaces were inserted. On the first floor an archway in the north wall, now built up, gave access to a recess over the earlier entrance, within which the mechanism for working the original port¬ cullis was placed. In the north wall of this recess is a narrow loophole or chase some 4 inches wide and 9 feet long, widening to a spade-like shape at the sharply-splayed base, through which quicklime or other offensive material could be poured on intruders, should they attack the portcullis. In the south-west angle of the gatehouse a small wheel-stair—originally the only access— communicates with the upper floors and the parapet walk. There is a shelved recess in the west wall for the purpose of containing two wooden lockers or cupboards, and a smaller one in the north wall. The windows and doors of the south wall of the gatehouse have been altered or inserted, as the details of their jambs show, when the gallery and its piers were built. It is not clear how access from the basement to the first floor was obtained before the later wheel-stairs outwith the gatehouse were built. In all probability there was a moveable wooden stair or ladder to the west against the south wall. In this wall, near the angle formed by it and the west wall, is a semicircular relieving arch, subsequently filled in, which suggests that such a stair may have led to an entrance within this arch. In each of the towers, at this level, is a chamber with windows anterior in date to the forebuilding, as the openings commanding the entrance are obscured by the later work. The east chamber has a fireplace and garderobe on the east side and the west chamber a garderobe in the thickness of the wall at the south-west angle. The second floor and the storeys above in the main building of the gatehouse had 15 Fig. 19.—Caerlaverock Castle (No. 33 (2)) 16 CAERLAVEROCK.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [caerlaverock. THIRD FLOOR PLAN EARLY 15- CENTURY MID CENTURY LATE CENTURY EARLY 1^^ CENTURY EARLY 17- CENTURY MODERN . SECOND FLOOR PLAN 10 ZO 30 4 0 5 0 60 70 80 90 100 110 FEET ^^^^—[ - T— t - I —^^ IP Fig. 20.—Caerlaverock Castle (No. 33 (2)). 17 2 CAERLAVEROCK.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [caerlaverock. floors of wood carried on beams borne on corbels projecting from the north and south walls. On the second floor the general arrange¬ ments are similar. An access is provided through the original north wall to a chamber beneath the forework, from which the later portcullises and the drawbridge were worked. The grooves for the windlasses and beams by means of which these were hoisted still survive on the reveals of the small window illuminat¬ ing the chamber and in the lateral waUs. The chambers on this floor in the towers have domed ceilings of stone—the western furnished with ribs of early 15th-century type, meeting at a central boss shaped as a shield. On the third floor there are no rooms in the towers. Access to the forework is obtained by an angled passage contrived within the west wall and proceeding over the haunch of the vault of the west tower. The fourth floor, at the level of the parapet walk, appears to have been renewed when the rear gallery was built. This structure, within which the mechanism of the south portcullis is placed, is entered from the west apartment in the main building. These apartments com¬ municated ^vith rooms which were situated in the roofs of the towers. Above the forework is a cap-house two storeys high, entered from the parapet walk, which returns along the north fagade and the southward extensions of the lateral walls. It was also carried across the south wall of the gatehouse until the later west staircase and the rear bartizan intervened. The west drum tower of the base is known as “ Murdoch’s ” Tower, as therein Murdoch, Duke of Albany, is said to have been incar¬ cerated before his execution at Stirling in 1425. Although of smaller dimensions, it is evidently contemporaneous with the gate¬ house towers. There is only one apartment in each of the four storeys. The basement would be used as a lodge for the porters of the postern, and was entered through the southmost chamber of a coeval range of buildings against the west curtain. This apartment and the two storeys above alone were retained when the present block on the west was built. The first and second floors of the towers were entered from the floor above this apartment and the third floor from the parapet walk along the curtain. These floors were of wood, and have long been demolished. The block of building against the west curtain (fig. 21) was erected towards the end of the 15th century, replacing an older range in the same position. The only traces left of these older buildings are sundry openings in the curtain now built up and the beam-holes and corbels for the floor joists of the southmost apartments. The present range contains two storeys beneath the wall-head ; there are traces of a garret within the roof, reached from the first floor by a wheel-stair, now demolished, at the south-east angle. On the ground floor there were three apartments, each with its entrance from the ward, and on the upper floor two. The windows, with the exception of one on the first floor, which has been inserted in the west curtain, look out eastward to the ward. These have, like the doorways, moulded jambs and lintels. The first-floor windows are of considerable size, and were divided by mullions and transoms. The wall-head is surmounted by a narrow cornice, and from this level there rise massive chimney- stacks terminating in moulded copes ; on the north skew-put is a shield charged with a bend sinister. The fireplaces are of a type common in this period (fig. 22) ; the jambs consist of three filleted rolls with hollow interspaces terminat¬ ing in moulded bases and capitals following the contour of the jambs under a high lintel sur¬ mounted by a narrow cornice. In one example the jamb is enriched with a continuous floral ornament. There being no internal communication between the ground and first floors, the latter was apparently entered from the ward by a wooden stair at the north end leading to a fine doorway with moulded jambs and lintel, now obscured by the later 16th-century stair¬ case built between this wing and the gate¬ house. This staircase is wide and well lit. It gave communication from the ground to the parapet walk and to each of the floors of the keep through doorways which, like the win¬ dows, have jambs and lintels ornamented with a quirked bead-and-hollow moulding of the period. The little court formed by this staircase, the gatehouse, and the curtain, was at this time provided with galleries, sheltered Ancient and Hisloricat ISIotuinients—Dmnfi ies. T<'face t, iS. Fk; 21 .—Caetlnverock Castle : Interior fiom the South (\o. 33 ( 2 ' ) 17 - CENTURY FIREPLACE EAST WINCi 12 9 6 I FT. I I I I I [ SCALE. FOR DETAILS LATE 15- CENTURY FIREPLACE, WEST WINQ 12 60 1 234-56789 FEET Fig. 22.—Fireplaces and details, Caerlaverock (No. 33 (2)). CAERLAVEROCK.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [caerlaverock. by a pent-house roof, which were entered from doorways in the staircase. The gallery which adjoins this stair is con¬ temporaneous. It projects over the pend, and is borne on a segmental arch set on lofty piers. On the ingoing of the west pier a date, apparenth' 1595, is roughly incised some 4 feet 6 inches above the ground. The soffit of the arch is moulded and is pierced by the slot in which ran the portcullis, which was worked from the galleiy. The mouldings of this arch and of the windows inserted in the south wall of the gatehouse, vdthin the recess formed by the piers, are of a common early 16th-century type. Against the east and south curtains were probably subsidiary buildings, which were converted in the early 17th century into the principal residential apartments in accordance with the enhanced requirements of that age (fig- 23). The courtyard fa9ades of these two wings must in their entirety have formed an ex¬ quisite little Renaissance composition, so admirable is the proportion and grouping of the surviving architectural detail. Un¬ fortunately only a small portion—that abut¬ ting against the gatehouse—stands complete to the wall-head. The remainder is less than one storey high. The setting out is symmetrical ; the small windows at the northern end of the east ving being repeated on the south. Be¬ neath the wall-head, which is surmounted by a moulded cornice, were three storeys, and an attic was contained within the steeply-pitched roof. The chimney-flues are taken into high square stacks set diagonally on their seating. The vdndows are extremely ornate ; those on the ground floor have moulded and fluted architraves. On the upper floors the jambs have slight rusticated engaged shafts, which rest on little corbel bases and terminate in Ionic capitals beneath cornices and pediments, segmental and triangular, containing heraldic achievements and representations of subjects from classical mythology. The hall occupied the eastern portion of the south vang, and is reached from the courtyard by an arched entrance projecting slightly from the face of the wall. The doorway (flg. 24) has splayed jambs, daintily moulded, termina¬ ting in stops at the step and in foliaceous carving under heavy imposts from which the arch springs. The mouldings of the arch are slight, but are enriched with the egg-and- dart, bead-and-spinnel, and other motifs. A recessed circular panel occupies each spandrel. The rude pilasters on each side of the door¬ way are modern ; the bases and east capital have apparently been taken from some other portion of the building. The keystone of the arch is incomplete. The original keystone would be shaped as a console, bear¬ ing the projection of the architrave, frieze, and cornice, which occurs here and over the pilasters. There are two windows east of the entrance. The nearer is incomplete and has fluted archi¬ traves. The farther, a blind window, is more elaborate and has fluted jambs and a moulded architrave and cornice, between which is sculptured, on the frieze, a cherub’s head with widely distended wings. The hall has been a noble apartment 17 feet high, as a moulded cornice on the east wall shows. The jambs of the doorways and windows were ornamented with bead-and- hollow mouldings. A partition on the west separated the hall from a withdrawing-room. A doorway in the south-east angle opens on a wheel-stair, through which the east basal tower was entered. The fireplace is situated in the north wall, in front of the blind window. The projecting jambs, shaped like trusses or consoles, are almost buried beneath the ground. East of the fireplace a lofty archway gives access to the main staircase, which is con¬ tained within the east wing, communicating with the first-floor apartments of that and the south wing. A doorway beside the entrance to the staircase leads under the upper flight of steps to the kitchen offices in the basement of the east wing. These consist of a well- room and bakery, a kitchen and a servery, all with barrel-vaulted ceilings. The two former chambers had separate entrances from the courtyard. A service wheel-stair is contained within the angle of the gatehouse and the east pier supporting the gallery, and communi¬ cates with the upper floors of the east wing and, by doors inserted in the south wall, with the gatehouse. The first floor is entered from a doorway off the main staircase leading through the well of a wheel-stair which ascends to the floors above. 20 ELEV/ATION OF EAST WING. SECTION ()7- century) lO S O lO 20 30 FEET Fig. 23.—Caerlaverock Castle (No. 33 (2)). 21 CAERLAVEROCK.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [caerlaverock. It contains two large apartments, lit by windows to the courtyard and others in¬ serted in the east curtain. The fireplaces of these chambers have projecting jambs, which are shaped like consoles. The fronts are moulded and in one fireplace ornamented towards the base with a fleur-de-lys. A wide lintel spans the jambs, and over it is a slight moulded cornice. The upper floor is similarly arranged. Dimensions. —The enceinte measures 130 feet from north to south ; the basal curtain is 137 feet long and the lateral curtains are III feet—all measured between the towers. These curtain walls vary in thickness from 4 feet 9 inches at the sides to 7 feet at the base. The gatehouse measures 60 feet along the south wall and 38 feet from north to south. The flanking towers have an ex¬ ternal diameter of some 25 feet, with walls averaging 4 feet 9 inches thick. The external diameter of Murdoch’s Tower at the south¬ west angle is 20 feet, and the wall is 5 feet in thickness. The west wing projects 19 feet from the west curtain and is 64 feet long. The wall is 3 feet thick and the partitions rather less. The west staircase has an internal diameter of 10 feet. The east wing projects 20 feet from the east curtain, and the wall is 3 feet 6 inches thick. The haU was 66 feet long and 25 feet 6 inches broad. Chronology. — The gatehouse, flanking towers, basal towers, and curtain walls are contemporary, and were erected in the early 15th century ; the room under the forework is later, probably mid-i5th century. The west wheel-stair, the rear gallery, the mid¬ partition, the floor above the parapet walk, and the windows in the south wall of the gate¬ house all date from the early i6th century. The buildings on the west side of the enclosure date from the end of the 15th century, and those on the east and south sides from the early 17th century. Condition. —Although of late years a con¬ siderable amount of gradual repairing has been undertaken, the present condition of the buildings is most unsatisfactory. The easter flanking tower is seriously rent, and calls for immediate repair. The wall-heads of the buildings, and in particular of the gatehouse, are covered with vegetation, under which the masonry must be in an extremely bad state. The circular staircases are bereft of the majority of their steps, and in the absence of these ties must ere long collapse. The west wing is slowly disintegrating ; the safe-lintels are falling in. The 17th-century buildings are fairly sound, and, if the cornices were denuded of vegetation, re-pointed, and weather-proofed, little else would be required. A serious fissure in the south gable, which latterly threatened the stability of the structure, is now tied in, and the vault under the gable supported by a pier at the south end of the kitchen. Outworks. — The enceinte is surrounded by a moat 42 feet wide at the north and 80 feet at the south, which is girt by a rampart 30 to 40 feet thick at base and 10 feet 6 inches at highest above the moat, which is 10 feet deep. On the south the rampart broadens and encloses a terrace 30 feet wide. The moat is drained by a sluice at the south-west angle. The swampy nature of the ground to the east, west, and south obviates the need of outer defences in these quarters. To the north, however, the ground rises and is firm. On this front there is an outer ditch and scarp, terminating at either end in the swamp and traversed by the pathway to the castle, which leads through a base-court, 3 acres in area, to a wide semicircular arched gateway of the i6th century, 10 feet 4 inches wide and II feet 9 inches high, with chamfered jambs. Historical Note. —The record of Caer¬ laverock Castle is complicated by the fact that no distinction is made or suggested between the present ruins and the buildings which must have stood on the strongly en¬ trenched site some hundred yards to the south. Yet, if the latter site is not that of an earlier castle of the same name, of what castle is it the site ? The earliest reference of importance to Caerlaverock seems to be of October 1299, when it is reported to King Edward I., from Loch- maben, that “ There is a castle near them, called Carlaverock, which has done and does great damages every day to the King’s castle (Lochmaben) and people.” They had, how¬ ever, scored a success, and the head of the 22 Ancient and Historical Monuments — Dumfries. To face f. 22. Fig. 24 .—Caerlaverock Castle: Entrance to Hall (No. 33 ( 2 )). CAERLAVEROCK.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [caerlaverock. Constable of Carlaverock “ was now set on the great tower at Lochmaben.” This experience just precedes what constitutes the best known and most remarkable episode of the castle’s existence—the siege of July lo-ii, 1300, directed by Edward I. in person. This episode was made the subject of a contem¬ porary poem, Le Siege de Karlaverok, in which details are given of the blazoning of arms of each of the 87 English “ companions ” or leading knights. This number of knights- banneret implies something under 2000 lances ; the poem gives 3000 ; but even the smaller number is probably over the mark. The army was, of course, a field army for a cam¬ paign and the siege merely an incident in the operations. Edward arrived before the place on the loth July, to find that the garrison were in mind to offer a stout resistance. He had to summon his siege-engines, some by ship, some from Lochmaben Castle, and with their battering the castle was reduced. The garri¬ son numbered only sixty survivors, but they had inflicted severe losses in men and horses upon their assailants, with apparently small loss to themselves. Walter Benechafe, the constable, and “ eleven other Scots, his fellows,” were sent to prison at Newcastle. Edward was back at Dumfries on the i6th. The chief interest is in the description of the castle and its position. In shape the building was like a shield {Com uns escus estoit de taille) ; that is, triangular, like the ” heater ” shield of the time, for, it is ex¬ plained, ‘‘ it had only three sides round about, and in each angle a tower ; but one of these (towers) was double, so high, so long, and so large that underneath was the gate with a drawbridge well-made and strong, and other defences in sufficiency. It had good walls and good ditches, quite full to the brink of water.” On the situation of the castle there is this ; It was beautiful, ” for on one side, towards the west, could be seen the Irish sea (the Solway), and to the north a fair country surrounded by an arm of the sea, so that on two sides no creature living could approach it without putting himself in danger of the sea. Nor is it easy to the south, for the many ways are made difficult by wood, by marsh, and by trenches filled by the sea where it is wont to meet the river {trenchies La ou la mere les a cerchies On seiilt la riviere encontrer) ; and, therefore, it was necessary for the army to come towards the east, where the hill slopes.” This description of the site is fairly general, and, as the castles are but a few hundred yards apart, is thus applicable to either. Similarly the description of the castle itself quite suits the present building, but that may have been constructed on the lines of the older one, the site of which is lozenge-shaped. The tests are not decisive : the first requisite is excavation of the site to determine the ground plan of the structure ; the work on it alluded to above (p. 11) appears to have been of a random nature. Another contemporary chronicler, describ¬ ing this campaign, speaks of its only result as the capture of a ” poor little castle ” {povere chastelet, Langtoft). Caerlaverock was one of the castles seized b}^ Robert Bruce after the murder of Comyn in February 1306, but soon recovered by the English. At the close of May in that year there was a garrison in the castle of eight men-at-arms and twenty foot archers. In 1312 the castle is still in English possession, with Sir Eustace Maxwell as keeper. Caer¬ laverock was the principal seat of that family. Sir Eustace turned to the national side, with the result that the castle had again to suffer a short siege, which, however, was unsuccess¬ ful. In the end the castle suffered the fate which Bruce had determined upon for such fortresses; it was levelled to the ground {pro fractione et pro prostratione castri de Carlaverok ad terram), and its owner, Eustace Maxwell, received compensation in reduction of the annual of £22 sterling, due to the Crown for these lands, to £12.'^ This is the first destruc¬ tion, but how far these destructions went is problematical. Caerlaverock was probably reconstructed, with other important castles in Scotland, during the English occupation under Edward III. The Maxwell of the time. Sir Herbert, a nephew of Sir Eustace, made his submission to that King, and surrendered the castle to English keeping in 1347 ; but early in 1356 it was beseiged and captured for the Scots by Sir Roger de Kirkpatrick, who levelled it to the ground {ad solum prostravit).- This is the second destruction. Apparently in the early part of the 15th 23 CAERLAVEROCK.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [caerlaverock. century a new Caerlaverock was being raised. A little after the middle of the century, Robert, second Lord Maxwell, is credited vdth ha\'ing “completed the bartizan* of Car- laverock.” 3 Thus, in the conflicts of the 16th century the castle is again a place of interest. In the autumn of 1545 negotiations were being carried on vdth Lord Maxwell for its transference to English hands, which ulti¬ mately^ occurred. For the purposes of Henry VIII. the castle was regarded as having the advantage of being accessible by sea. The Scots again recovered it, but Maxwell’s support of Queen Mary brought an English force in 1570 under the Earl of Sussex to harry the district, and Caerlaverock is included in the list of castles which that Earl reported he " threw down.” ^ But in 1593 we find that the Catholic Maxwell has “ many men work¬ ing at his house, five miles from Dumfries.” ® So repairs were in hand again. The final reconstruction dates from the time of Robert, first Earl of Nithsdale, in the first quarter of the next century, when the Renaissance wing to the east and the southern work were added. In 1640 the place was beset by the Covenanters, and, after the longest siege on its record—three months and a week—fell for the last time and was dismantled. Le Siege de Karlaverok (Harris) ; Calendar of Documents relating to Scotland, vols. ii., iii. ; The Book of Caerlaverock, vol. i.; Grose’s Anti¬ quities, vol. i.; ^Reg. Mag. Sig., i. (1912), p. 456; - Scotichronicon, Lib. XIV. cap. V. ; ^ Book of Caerlaverock, i. p. 56 ; ^ Cal. Scot. Papers, ii. p. 327; “ Calendar of Border Papers, i. p. 470. Ixi. S.W. October 1914. 34. Bankend or Isle Castle.—This ruined castle occupies a spit of marshland on the west bank of the Lochar Water, by which it is almost surrounded on three sides, while indica¬ tions of a ditch remain on the south-western side. The site is near the south-eastern ex¬ tremity of the Lochar Moss, distant some miles by road from Dumfries and about 2 miles to the north of Caerlaverock Castle. The building is of the type known as the T plan, * Apparently the battlementing, “bartizan” or “bertisene” being, by metathesis, for “ bratticing ” ; but sometimes, in the 17th century, applied to a high castle wall. consisting of an oblong, measuring some 221 feet by 15-I feet, within walls fully 3 feet in thickness, and a staircase wing in the centre of the north-west wall with a projection of 9 feet 6 inches and a width of about ii feet. The doorway is on the ground level at the eastern re-entering angle, and has a deep bar-hole formed in the north-western jamb, also a lamp- recess adjoining it at the stair-foot. The ground floor appears to have been vaulted and defended on three sides by circular loopholes widely splayed to the exterior and square within, while on the north-east side facing the Lochar Water is a window measuring about 18 inches in width. The castle is now a complete ruin ; almost the whole of the south-east and south-west wall has fallen, but the remaining fragments of the staircase wing and of the north-west wall indicate that it was originally four storeys in height. The entrance doorway appears to have been defended by a bretasche supported by moulded corbels placed near the level of the waU- head. On the north-eastern wall of the staircase wing is a panel containing the arms and initials of Edward Maxwell and Helen Douglas, his wife, with the date 1622 carved in relief. “ The Bank Ende ” was selected in the Military Report of some date between 1563 and 1566 as one of the places which would strengthen an English occupation of the district.^ “ It will havand Annande forti- fyed . . . may (make) that way to Drum- freis for Englande to be free, and bring all Nythisdale in subjection. It is a straite passage, and may be well kept being ones fortifyed.” So important was its position considered that Caerlaverock was to be relegated to the position of “a garrisone assistant.” 1 Armstrong’s Liddesdale, App. Ixx. p. cix. Ixi. N.W. August 1913. Defensive Constructions. 35. Fort, Ward Law.—The Wardlaw Hill, which rises to a height of 313 feet over sea- level, overlooks the Castle of Caerlaverock, from which it is distant about i mile, and commands also a prospect over a great extent of surrounding country. It is surmounted by 24 CAERLAVEROCK.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [caerlaverock. an oval fort (fig. 25) surrounded by a rampart of stone and earth, with a terrace or trench now filled in, before it, some 18 feet broad, having a mound on the outer edge or counterscarp. The enceinte has its longest axis north and south, measures some 210 feet by 180 feet, and rises in elevation towards the north. The rampart along the north arc at the edge of the interior is scarcely perceptible, but around the lower part of the periphery it Fig. 25.—Fort, Ward Law Hill (No. 35). has a height of 3 feet 6 inches or thereby on the inner side, and a ramp some 8 feet in height to the terrace, which lies at a general level of 6 feet above the ground outside. There is an entrance from the west some 5 feet in width. On the north arc, in front of the highest point of the fort, and where the parapet is not observable, a slight mound is carried along the terrace some 15 feet out and 4 to 5 feet back from the edge, is brought forward to the edge as it passes eastward, and eventually merges in the inner mound beyond the pro¬ minence to the north on the east side. Ixi. N.W. 24 July 1912. 36. Fort, Craig Wood, Highmains Hill.—On Highmains Hill, and within the Craig Wood, to the south of Craig and some f mile to west by south of Bankend, are the remains of a curvilinear fort. The hill rises abruptly from the north, and slopes away from its highest level on the south-east by an easy gradient to the north-west. From a point adjacent to the steep face on the north a bold rampart, some 20 feet wide at base, curves segmentally across the summit where the ground commences to decline towards the west, thence disappear¬ ing in the slope. To the outside the rampart has a height of from 4 to 5 feet, and is covered by a slight trench. On the inner side it merges gradually into the natural slope of the ground. Ixi. N.W. 24 July 1912. 37. Kelwood “ Mote,” Bailie Knowe.—Situ¬ ated immediately to the south of the glen of Kelwoodburn, and some 200 yards east of Kelwoodburn cottage, are the remains of a small circular fort. It lies on cultivated land, and appears to have been surrounded by a single rampart and trench. The interior, which is on falling ground, is somewhat basin-shaped, dipping towards the centre from the surrounding rampart. The contour of the rampart is now rather indefinite, but the diameter of the enclosure appears to have been some 150 feet. Ivi. S.W. 25 April 1913. Sites. 38. Earthwork, Ward Law.—About 300 yards north by west of Wardlaw Fort (No. 35) is the site of an earthwork. It has been almost entirely obliterated by the plough, and only at the north end does any trace of it remain, and that a short ill-defined stretch of straight rampart with a slight depression in front of it to indicate a trench. Were it not for a dip in the field dyke this fragment might escape observance. Ixi. N.W. 25 April 1913. 39. “ Earthwork,” Blackshaw.—The O.S. map marks an earthwork on the south side of a farm road and about I mile east of Blackshaw. 25 CAERLAVEROCK.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [canonbie. No trace of this is to be found. The field in which it is situated lies behind the farm of Neu'field. Ixi. S.W. 24 Jul\' 1912. The O.S. maps also indicate sites as under ;— 40. St Columba’s Chapel and Well, about I mile south of Glencaple. lx. N.E. 41. “Kilblain,” South Kilblain. Ixi. N.W. CANONBIE. Ecclesiastical Structure. 42. Tomb, Churchyard, Canonbie.—A frag¬ ment of 13th-century ecclesiastical work is preserved vithin the churchyard of Canonbie Parish. It lies to the south of the modern church and vdthin recent years has been utilised in the construction of a tomb en¬ closure—that of the Rev. James Donaldson, late minister of Canonbie. It consists of a recess, 4 feet 8 inches wide and 4 feet 4 inches high, with a segmental arch of roll-and-hollow mouldings enriched vith a dog-tooth ornament. These mould¬ ings are carried down the jambs to a stop at a sill; above is a hood-moulding terminat¬ ing at either side in label-stops—one knotted and zoomorphic, the other floriated. The back appears to be modern. From its posi¬ tion in a portion of the south wall of the old parish church it has been suggested that this fragment was the sedilia, although the design is unusual for such a construction. Several graveslabs of no great interest, dating from the end of the i6th century, are built into the north wall of the churchyard near the entrance. liii. S.E. Visited 27 March 1915. Castellated and Domestic Structures. 43. Hollows Tower.—This tower (fig. 26) is situated on the right bank of the river Esk, midway between Langholm and Canonbie. It has been completely defended by the river bank to the north and partly by the sloping marsh¬ land to the west. On the eastern side, where the ground is level and unprotected by nature, the tower was probably enclosed originally by the walls of an outer courtyard. On plan (fig. 27) the building is oblong and measures some 23 feet 2 inches by 15 feet 3 inches within walls averaging 6 feet in thickness, and the total height from the step at the entrance to the top of the corbel-course measures nearly 40 feet. The doorway is at the south end of the west wall and gives access to the wheel-stair, which has com¬ municated directly with the upper floors and with the parapet walk. Originally it was equipped with a strong outer door and an iron yett, neither of which now remains. The wheel-staircase projects on the interior floor space. The ground floor is vaulted and GROUND FLOOR PLAN FIRST FLOOR PLAN 10 5 O 10 ao JO FEET Fig. 27.— Hollows Tower (No. 43). lighted by narrow shot-holes, the north wall having two such openings, one above the other. On the first-floor level is the hall, measuring about 24 feet 2 inches by 16 feet 3 inches and having a window with stone seats in the east and west walls. A wide fireplace with moulded jambs is formed in the north wall, with an aumbry on each side, and there is a narrow opening to the south. The two upper floors and the attics have each also consisted of a single apartment. The stone corbels for carrying the floors remain. There are no fireplaces above the level of the second floor. The corbel-table, which has supported the stone parapet, is of the ornate type charac¬ teristic of the i6th century. The upper member consists of a bold cable-moulding, returned at intervals where gargoyles have occurred. Below this is a continuous band. 26 Ancient and Historical Monwnents — Dumfries. Fig. 26 .—Hollows Tower (No. 43 ). To face p. 26 CANONBiE.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [CAKONBIE. decorated with a series of projecting and sunk enrichments, and the lowest member takes the form of a simple double roll. A circular turret, resting on corbelled projections, is constructed at each angle. A feature of this tower is a watch or beacon stand corbelled out at the apex of the south gable; cf. the analogous structure at Elshie- shields (p. 155). A house was built at Hollows (Hole-house, Hollace, Hollas, Hollis, are some other forms) in or soon after 1518 by one of the migrating Armstrongs (see Introd., p. xxxv.), and burnt by Dacre in 1528. The present building, how¬ ever, in its upper portion at least, seems to be of later date. It requires only a roof to assure its preservation. Spiral-Marked Slab. — The sill of the doorway into the vaulted chamber in the basement of Hollows Tower is a slab of sandstone (illustrated in the Introduction, fig. 2), measuring 3 feet in length by i foot 7 inches in breadth at the centre, which is incised on its surface with spiral and other markings. At the upper and slightly narrower end of the stone, and towards the outside, is a spiral figure, consisting of two complete turns and half of a third, the line thereafter passing divergently across the stone to the other side, being surmounted near the middle of its course by a single key-like symbol or ornament. Between the free end and the spiral is a single incised line which may have been connected with it. Immediately below the spiral there is visible a semi¬ circular incised line, and at the lower end of the stone, partially hidden by the architrave of the door, is another and smaller spiral with certain indefinite markings springing from it at one side. The stone is much worn, and the figures are now probably incomplete. The marks on the lower corner, opposite to that on which the spiral appears, are natural inequalities of the surface. liii. S.E. ("Gilnockie Tower”). 18 July 1912. 44. Auchenrivock Castle.—This fragment stands at a considerable elevation on the south side of the main road to Canonbie and some 3 miles from Langholm. It is built of irregular boulders and now forms part of the garden wall to the north of the adjoining farm-buildings. It measures 33 feet 3 inches from north to south over walls averaging 4 feet in thickness and 7 feet in height. The north and south walls are respectively 13 feet 6 inches and 10 feet in length. Shot-holes, one in the north wall and another in the west wall, with splayed outer and inner jambs and with circular openings some 4I inches in diameter, are the only features now remaining. The inner surface of the western wall is very indefinite, but it seems probable that the basement was vaulted. This place, near the Irvine Burn where it falls into the Esk, was of old known as Stakeheugh,^ and was the original seat of the Irving family. In October 1513 Sir Christopher Dacre burned ” the Stakehugh, the manor place of Irewyn, and the hamlets down Irewyn Burn.” - ^ Langholm as It Il'as, p. 353 ; 2 Letters and Papers, For. and Dorn. Henry VIII., i. No. 4529. liii. N.E. 4 July 1912. Defensive Construction. 45. Roman Camp, Gilnockie.—Immediately in rear of the farm cottages at New Wood- head, and about J mile due north of Gilnockie railway station, is a large rectangular oblong enclosure with rounded angles, lying partly on grass land and partly within a wood, which presents characteristics of Roman castrameta- tion (fig. 28). The site is a plateau rising gently on all sides to an elevation of some 390 feet above sea-level, not in itself very prominent or greatly exposed, but commanding an ex¬ tensive prospect over a wide area of country in all directions. The fort lies with its main axis north-east and south-west, and has measured within its defences some 1450 feet by 750 feet, or about 25 acres. It has apparently been surrounded by a single ditch and ram¬ part, the former having now from crest to crest a width of some 18 feet and the latter a breadth at base varying from 20 to 30 feet. Along the north-east end the vallum across the grass field, though much ploughed down, is easily traceable. On the south-east flank from the east angle for a distance of 220 feet it is no longer to be seen ; but there¬ after across an adjacent meadow through a young plantation and an old wood it can be 27 CAXONBiE.] HISTORICAL MONUxAIENTS (SCOTLAND) COMMISSION. [canonbie. followed \Hth ease, till it makes a return in a north-westerl}^ direction close to the railway from Riddings Junction to Langholm. With the railway line it gradually converges, and is eventuall}' lost beneath it. On the north¬ west flank it has been great^^^ interfered with in the formation of a road, ditch, and hedge, which more or less occupy its position, leaving 100 O 100 200 300 400 500 600 FEET. ^^^^1 Fig. 28.—Roman Camp, Gilnockie (No. 45). it only partially and intermittently recog¬ nisable. On the south-east side, at 520 feet from the east angle, is a weU-defined entrance some 72 feet in width, which is covered at a distance of 36 feet in front by a traverse, a mound 56 feet in length and 22 feet in breadth at base at centre, tapering slightly to each end and 3 feet 7 inches in elevation, with a ditch at its base on the outer face. At 126 feet to the north-east of this entrance is a gap in the vallum 10 feet wide, which, if not original, does not seem to be modern. At 400 feet further to the southward, and 478 feet from the south angle, is another entrance 40 feet vide, which is likewise covered 30 feet in front by a traverse 50 feet long and 33 feet broad at the centre, tapering slightly to either end and 3 feet 6 inches in elevation. Any entrance which may have formerly existed through the south-west end does not seem now to be recognisable, the rampart being destroyed at a number of places and the ditch much filled in. Near the centre of the north-east end, immediately in rear of the cottages, a slight break is apparent in the rampart ; and some 25 feet in front lies a circular area, measuring in diameter some 33 feet by 27 feet, on which the vegetation is markedly poorer than elsewhere in its vicinity —a condition which may possibly be due to the former existence of a traverse on the spot, the clay from which has deteriorated the soil. The greatest existing height of the rampart above the ditch is about 5 feet. There are no indications of any foundations in the enceinte ; and, though the woodland area has been trenched in all directions for drainage, there is no record or tradition of any traces of buildings having been observed or of relics recovered. liii. N.E. (“ Earthwork ”). 16 July igi2. Enclosure. 46. Enclosure, Macrieholm Knowe. — Cut through by the old road which, traversing the moorland, leads from Old Irvine to Solway- bank, is a circular enclosure measuring some 120 feet in diameter, surrounded by a trench 20 feet wide and 3 feet deep, without any conspicuous mound on scarp or counterscarp, though on the north-west there is a slight swelling on the outside, and along the north half the level of the crest of the counterscarp is higher than that of the scarp and interior. The site is the west end of a hillock dropping some 20 feet on the north face to boggy ground. There are numerous excavated hollows be¬ tween the construction and the edge of the bank on the north, probably made however for the purpose of obtaining soil or gravel for the road. liii. N.W. 17 July igi2. 47. Long Cairns, Small Cairns, and Standing- Stones, Windy Edge.—This group of monu- CANONBiE.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [CANONBIE. ments is situated just within the fence that forms the boundary between Dumfriesshire and Roxburghshire, about 2 miles east-south¬ east of Peterburn farm and the same distance north-east of Bruntshielbog. At 999 feet above sea-level the position commands a wide pros¬ pect extending from Criffel at the mouth of the Nith on the west to Carter Fell in the Cheviots on the east. Long Cairns. —The main constituent of the group is a regular but much disturbed line of heaped stones of fair size, which stretches, over all, for 248 feet from east to west and does not anywhere exceed 5 feet in height. First, from the west, comes a long cairn measuring 115 feet with a breadth varying from 25 to the outside of the structure, and may thus be of secondary origin. Some at least must have projected beyond the outer edge of the cairn. (C/. “Cairn na Gath,’’ Balmurrie Fell, No. 281, in Inventory of Monuments, Wigtown). The circular expansion shows no stones larger than the average of those in the cairn proper, but much of the material has been removed. The circle is 45 to 47 feet in diameter, and the interior is marshy. About 5 feet of clear space intervenes between this structure and that to the east, which seems to consist of two circular heaps, some 30 feet in diameter, connected by a belt of building, on the north of which one chamber is clearly distinguished, while on the south STANDING STONL > a 10 5 0 10 20 30 40 50 ^ 70 80 90 lOO liO (20 130 140 )50 ifaO iTO I 8 O 190 200 FEET I I I I I I I I I I I I I I I I I I I I I t - t-- Fig. 29. —Long Cairns, etc.. Windy Edge (No. 47). 30 feet, which is rounded at the west end and at the east passes into a circular foundation about 4 feet 6 inches wide where sufficiently preserved. A passage inwards is indicated at the western extremity by slabs set on end, between which it measures 2 feet 8 inches. A surveyors’ cairn has been raised above this and blocks further examination. On the north side, however, are the inner halves of two round built chambers, and the same portion of another is obvious, opposite one of these, about half-way along the south side. The latter measures 6 feet 6 inches across the mouth and 5 feet 8 inches from front to back, while the built interior is 4 feet high ; it would appear to have been roofed beehive fashion, as there is no sign of large roofing slabs. These chambers were not entered from the inner passage, but apparently directly from there is a deep bend inwards. This may indicate the position of another chamber enlarged by destruction. An entrance from the east end is suggested by the position of certain upright flat stones. Small Round Cairns. —Rather less than 50 feet south of either extremity is the site of a round cairn, marked by a smooth grassy sward, a slight rise above the level, and some scattered stones. That on the east measures about 20 feet in diameter, and that on the west about 13 feet. Standing-Stones. —Some 90 feet to the east of the long cairns is a large flat stone lying over at an angle of about 30 degrees and 4 feet 2 inches out of the ground on its southern face, which is 3 feet 6 inches wide at most, with a few smaller stones scattered on a slight grassy rise 30 feet in diameter. CANOXBiE.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [closeburn. Another similar lumpish stone lies 130 feet to the south-south-east of the first in a water¬ holding hole which is partly filled in with smaller stones. It lies over to the north at an angle of 45 degrees, and its longer face is 7 feet 9 inches above ground, while it measures 4 feet 2 inches across and averages about half this in thickness. The writer of the Statistical Account, 1795 (vol. xvi. p. 85), speaks of one stone near the south end of the large cairns “ standing perpendicular ... 7 feet above the moss,” and states that he found ” five other stones, nearly of an equal size uith the former, all inclining to, or lying on the ground, forming a circle, the diameter of which is 45 j^ards.” This is just about the distance between the two stones already described. The writer treats the group as being in Roxburghshire. The stones in all cases are of hard sandstone, and, as the moor is entirely boggy with tussock grass and has no scattered stones upon it, these must have been brought from some distance. For the same reason the cairns have been subjected to much spoliation probably on behalf of a dry-stone dyke less than a mile away. xlvi. S.W. and S.E. 4 June 1920. Miscellaneous. 48. Scots Dike.—See p. xix. lix. N.W. and N.E. 5 June 1920. Sites. 49. Cairn, “The Haunches.”—On the S.W. corner of ” The Haunches ” (1002 feet) a rush-grown rise of green sward appears to mark the site of a cairn, the remaining stones of which have been gathered into a surveyors’ cairn, with the exception of a few still left here and there round the base. It is roughly 23 feet in diameter. xlvi. S.W. and N.E. (unnoted). 4 June 1920. 50. Mound and Ditch, Gilnockie Bridge.— At the east end of Gilnockie Bridge is a high mound running from the road in, a northerly direction. Midway there is an entrance, and a fosse runs parallel to the whole front. The inner scarp of the mound has been faced with stones. The enclosure so cut off is a pro¬ montory with precipitous sides to the river and narrowing just in the line of mound and ditch. The prolongation of these to the south bank has been destroyed in making the road at a lower level, but the construction clearly suggests a promontory fort of familiar type. The O.S. map indicates “ Gilnockie Castle (site) ” E. of No. 50. liii. S.E. (” Moat”). 27 March 1915. The O.S. maps also indicate sites as under ;— 51. Priory (Canonbie), Hallgreen.^—Of the Augustinian Priory of Canonbie (” Canons’ hamlet ”), founded in the 12th century and subsequently a cell of Jedburgh, no part now remains. The site is ^ mile south-south-east of the parish church, liii. S.E. 51a. Chapel, near Pingle Bridge, liii. S.W. 52. Morton Church, Tower of Sark. lix. N.W. 53. Tower, Tower of Sark. lix. N.W. 54. Mumbie Tower, liii. N.E. 55. Kinmont Tower, lix. N.E. 56. Harelaw Tower, Harelawgate. liv. N.W. 57. Tower, Outer Woodhead. liv. N.W. CLOSEBURN. Ecclesiastical Structure. 58. Old Church.—The old church, according to the Statistical Account, was rebuilt in 1740 with a north transept: of this building little more than the east gable remains, its place having be*en taken by a new church built on the south side of the old churchyard in the 19th century. The remaining east gable of the old church is 30 feet wide and the wall is 3 feet thick. It contains a doorway 4 feet wide, and has a semicircular arched head, with moulded archivolt, keystone, and imposts. In the upper part of the wall is a circular window, as at Morton and Dalton, and on the top is a belfry. The remains of the north and south walls are 2 feet 6 inches thick. Bell.- —The bell still remains in the belfry, and is inscribed :— -l-TREGINTA-|-DE-l-AGVST-f APVD • POTERRAW + ANNO+DOMINI +1606 On waist a crown, with hammer below, and letter G on one side, H on the other. Diameter 15J inches. 30 CLOSEBURN.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [closeburn. The inscription is in two lines, as shown ; there are single rims above, below, and between these lines. The lower line has smaller lettering. The lettering is very rude and irregular, the n’s and the s are reversed, and all the letters, figures, and ornaments appear to have been made by marking the mould with a sharp instrument instead of by the use of stamps. The initials are those of George Hog, who cast several bells during the earlier part of the 17th century, including one at Keith Marischal, Haddingtonshire. The hammer and crown are the insignia of the Incorporation of Hammermen of Edinburgh, and “ APUD POTERRAW ” in the inscription evidently refers to the street of that name. Brass Alms Dish.^ —In the manse is pre¬ served a brass alms dish 13 inches in diameter, bearing in the centre, in repoussee work, a representation of the Annunciation. It is German work, of probably the 15th century. Font from Dalgarnock.— In the porch, beneath the tower of the modern church, is the basin of the font of the old church of Dalgarnock. It is a plain octagonal basin, with a drain in the bottom, 2 feet 3 inches in diameter over all, i foot g inches in diameter across the actual basin, i foot 3 inches in depth outside and 8 inches inside. The name “ Dalgarno ” has been cut on the edge in modern lettering. Cross-shaft, etc. —Beside the font lies a fragment of a cross-shaft, with two vertical panels of interlaced work formed from a four-cord plait. A fine beak-head, evidently from the cornice of a church of the later Norman period, is also preserved here. xxxi. S.E. 24 May and 12 June 1912. Castellated and Domestic Structures. 59. Closeburn Castle.—This tower (fig. 4 of Introduction), dating from the end of the 14th century, stands in what is now one of the fertile and wooded parks of Nithsdale, some 12 miles by road north-north-west of Dumfries. The site has been originally a peninsula at the south-east end of what once was Closeburn loch, and the approach to the castle from the east or landward side has been defended by a wide moat cut across the neck of the peninsula, which would in this way be converted into an island, as it is shown in Grose’s Antiquities of Scotland even as late as 1789. On plan (fig. 30) the tower is of the simple rectangular type, measuring, on the ground floor, some 27 feet 6 inches by 15 feet 6 inches within walls nearly 10 feet in thickness. The total height from the ground to the level of the parapet measures some 50 feet. The building com¬ prises four storeys and an attic, the two lower GROUND FLOOR SECTION FIRST FLOOR Fig. 30.—Closeburn Castle (No. 59). storeys, including the basement, and the attic being vaulted. A doorway in the west wall gives access to the basement, which is now subdivided into three dark cellars. No windows appear to have been formed in its massive walls, probably with a view to greater security, nor has there been any internal com¬ munication with the upper floors. The main entrance is situated at the north end of the west wall at the first-floor level, some 10 feet from the ground ; access to the tower would be gained originally by a move- able ladder, which has been replaced in later times by the present forestair of stone. The doorway measures 4 feet 6 inches in width, has splayed outer jambs and a semicircular arch-head, and is still secured by an excep¬ tionally well-preserved iron “ yett.” There is also a bar-hole in the south jamb, some 6 feet in depth. Originally the first floor would serve as the hall, measuring about 30 by 18 feet within walls averaging 8 feet in thickness ; but it has been subsequently divided by a central partition some 3 feet in thickness, which contains two fireplaces. The windows have been enlarged to suit modern requirements. A wheel-stair in the north wall adjoining the entrance gives access to the upper floors and to the parapet walk. The CLOSEBL-RX.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [closeburn. upper floors have been altered and adapted for convenience of occupation, and an attic has been formed within the uppermost vault. The crenellated parapets of the main building and of the cap-house are e\ddently of recent date. The castle is still inhabited, and is in excellent repair. A charter by Alexander IT, of the period 1231-1232, coirfirms a grant of the lands of Closeburn to Ivo de Kilpatrick. It is copied into the GlenriddeU MSS.^ In 1296 William de Kilpatryk of the valley of Annan is a prisoner at Windsor ; - and in 1299 Ivo, son of Stephen de “ Killeosborne,” died a hostage at Carlisle.^ ^ Cf. C. Kirkpatrick Sharpe’s Correspondence, vol. i. p. 552 ; Grose’s Antiquities of Scotland, i. p. 150 ; “ Calendar of Documents relating to Scotland, iv. p. 358 ; ^ ibid., ii. No. 1179. xxxi. S.E. 6 June 1912. 60. Low Auldgirth.—Adjacent to the farm of Low Auldgirth are the ruins of a small keep, consisting of portions of two contiguous walls some 3 feet 6 inches in thickness. The base¬ ment has been vaulted with a simple barrel- vault. xl. N.E. (unnoted), i May 1913. Defensive Constructions. 61. Fort, Townfoot.—On the western slope of the moorland, about | mile to the south¬ east of Townfoot farm, is a fort. It is elliptical in form, lying north-north-west and south-south-east, and shows a broad stony rampart rising to a height of some 5 feet above an encircling trench, now scarcely apparent except at the ends. The rampart has been considerably despoiled for stones along the west side, and there are several gaps in it. The principal entrance, however, has evidently been at the north end of the west side, opening on a slight hollow in the interior. A stony bank faces the gap and passes southward, flanking it for some 40 feet, thereafter passing across the interior towards the rampart on the east side. On the south side of the entrance a low stony mound runs outward as a traverse for a distance of some 30 feet across the front of it. The length of the interior is some 230 feet and the breadth at the centre 175 feet ; the rampart at base has a breadth of 20 feet where prominent at the south-east angle, and the trench a width of 25 feet from crest to crest. The width of the entrance at ground level is about 7 feet. The interior is very uneven and stony, showing in one or two places evident remains of divi¬ sional banks or walls, and rushes growing in several spots suggest the presence of water. The elevation of the site above sea-level is 800 feet. A section of the " Deil’s Dyke ” runs parallel with the west face of the fort, some 60 feet distant (see No. 80). xxii. S.E. 7 June 1912. 62. Fort, Trigony Wood.—At* the end of a ridge which rises steeply from the northward, overlooking Trigony House, are the remains of an oval earthwork. It is situated within a thick fir wood, so that its outline is now difficult to follow, while measurements are almost unobtainable. According to the O.S. map, the dimensions are approximately 240 by 200 feet. The defence consists of a single trench, where best preserved some 13 feet in breadth and now nowhere of greater depth than 2 feet, with a slight mound above the scarp, and probably the same above the counterscarp. Along the east side the trench is barely traceable, and on the north-west* there appears to be a gap of some 60 feet, where it has disappeared entirely. xxxi. N.E. 12 June 1912. 63. Fort, Crichope Linn.—In the field on the north and near the head of Crichope Linn, are the remains of a fort. With its base resting on the precipitous bank of the Linn about 80 feet in height, it extends to the north-east with two straight sides some 136 feet apart, formed of a trench and inner rampart, now imperfect on the north-west side, which are connected by a segmental curve towards the north-east. The plan appears to be an irre¬ gular ellipse, with its major axis from north¬ east to south-west. Along this axis it measures to the stone dyke which cuts across it at the side of the Linn 212 feet, and where widest it extends some 40 feet farther to the edge of the bank. The surrounding trench has had a width of about 28 feet from crest to crest, and, on either side of the entrance, where it is 32 CLOSEBURN.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [closeburn. now best preserved, it has a depth of lo feet below the top of the scarp and 7 feet below that of the counter-scarp. The rampart at greatest height rises to about 3 feet above the interior level and is very stony. On the west there is a considerable gap in the defences at a point where the ground on the interior rises sharply to a height of 7 feet, and the trench, less deep at this point, may have been filled in and obliterated by cultivation. The en¬ trance, 12 feet wide, has been from the east, passing across the trench on unexcavated ground and through the rampart. Immedi¬ ately on the right of it, in the interior, on slightly elevated ground, is a circular area enclosed by a bank and measuring 41 feet in diameter. It is entered from the west dia¬ metrically opposite to the entrance to the fort. xxxi. N.E. (unnoted). 20 June 1912. 64. Fort, Benthead.—About J mile north- north-east of Benthead, on the south-east bank of the Linn and some 60 yards down from the fine waterfall known as the “ Grey Mare’s Tail,” is a small circular entrenchment. The ground falls from the southward towards the edge of the Linn, and directly overlooks the con¬ struction, which forms a small plateau above the precipitous bank 60 to 70 feet in height. The mound is encircled by a horse¬ shoe trench, some 20 feet wide, the ends of which rest on the bank, and it has nowhere a height of more than 5 feet above it, while on the upper side it does not exceed 3 feet 6 inches. The summit area measures 40 feet in diameter, and is not very level. On the south¬ west face is a depression some 12 feet across which gradually falls to a depth of 4 feet be¬ low the summit level. This appears to be secondary, and the soil from it has seemingly been used to level a platform at the edge of the bank of the Linn and partly to close the end of the trench in that direction. This construction is very similar to that on the Wanlock Water up the Crawick Pass, in Sanquhar parish (>^^0. 553). xxii. S.E. (unnoted). 20 June igi2. 65. Mote, Dinning.—This mote (fig. 31) lies some 200 yards north of the farm of Dinning, and with its base-coiirt or bailey is fashioned on a long natural hillock lying north-west and south-east, which rises out of a stretch of haugh- land reaching to the Nith, about J mile to the westward. The east slope of the valley com¬ mences to rise sharply some 50 yards distant from the base of the hillock. From the south¬ east end the knoll rises gradually to its north¬ west extremity, on which the mote itself has been erected. With the base-court lying at the south-east end, the whole construction occu¬ pies about half of the length of the hillock. In form the mote is a simple truncated cone. SECTION ON LINE A-B PLAN FEET Fig. 31.—Mote, Dinning (No. 65). composed, as far as it is possible to tell, of earth, and rising to a height of some 14 feet above the level of the base-court, while a steep gradient with a vertical height of 45 feet reaches to the base of the hillock on front and sides. On the terminal slope, some 20 feet up from the base, is a slight terrace which is possibly artificial. The plat of the mote has been circular with a diameter of some 20 feet, and shows a shallow bowl-shaped hollow, the wall of which has been slightly broken down on the north-west. The base-court is oblong on plan, measuring 66 feet in length by 57 feet in breadth, and is enclosed by an earthen rampart, somewhat slight on the sides but massive to the front, which impinges directly on the sides of the mote hill, uninterrupted by any intervening trench. The entrance to the court has been through the centre of the south¬ east front. A trench 34 feet in width, 12 feet in depth below the crest of the scarp, and some 4 feet below that of the counterscarp, has been dug across the hillock from side to side in front of the rampart. There is no gangway across 33 3 CLOSEBURX.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [closeburn. the trench, which has probably been covered by a bridge or drawbridge. xxxi. S.E. (“ Earthwork ”). 7 Maj^ 1913. Hut-Circles. 66. Hut-Circle, Townhead.—About J mile due east of Townhead farm buildings, and some 40 3'ards down from the road to Fellend, is a well-defined hut-circle. It is a small circular enclosure, surrounded by a turf bank measur¬ ing interiorly 7 feet 6 inches by 6 feet 6 inches, with its longer axis towards the entrance, which, vith a width of about 2 feet 6 inches, faces south bj^ west. The bank has a thick¬ ness of about 3 feet 6 inches and, at most, an elevation of i foot. The structure has been placed vithin a larger circle, the stony founda¬ tion of which is just traceable, measuring in diameter some 26 feet, close to the back but against the waU—presuming the front of this outer circle to have been in the same direction as that of the interior construction. Though no small cairns are actually adjacent, there are a number sparsely scattered to the west and south. xxii. S.E. (unnoted). 20 June 1912. 67. Hut-Circle, Townhead.—On a natural terrace on the hillside to the east-north¬ east of Townhead farm, and some 400 yards distant, is a small hut-circle excavated by a former tenant of the farm. It measures in¬ teriorly 8 feet by 6 feet, the longer axis being towards the entrance, which has been from the south-south-east. The floor was sunk about I foot below the adjacent natural level, and a low stony bank surrounded the edge. A small hearth was found in the centre of the hut, formed of thin stones set obliquely in the soil, as , well as wood ashes, but no relics were recovered. The hut-circle is within a large oval walled enclosure in an extreme state of ruin, which, however, may be of later date. In one of the fields, at a lower elevation to the south of the farm, it is said that numerous hut sites— recognisable by the charred deposits—were disturbed in a deep ploughing some years ago. Numbers of flint flakes have been found in' the vicinity. xxii. S.E. (unnoted). 20 June 1912. Cairns. 68. Small Cairns, Nether Dod.—At the south-east end of Nether Dod, on the lower end of the haunch of the hill overlooking the Capel Water, is a small group of cairns measuring some 12 feet in diameter and very low in elevation. They are at an altitude of some 850 feet above sea-level. xxxii. N.E. 7 June 1912. 69. Small Cairn, Fellend.—On the upper side of the road from Townhead to Mitchellslacks, about 350 yards to the east by south of Fellend, is a small cairn with a diameter of 24 feet and an elevation of 2 feet. It has not been excavated. It lies at an altitude of nearly 1000 feet over sea-level. xxiii. S.W. 7 June 1912. 70. Small Cairns, Knockbrack.—On the west side and around the summit of Knockbrack, a green grassy hillock, which rises up from the moorland about | mile to the south-east of Townfoot Loch, is a large number of small cairns, which measure from 10 to 20 feet in diameter, are overgrown with grass and low in height. They are situated at an elevation above sea-level of about 900 feet. One near the summit, measuring some 20 feet in diameter, was opened some ten years ago by the tenant in Townhead, who found two flint chips, a round thin disc of stone about the size of a penny with a small depression in the centre on one side, and some charred wood. xxiii. S.W. 7 June 1912. 71. Cairn, Capel Glen.—About f mile to the north-east of the farm of Locherben, on a plateau which interposes between the slope of the high land to the westward and the precipitous right bank of the Capel Burn which flows by 100 feet below, lies a cairn, measuring some 25 feet in diameter and low in elevation, formed of large stones, considerably overgrown. It does not appear to have been excavated. A number of large slabs, lying flat on the surface or set upright and just protruding between the cairn and the top of the bank of 34 CLOSEBURN.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [closeburn. the burn, are suggestive of a ruined cist. The altitude is about goo feet. xxiii. S.W. 7 June 1912. 72. Cairn, Threip Moor,—This cairn is situ¬ ated on the crest of the watershed between the Poldivan Burn and the Capel Water, about J mile west of the confluence of the streams and nearly J mile east-north-east of where the road running southward from Mitchellslacks crosses the former. It has not been excavated, and measures some 50 feet in diameter and 6 feet in elevation. xxxii. N.W. 7 June 1912. 73. Cairns, Nether Dod.—On the southern end of the long grassy hill which lies to the east of Mitchellslacks and bears the name of Nether Dod, at an elevation of some 950 feet over sea-level, are two cairns within 150 yards of each other. The one measures 39 feet in diameter and 4 feet in elevation, while the other measures 30 feet in diameter and 4J feet in elevation. xxxii. N.E. 8 June 1912. 74,. Cairn (remains of), Auchencairn.—In a grass park about | mile to the north of Auchen¬ cairn farm house are the remains of a very large circular cairn. Only a small segment remains, and the site of the remainder has been covered with gathered stones from the field. The interment has probably long since been disturbed. The field is known as the “ Witches Wa’s.” xxxii. S.W. (unnoted). 12 June 1912. 75. Cairns, Gawin Moor.—Situated on Gawin ]\Ioor, rather more than a mile to the north¬ east of Auchencairn, is a group of eight or nine small cairns. The largest, which has measured some 18 feet in diameter, has been excavated. Some 300 yards to the southward, on a slight ridge overlooking a stretch of boggy moorland to the west, is a much larger cairn, measuring in diameter 62 feet and in elevation 8 feet. At one or two places excavations have been made in it, but no cist or chamber has been reached. To the north and north-west of it lie several small cairns measuring in diameter from 12 feet to 14 feet. Some 200 yards to the southward, and near the edge of the boggy land, lie the remains of another circular cairn which has been ex¬ cavated, and the remains of a short cist lie exposed in the bottom of it. The cairn has measured some 44 feet in diameter. The ex¬ cavation, which was conducted in 1894, re¬ vealed three cists : one, measuring 3 feet 6 inches in length by i foot 9 inches in width, contained burnt earth and ashes; the second, of which no dimensions are recorded, contained similar remains ; the third, which is said to have measured only 2 feet by 11 feet, yielded fragments of a beaker urn and a flint implement. No osseous remains were found [Dumfries Standard, 25th July 1894). Sixty yards or thereby to the north are the remains of another and much smaller cairn, also excavated. The “ Mid Cairn ” is a large circular cairn on the crest of the moorland, i mile due east of Auchencairn and 80 yards or thereby to the east of the Drove Road. It measures some 54 feet in diameter and about 9 feet in eleva¬ tion. Though it has been dug into to a small extent it still remains unexcavated. On the moorland, to the east of the Drove Road, and about | mile north-east from Auchencairn, is a cairn which has been excavated. It has measured about 20 feet in diameter. xxxii. S.W. 12 June 1912. 76. Cairn, Gufhill Rig, Knockenshang.—Some 200 yards south-south-east of the summit of the hill which overlooks the road from Annan- dale into Nithsdale by Loch Ettrick, J mile west of the farm of Knockenshang, and just under an altitude of 900 feet over sea- level, there is a large oval cairn overgrown with grass except at the north end, where the stones are exposed. It measures 93 feet from north to south by 84 feet from east to west, and rises to a height of 5 feet. Though a slight excavation has been made in it at the north end, the interment does not appear to have been disturbed. xxxii. N.W. 2 May 1913. 35 CLOSEBURN.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [closeburn. Miscellaneous. 77. Gravestones, Closeburn.—At the east side of the churchj'ard is a slab heavily carved along one side, with a pillar-like figure enriched with festoons and terminating in a human head. Incised is an inscription : here LYETH THE CORP OF JOHN KIRKPATRIC IN BARNMILL WHO DIED JAN. 1696. AGED 76. At the foot of the slab, in relief, within a car¬ touche, is a shield bearing a saltire and chief, the latter charged with three cushions, all surmounted by a helmet and mantling. These are Kirkpatrick arms. xxxi. S.E. 12 June 1912. 78. Gravestones, etc., Dalgarnock Church¬ yard.—Within the churchyard at Dalgarnock are one or two late 17th-century gravestones of no particular interest, and two dating from the first half of the i8th century, which show figures in contemporary costume carved in relief—one commemorating a schoolmaster from Glencairn and the other a “ Chirurgeon ” from Thornhill. On the left of the entrance stands the socket-stone of a cross, measuring 2 feet 2j inches by i foot 9 inches by i foot 4 inches, with a rectangular sinking in the centre. xxxi. N.E. 12 June 1912. 79. Mound, Knockhill.—At the west end of the wall which comes down by the south side of the Knockhill, forming the boundary between Townhead and Townfoot, is a grass-covered mound, evidently artificial, lying with its longest axis east-south-east and west-north¬ west, and measuring in diameter 24 feet by 15 feet. Without excavation its character cannot be determined. xxii. S.E. 20 June 1912. 80. “ Deil’s Dyke.”—Running parallel with the west face of the fort (No. 61), some 60 feet distant, but at an elevation about 20 feet lower, is a section of the “ DeiTs Dyke.” It is here an earthen mound some 12 feet wide at base, with a certain amount of stone protrud¬ ing at places through the top, rising to a height of from 2 to 3 feet, and with a slight and narrow trench some 7 feet wide on the upper side. It runs in an irregular line along the face of a steepish slope some 20 feet down from the crest. The Dyke passes along the lower slope of the hillside, just above the enclosed land, 200 yards to the north of Townhead farm. In appearance it is an earthen bank, with a trench on the upper side, running irregularly across the brae face, measuring some 8 feet broad at base, narrowing upwards, and some 2 feet in height, while the trench has a breadth of about 7 feet and is now shallow. Where the bank has been broken by sheep it is shown to be formed with a core of boulders laid horizontally. The stony structure of the rampart becomes much more pronounced as it turns down the slope above Burn farm. Here, indeed, it has the appearance of being wholly formed of slabs, generally about 2 to 3 feet long by 18 inches wide, while in the usual earthy structure the stones are mostly such as may be carried in the hand. xxii. S.E. 7 and 20 June 1912. A section of the Dyke is also to be seen crossing the field between Benthead and Crichope Linn. It is an earthen bank, 3 feet 6 inches in height and 12 feet wide at base, with a trench on the east some 14 feet wide, from which the soil has been upcast. xxxi. N.E. (unnoted). 20 June 1912. 81. Standing-Stone, Kirkbog.—On the crest of a broad-backed ridge, J mile east of the farm of Kirkbog, stands a single upright whinstone boulder in the middle of a cultivated field. It measures 4 feet 3 inches in height, and about half of its thickness has been broken off at no distant date. There is nothing in the character of the stone, nor in its situation, to contradict the statement in the History of Closeburn, that originally there was a stone circle here. xxxi. N.E. (unnoted). 7 May 1913. Sites. 82. “Cairn,” Benthead. — About J mile north-north-east of the Grey Mare's Tail waterfall, near Benthead, on a slight eminence towards the crest of a ridge, is the site of a dilapidated construction, probably a large CLOSEBURN.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [cummertrees. circular cairn, from which the larger stones have been removed. xxiii. S.W. (unnoted). 20 June 1912. 83. Cairn, Auchencairn Height.—The cairn on the top of Auchencairn Hill is a mere site ; the interment has probably been removed long ago. xxxii. S.W. 12 June 1912. 84. Tumulus, M‘Mount, Knockenshang.— The “ tumulus ” noted on the O.S. map on the summit of the hill across the valley to the west-south-west from the Gufhill Rig is now a low, stony mound, with a diameter of some 8 feet, and with a few large stones lying on the surface. xxxii. N.W. 2 May 1913. Sites are also indicated on the O.S. maps as under ;— 85. Dalgarnock Church and St Ninian’s Well, about 700 yards north-west of Kirklands cottage, xxxi. N.E. 86. Chapel, Nether Mains.—xxi. S.E. 87. St Patrick’s Chapel, Kirkpatrick.—xxxi. S.E. 88. Royach Cairn.—xxxii. S.W. CUMMERTREES. Castellated and Domestic Structures. 89. Repentance Tower.—This tower (fig. 4 of Introduction) stands on rock within an old graveyard on the summit of a small hill (350 feet), which is about half a mile to the south of Hoddom Castle, and commands an extensive view on all sides. It dates from about the middle of the i6th century, but has evidently been repaired in later times. The walls are built of a local pinkish sandstone set in courses with dressed margins and jambs. On plan (fig. 32) the tower is oblong, measuring externally 24 by 21 feet. It is three-storeyed and terminates at a height of 30 feet above ground level in a parapet walk, within which rises a roof covered with over¬ lapping flagstones and surmounted at the apex by a central chimney. The entrance is at the east end of the north wall, at the level of the first floor, 3 feet above the ground. On the door lintel is the word repentence, executed in raised Gothic lettering and flanked on the dexter by a carving of a bird and on the sinister by a scroll. The ground floor has evidently been entered through the first floor but is at present in¬ accessible ; from outside can be seen two built-up gunloops on each wall. The windows are small, with rounded jambs, save where these have been reconstructed with rect¬ angular rybats. At first-floor level a large window checked for outer shutters has been inserted in the south wall, probably in the i8th century, and in its turn is now built up. The lower member of the continuous corbel course is original, the upper member and the projecting gargoyles are later, if not modern. The parapet has been reconstructed in the early i8th century and has ashlar quoins with channelled joints. The chimney is similarly jointed, as is the doorway opening from a staircase to the parapet walk, and the south-west angle of the tower has been repaired at ground level with channelled quoins. The entrance is reached Rom the exterior by five modern steps and has been closed by two doors. The interior of the first floor is lighted by several narrow windows or loopholes and by the later window in the south wall. Vithin the south-east angle is a recess, its stone sill raised about 2 feet above the floor level. The 37 CUMMERTREES.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [cummertrees. upper floor has been reached by a ladder. This room, measuring 14 feet by 10 feet 8 inches, is now fitted up as a dovecot. It is lighted by a narrow loophole in each wall, and a fireplace is formed in the north-west angle; the upper part is enclosed with a semicircular stone vault below the flagged roof. From the level of the second floor a stone stair, starting on a corbelled projec¬ tion intruding on the room below and resting partly on the east wall, leads to the low doorway which gives access to the parapet walk. The origin of the superscription repentence has been connected with a certain chain of episodes in the career of John, Master of Max¬ well, afterwards Lord Herries, about the middle of the i6th century.^ The same person seems to have built Hoddom Castle. One of his proposals in 1579 to the king for the proper rule of the Borders definitely fixes the character of the building : “ The wache toure upoun Trailtrow, callit Repentance, mon be mendit of the litill diffaceing the Englische army maid of it (? 1570) ; and, according to the formar devise, the greit beU and the fyir pan put on it; and ane trew man haiff ane husband land adjacent for the keping of the continuall wache thairupoun.”^ At this time tower and lands were the property of Lord iMaxwell, who had purchased the feu from Lord Herries.® The “ devise ” in question was no doubt that engrossed in the Border Laws, which enjoined “ ever in Weir and in Peace, the Watch to be keeped on the House-head; and in the Weir the Beaken in the Fire-pan to be keeped, and never faill burning, so long as the Englishmen remain in Scotland ; and with ane Bell to be on the Head of the Fire¬ pan, which shall ring whenever the Fray is, or that the Watchman seeing the Thieves dis¬ obedient come over the Water of Annand, or thereabout, and knowes them to be Enemies ; And whosoever bydes fra the Fray, or turns again so long as the Beaken bums, or the BeU rings, shall be holden as Partakers to the Enemies, and used as Traitors to the Head-Burgh of the Shyre, upon ane Court-day.” ^ ^ Transactions, Glasgow Archceol. Society, 1896 ; Repentance Tower and its Tradition, by George Neilson, p. 340 ff. ; ® Register of Privy 38 Council, hi. p. 81 ; ® ihid., p. 84 ; * Leges Mar- chiarum, p. 198. Ivii. S.W. 14 August 1914. 90. Hoddom or Hoddam Castle.—This castle (fig- 33) stands on the eastern edge of a plateau on the right bank of the River Annan, 6 miles south of Lockerbie. The buildings are en¬ closed on west and south by a courtyard wall, much modernised but showing traces of 17th- century detail. Outside the south waU is a deep artificial ditch. The castle is buUt of reddish ashlar and towers above the modern additions which encompass it on all sides save the north and east. It is buUt on an L plan (fig. 34), the shorter wing lying to the west. Externally it measures 51 feet from north to south and the same from east to west ; the longer wing is 36 feet broad, the shorter 29 feet. The former contains four storeys beneath the waU head, where it terminates 52 feet above the ground in a parapet walk, within which rises a slated roof containing a garret ; the other wing is carried up to a height of 72 feet, where it terminates in a flat roof within a parapet ; it contains the main staircase. The parapet of the larger wing is borne on a continuous moulded corbel course of a late decorative type, with moulded interspaces and corbels, which breaks upward and returns around the staircase wing at a higher level; rounds are set at the angles of the lower parapet, and turrets of slight projection rise at the western angles of the higher corbel course to above the flat roof covering the staircase wing, where they terminate in conical roofs. Around this flat roof runs a parapet, largely modern but pro¬ vided with embrasures and with rounds at the eastern angles, as on the main tower. The entrance to the castle is in the short arm of the re-entering angle by a doorway with an arched head of segmental form ; on the jambs and head is wrought a bold quirked edge-roll moulding with a fillet ; the label takes the form of a monstrous cable with knotted stops, now mutilated. The lower storey is unlit save by two gun-loops in the east wall; the other floors are lit by windows of moderate size, some of which have been heightened ; the edge-roll moulding is repeated on the jambs and lintels. Ancienl and Historical Monuments — Dumfries. A 3S. Fig. 33.—Iloddom Castle (Xo. 90). CUMMERTREES.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [cummertrees. Just within the entrance a doorway on the the east gable. The mid-partition between right admits to a small lobby, formed in the these chambers is carried up through all the thickness of the intermediate wall, which gives storeys to the roof, forming two apartments on GROUND FLOOR SECOND FLOOR Fig. 34.—Hoddom Castle (No. go). access to the northern of two intercommunica¬ ting chambers contained within the main wing. These are ceiled with a semicircular barrel- vault, and a mural closet is formed within each floor. The main staircase is spacious and of easy ascent ; from it is entered at first-floor level a lobby formed within the intermediate wall, which contains on the south a mural cuMMERTREES.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [cummertrees. chamber and gives access on the east to a modern passage encroaching on the apartments which originally occupied the full width of the main wing. These and the corresponding apartments on the floors above have been modernised and present no special features of interest. In several rooms mural chambers are contrived within the massive walls. The main staircase terminates at the second floor, and from this level a turret staircase, contained within the walls at the re-entering angle, con¬ tinues the ascent, and gives access to the upper floor of both wings and to the roofs. The room over the staircase has a stone-vaulted ceiling at the level of the corbelling, while the two storeys above have been subdivided by modern partitions and the walls covered with lining. The small circular chambers contrived within the turrets are retained. The north room on the fourth floor contains a fine 16th- century fireplace in the w^est wall, which, having been plastered over, was only dis¬ covered some few’’ years ago. A bold quirked edge-roll is w^orked on the jambs and is carried in tw'o impinging semicircles across the lintel. A modern shelf and fire-grate have been added, but without interfering with the old w'ork. In the north-east angle of this room is a garderobe with an outlet through a hollow corbel ; from the garderobe the parapet walls of the main wing could be reached, but the access is now filled up. From the parapet w^alk machico- lated projections are corbelled out at the centre of the east wall and at the junction of the wings. That over the re-entering angle is evidently for defence of the doorway; the others w'ere possibly for disposal of sew^age. In addition to these outlets gargoyles project from the corbel-table to carry off roof water. The gables of the garret are crow-stepped; the southern chimne}^ stalk has been added to. The north gable is surmounted by a w^atch or beacon platform borne on a corbel-table. The great size of Hoddom Castle, and the thick w'alls (8 to 9 feet on the lower floors) with w'hich it is built, together with the presence of mural chambers, would suggest its origin in the 15th century, but an analysis of plan and decorative and structural detail show's that the structure was erected towards the close of the i6th century, in all probability by John Maxw'ell, Lord Herries. (C/. No. 89.) The position w'as know'n as “ Hoddomstanes,” and the fortress on this the manor place of Hoddom is the subject of various record references in the early part of the 17th century ; as, in 1610, lie Hodaniestanes cum fortalicio et manerie, etc.^ Hoddom w’as accounted a very strong place ; it surrendered to the Regent Murray on 21st June 1568, after a day’s siege, but it is remarked that the garrison “ mycht haif holdin long enewcht, yf thar had bene gud fellowes within it.”- It then became the military base for the Regent’s Warden, Drum- lanrig .3 By March 1569 it had been recap¬ tured by the adherents of Queen Mary.^ In 1570 Lord Scroop ‘‘ blew up with pow'der the Castle of Hoddom.” ® 1 Reg. Mag. Sig. (1609-20), No. 295 ; ^ Cal. Scottish Papers, ii. No. 717 ; ® ibid., 723 ; ^ ibid, 1017 ; ® Herries’ Memoirs, p. 127 ; cf. Introd., p. xl. Ivii. N.W. 25-26 March 1915. Defensive Construction. 91. “ Moss Castle,” Murraythwaite.—Within a thick fir plantation, about J mile south by west of Murraythwaite House, is a circular enclosure w'ith an internal diameter of some 180 feet (O.S. measurement), surrounded by a bold rampart of earth and stone some 5 feet in height above the general level of the interior, and on the exterior attaining to its greatest height towards the w'est, w'here it is 9 feet high, with a trench in front some 32 feet in width from crest to crest, but now very shallow, and a mound above the counterscarp. The entrance has been from the east, approxi¬ mately 7 feet in w'idth at the lowest point of the interior, and about 3 feet below that of the highest part. Ivii. S.W. (” Fort ”). 2 October 1912. Miscellaneous. 92. Graveslabs, Hoddom Castle.—The under¬ noted graveslabs are lying on the terrace at the N.E. angle of Hoddom Castle:— (i) A slab 6J inches thick, 5 feet 8 inches long, I foot 5 inches broad, with a raised margin 2| inches broad around the edges, framing a raised cross-patty set on a shaft qj inches broad, bifurcating at base and terminat¬ ing in opposed spirals. The ground of the slab is recessed | of an inch, leaving the margin and cross in relief. 40 cuMMERTREES.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [dalton. (2) A slab 6 feet 2 inches long, i foot 8 inches broad at top, tapering to i foot 5 inches at base with a chamfer wrought on the edges, and bearing a cross in relief. The cross¬ head is formed of four penannular rings within which is a rectangle. The lozenge¬ shaped space formed by the impinging of the circles is occupied by a small cross. The edges of the upper portion of the shaft are shaped at neck, and thence descend straight to a graduated base of three steps. On the sinister side of the cross is a sword with straight quillons, and on the dexter a panel. (3) A slab 4 feet ii inches long, i foot 7 inches broad, with a chamfer on the edges and a raised margin framing a cross-patty set on a shaft terminating at base in a mount. (4) A fourth slab is so overgrown as to be indecipherable. Ivii. N.W. 14 August 1912. 93. Carved Lintel, Murraythwaite House.— Built into Murraythwaite House above the south garden door is a stone on which are carved the initials W. M. and I. G. for William Murray and Jean Grierson, and the date 1660. Above, on a pediment-shaped stone, are two shields bearing arms:— dexter, three stars (Murray) ; sinister, three quadrangular locks (Grierson). Ivii. S.W. 2 October 1912. Sites. The O.S. maps indicate sites as under ;— 94. Church, Cummertrees. Ixii. N.W. 95. Trailtrow Chapel near Repentance Tower (No. 89). Ivii. S.W. DALTON. Ecclesiastical Structures. 96. Church, Mickle Dalton.—The present church of Miclde Dalton was built to supersede an older building, which stands on rising ground to the south of it and is now roofless. This older church was erected in 1704, and is an interesting example of a Renaissance building of superior type to most Scottish churches of the time. It may be compared with the still more elaborate church of Durisdeer (No. 152), which was built five years previously. It is oblong on plan (fig. 35), 53 feet 3 inches by 20 feet 4 inches, internally, and is built of local sandstone, with carefully finished rustic¬ ated dressings to the corners and openings and a simple but good Renaissance cornice. There are at ground level two doorways in the north wall, one in the east wall, and one in the west wall. In the middle of the south, west and east walls are doorways at a higher level which led to galleries. The plat at the top of the west gallery staircase is supported on one side by a circular pillar, thus forming a porch to the west doorway below. In the upper part of the east wall is a circular wdndow, and there are large square-headed windows in the ^ ^^ ^ ^ 0 _10_M_^TTET Fig. 35.—Dalton Church (No. 96). north and south waUs, The west gable is crowned by a small belfry, surmounted by the remains of a vane. The east gable formerly had a large finial, consisting of a pedestal i foot II inches square at the foot supporting a spherical ornament, on which is a kind of circular capital that forms the support for a slightly tapering upright stone, 2 feet 4 inches high, carved with w’avy lines as if to represent flames. This finial is now erected upon a new stone base in the churchyard attached to the present parish church. Its total height is 5 feet 7 inches. Another such finial is in the garden of the Thornhill Museum. That the north side of the church and the east and w'est ends are built on the base of the walls of a previous building is shown b}- a massive plinth or base-course, with a plain chamfer at the top, which extends all along the north side of the church, except where it has been cut through to admit the doorways. On the low’er part of the west wall several courses of masonry of much earlier date than 41 DALTON.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [DALTON. the rest of the chiuxh are visible above it. These remnants of an older chnrch may perhaps date from the 13th centm-y. Bell. —The bell of the present parish church is inscribed : ; GEORGE CARRUTHERS • OF • HOLDMAINS • PATRON • OF • THE • UNITED • PARISHES • OF • MEIKLE • AND • TITLE • DALTOUNS • EDINBURGH • I704 • On waist, repeated on opposite sides : Shield with Carruthers arms, surmounted by helm and crest with mantling, and above the motto • PROMPTUS • ET • FIDELIS • Diameter 2iJ inches, height 14! inches. A clean and good casting by John Meikle of Edinburgh. Inscription bands close to¬ gether and separated by very small rims. Strawberry-leaf frieze above, often used by Meikle, and copied from the work of Peter Ostens of Rotterdam. The coats of arms on the waist are very carefully done, and the heraldic tinctures are indicated. hdi. N.W. 9 June 1912. 97. Little Dalton Church. —if miles west- north-west of Dalton village and 200 yards 10 5 0 10 20 30 FEET |iii i| ii i i | -1-1-1 Fig. 36.—Little Dalton Church (No. 97). south of the farm of Kirkhill are the ruins of Little Dalton Church. The church is oblong on plan (fig. 36), measuring exteriorly 25 feet 6 inches from north to south and 52 feet 6 inches from east to west ; from the north wall a sacristy projects 16 feet 9 inches and is 17 feet wide. The walls are 2 to 3 feet thick and 8 feet high. The windows are all filled in and few traces of detail are visible ; the westmost window in the south wall has a daylight 6 inches wide and i foot 9 inches high, with a groove for glass. The head is pointed and formed in a single stone. The church is entered from a semicircular-headed doorway, 4 feet 7 inches wide, in the north wall. A semi¬ circular archway in this wall affords access to the sacristy—a small apart¬ ment measuring internally 7 feet 7 inches from north to south and II feet 7 Fig. 37.—Window, inches from east to west, with an outer door in the east wall. There is now no trace of a piscina or aumbry in the building, so dilapi¬ dated is its condition. The church is very like Kirkbride in Durisdeer Parish (No. 155) in plan and detail, and probably dates from the early i6th century. Ivi. N.E. 9 June 1912. Defensive Construction. 98. Fort, “ Range Castle,” Holmains.—This fort (fig. 38) occupies the summit of a promi¬ nent hillock rising from moorland, about J mile east-south-east of Holmains farm. A shoul¬ der of this hillock, projecting to the east, is isolated by the surrounding trench, to be hereafter described, while to the south-east an eminence rising to a higher elevation is separated by a natural hollow. From the summit of the hillock the ground falls by a steep declivity towards the north-east, and though in the other directions the gradient is easier, it is still considerable. The main defence of the fort consists of a trench cut for the most part through rock, with an average breadth of 30 feet from crest of scarp to that of counterscarp, and a depth of 7 to 8 feet, and seemingly surmounted on the scarp in places by a .stony parapet mound. Within the trench the ground on the east rises to the summit rather abruptly to a height of 15 feet, and on the west by a longer slope to a height of 20 feet. The area enclosed is 42 DALTON.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [DALTON. roughly circular, though coming to a point towards the south, and measures 300 feet by 280 feet. A deep hollow on the north end forms a re-entrant in the summit area. Some 45 feet back from the trench on the west, 90 feet on the south, and 104 feet on the east, has been a wall running southward from the edge of the re-entrant, and thence from the south end northward, following the crest of a steep natural scarp. Some 70 feet within this wall on the west, and 81 feet on the south, and 100 O 100 200 300 FEET HH-f-H=f Fig. 38. —Fort, " Range Castle ” (No. 98). resting on the previously mentioned wall to the east, is an inner wall of low elevation and greatly dilapidated, encircling the actual summit, which measures some 86 feet by 79 feet. The main entrance passes over the trench on the west on unexcavated ground wdth a width of 27 feet and ii feet as it passes through the mound above it, and proceeds by a track up the steep slope to pass through the intermediate wall, being faced where it does so by a rocky outcrop which continues to flank it as it turns in a southerly direction. There appears to have been a “ postern ” on the south, where a gangway crosses the trench in front of a point of rock, to either side of which the track bifurcates. A third entrance seems to have existed through a narrow gap opening into the recess on the end of the hillock. The fort commands an extensive prospect to the north at an elevation of 600 feet. 1 . S.E. 4 October 1912. Cairn. 99. Cairn, Holmains Moor.—On an emin¬ ence rising from a ridge on Holmains Moor, and about J mile west-south-west of Holmains, is a cairn with a diameter of 70 feet, which has evidently been systematically excavated. To the south of the centre a large upright slab is partially exposed, with a breadth of 3 feet 9 inches, which has apparently formed the end of a cist or chamber, but which of these it is impossible to say in the now dilapi¬ dated state of the cairn. 1 . S.E. 5 October 1912. Miscellaneous. 100. Standing-Stone, Dalton Village.—Stand¬ ing in the centre of Dalton village is a four¬ sided pillar 6 feet 5 inches in height above ground. Ivii. N.W. (unnoted). 2 October 1912. Sites. 101. Fort, Dalton.—This fort has occupied the summit of a prominent hillock, which rises to a height of some 30 to 40 feet, and overlooks the road between Kirkwood Lodge and Dalton village. The site is under cultivation, and, except for a slight bank which drops with a long scarp for some 4 or 5 feet of elevation to a shoulder on the west, there is no trace left of defences. The O.S. map shows the fort as oval, measuring 275 feet by 200 feet. Ivii. N.W. 2 October 1912. 102. Fort, Morrison House.—Above the steep slope which rises from the Dumfries and Annan road, just by the farm to the westward of Morrison House, is the site of a circular fort, now only discernible by a slight glacis, where a rampart has been. Ivi. S.E. (unnoted). 25 April 1913. The O.S. maps also indicate sites as under :— 103. Holmains Tower, Holmains. 1 . S.E. 104. Little Dalton, ^ mile north-west of Kirkhill. Ivi. N.E. 105. St Bridget’s Chapel, Isle of Dalton. Ivii. N.W. 43 DORXOCK.] HISTORICAL I^IONUMENTS (SCOTLAND) COMMISSION. [dornock. DORNOCK. Castellated and Domestic Structures. io6. Stapleton Tower,—This tower (fig. 4 of Introduction) is situated some 3 miles to the north-east of Annan. A modern man¬ sion house abuts against the north-eastern and north-western walls of the tower, which has been adapted to suit modern requirements and communicates Muth the modern building on the upper floor levels. On plan (fig. 39) the tower is a simple oblong measuring externally 43 feet by 27J feet; vfith walls varying from 5 feet 8 inches to about 12 feet in thickness at the north-eastern end ; while the height from the ground to the top of the parapet measures 41 feet. The entrance, which is on the ground level near the south-eastern angle, gives access 10 5 0 10 20 FEET Il ium I- H -j-h Fig. 39.—Stapleton Tower (No. 106). to a vaulted basement and to a wheel-staircase which communicates directly with three upper floors and with the parapet walk. The original shot-holes remain in the side walls of the basement, but the window in the south¬ western wall and the fireplace at the opposite end of the room are modernised. The entrance doorway has a bold bead-and-hollow moulding wrought on the jambs and continued around the segmental arch-head, the hollow being decorated with a leaf enrichment of unusual type. The section of the upper window jambs consists of a double roll separated by a hollow containing a dog-tooth enrichment, showing the return in the i6th century to this early form of ornament. The type of corbelling which supports the stone parapet is also characteristic of the same period. As in the case of Hollows Tower, near Langholm (No. 43), it consists of a series of miniature corbels. with interspaces of the same section, below which is a continuous moulding of small pro¬ jection springing from the surface of the main wall. This corbel-table is carried round the circular turrets, which rest on corbelled pro¬ jections at each angle. The top of the parapet is crenellated and appears to have been re¬ newed within recent years. On the face of the wall above the doorway is a small recess, which originally contained a panel bearing the arms of the owner. Like the windows it has a moulded margin en¬ riched with the dog-tooth ornament. Two small openings or machicolations occur at the level of the corbel-table immediately above the entrance doorway. From their insignificant size it seems doubtful that they could have been designed for defensive purposes. A splayed basement course returns around the tower. In 1626, and for two or three years previous, Stapleton or “ Stabiltoun ” was in possession of Fergus Graham of Blawatwood “ be vertew of gude and undoubted rightis,” when, at the end of January in that year, Christie Irving, son of the deceased Fdward Irving of Bonshaw and “ sometime of Stabletoun,” accompanied by “ a number of the fugitive lymmaris of the lait Bordouris,” surprised and captured the place " airlie in the morneing afoir the break of day ” and fortified and provisioned it " as ane house of warre.” Instructions were given on 26th February, to Sir John Charteris of Amisfield, as one of the Commissioners of the Middle Shires, to recover the place and restore it to Fergus Graham ; then on 21st April to Charteris along with Sir William Grier of Lag and James Maxwell of Kirkconnell. Irving refused the demand of these emissaries, who found further that to take the place by force would mean a “ long and continueand siege,” and on 2gth July commission was given to four earls and twelve knights and lairds to evict Christie Irving and his fellows. The end of the story is not given, but it may be presumed that, with the aid of the levies of the shire, this powerful company accomplished its purpose (Reg. P.C., vol. i, 2nd series, pp. 236, 286, 378, 667.) Ixiii. N.W. 15 August 1912. 107. Robgill Tower.—This tower, which is incorporated in the modern mansion of the 44 'tf *•-«# ■ Ancitu: itnJ Monunwnts — Ditutfrics. NORTH SIDE SOUTH SIDE Fig. 41.— Coped Stone, Doinock : second stone (No. 109). NORTH SIDE SOUTH SIDE Fig. 42. —Coped Stone, Dornock : third stone (No. 109). 'J'ofacep. 45 DORNOCK.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [dryfesdale. same name, stands on the high south-western bank of the Kirtle Water, almost opposite to Woodhoiise Tower. The vaulted basement, which has been used as a kitchen, is the only part of the original work now remaining, the upper storeys being obviously part of the modern reconstruction. On plan (fig. 40) the tower measures externally 34 feet 6 inches by 24 feet, with walls averaging some 5 feet in thickness. The entrance in the south-east wall gives access to a wheel-stair and to the kitchen, which is now divided into two apartments by a modern partition. The Fig. 40.—Robgiii original fireplace, measur- Tower (No. 107). ing 8 feet in width, with an aumbry adjoining, remains in the south-west wall. A lamp recess is formed in the wall opposite the stair-foot, and in the vault of the entrance passage is a square hatch. The open¬ ing of this hatch has been built up, and in the centre is placed a pendant, similar to the one in the entrance at Bonshaw, with the mono¬ gram TH-S carved in relief on the exposed surface. Indications of a splayed base remain on the exterior of the walls. This tower appears to be of the same type and period as Bonshaw Tower (No. i), and Woodhouse Tower (No. 368), i.e. i6th century. It belonged to a branch of the Irvings, and was burned by Wharton in Sept. 1544.’^ “ Cuth- bert Vrwen of Robbgill ” is on Wharton’s list of names of local gentlemen pledged to Eng¬ land in 1547.^ 1 Hamilton Papers, ii. p. 456 ; ^ Armstrong’s Liddesdale, App. xxxix. Cf. Introd., p. xxxviii. Iviii. S.W. 25 July 1912. Cairn. 108. Cairn, Robgill.—In cultivated land some 300 yards south by east of the east lodge of Robgill stands a cairn overgrown with grass and surmounted with trees, having a diameter of 90 feet and an elevation of 12 feet. A small amount of stone has been removed at some distant date from the sides, but it appears to be unexcavated. Iviii. S.W. (unnoted.) i October 1912. Miscellaneous. 109. Coped Stones, Dornock.—Some 40 feet to the south of the east end of the parish church there lie three richly carved coped stones. The stone nearest the church has suffered much from exposure to the elements, and its carving, which has been similar in character to that on the other two stones, is greatly worn away. It is partially under¬ ground and measures 6 feet 7 inches in length. The second stone (fig. 41), some 6 feet to the south of the last, measures 6 feet 9 inches in length, I foot ii inches in breadth at base, 5 inches across the flat top, and i foot in breadth on the sides. Each side is divided into four panels, those on the north side containing a fleur-de-lis carved in high relief, two designs being twice repeated. On the south side three panels contain fleur-de-lis, in this case combining with oak leaves ; the fourth encloses a cross within a quatrefoil. The east end of the stone is carved with a floriated device, and the west with a shafted cross. The third stone (fig. 42) is enriched with similar though not identical ornamentation, and measures 6 feet 5 inches in length, i foot 9I inches in breadth at base, 8 inches across the flat top, and i foot i inch in depth. It differs from the others in having a prominence at each end, some 10 inches in length and 3 inches in height, and from the stone last described in being undecorated at the ends. Each stone appears to have rested on four small round pillars with square caps, several of which remain. On the top of the stones at the east end holes have been sunk as if for an iron post. Ixiii. S.W. 30 September 1912. Site. The O.S. map indicates a site as under :— no. St Marjory’s Church, Dornock. Ixiii. S.W. DRYFESDALE. Castell.\ted and Domestic Structure. III. Tower, Lockerbie.—In the town of Lockerbie, on the east side of Main Street 45 DRYFESDALE.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [dryfesdale. and adjoining the police station, is the rem¬ nant of a tower. The structure has been reduced to two storej^'s in height and is covered with a modern roof. It is oblong on plan, measuring exteriorly 20 feet from north to south by 25 feet from east to west, and is built of local whinstone ; the corners are rounded. The windows have been re¬ newed and have modern freestone jambs and lintels. In the west wall, some 15 feet above the ground and over the entrance, which is there situated, may be seen four moulded corbels, apparently not in situ. These are probably the remnants of a corbel course bearing a parapet walk around the structure when entire. The walling above these appears to have been renewed. The building is now entered through a contiguous dwelling ; the wall at the doorway is 3 feet 9 inches thick. On the basement floors there were originally two vaulted chambers unequal in size, vith the lesser on the west. The barrel vaults run in opposite directions. The building was in use for a considerable period as a prison, until superseded by the neighbouring modern jail. There is no trace of an internal staircase. The building is in fair repair, although unused for any specific purpose. This was a tower belonging to one of the Johnstone families. In 1585 " two of the principalles of the Johnstons ” had each a stone house in the town, when both places were besieged and captured by the Earl of Morton.^ These gentlemen of the clan were Andrew Johnstone and Mungo Johnstone in Lockerby .2 In 1593 the “ howse at Locker- bye ” belonging to “ one Mongo Johnston ” was attacked by Lord Maxwell with the intention of “ demolishing and casting downe ” the same. The laird of Johnstone raised the siege, when Lord Maxwell and many of his company were slain.^ Which family owned the surviving tower cannot be said, but Mungo Johnstone is a conspicuous figure in the district in the closing years of the i6th century. Obviously this is the " old square tower still standing called the Mains,’' described in the New Stat. Acct., 1845 (iv. p. 455), as having " been placed on a ridge, between and almost surrounded by two large lochs or lakes . . . which are now drained.” ^ Calendar of Border Papers, i. No. 312 ; ^ Hist. MSS. Cojuin., Rep. XV., Appendix, part ix. p. 27 ; ^ Border Papers, i. No. 918. li. N.W. 14 September 1915. Defensive Constructions. 112. Fort, Hallmuir.—This fort is situated at the edge of a billowy ridge, about J mile south of the farm of Hallmuir and mthin a wood. It lies with its longest axis north and south, measuring some 390 feet by 255 feet (O.S. map measurement). To the east the ground is level, and rises but little above the interior ; but vdthin the fort towards the west side it dips considerably. Except on the lower side, where the defences are now difficult to follow, the fort is surrounded by a broad rampart of earth and stone, measuring some 20 feet in width at base and 4 feet in height on the interior, with a trench in front some 33 feet wide and 5 feet deep. li. S.W. 30 July 1912. 113. Fort, Castlehill, Pilmuir Common.—At the extremity of the ridge which reaches north¬ ward from the enclosure at Castlehill, and some 200 yards distant from it, is a fort, irregularly circular in form, with a diameter of some 280 feet from north to south by 285 feet from east to west (O.S. measurements). The ground has a slight fall from the south, tails away gradually on each flank, and at the northern extremity dips steeply to the level some 30 feet below. The enceinte has been surrounded by a rampart, apparently of earth, measuring some 30 feet in thickness at base and rising not more than 3 feet on the inner face, with a trench some 30 feet in vidth from crest to crest, 6 feet deep below the scarp, and some 4 feet below the counterscarp at the south end. Along the flanks and round the north end the parapet diminishes, and the trench becomes shallower. On the north¬ west, where is a slight depression on the face of the ridge, there appears to have been an entrance leading into the interior at its lowest point, and there has also been an entrance of less width, seemingly about 5 feet, by a ramp on the east side. The situa- DRYFESDALE.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [dryfesdale. tion, though only at an elevation of some 209 feet over sea-level, commands a consider¬ able prospect up Annandale. li. S.W. 30 July 1912. 114. Dryfesdale Fort, Dryfesdale Gate.—This fort has been situated immediately to the south of the farm of Dryfesdale Gate, on a point of land formed by the meeting of two high banks on the north and west. The plateau has risen to a slight eminence towards the point, and on this the fort has been formed. On the north and west there is an elevation of from 30 to 40 feet above the low haugh- land that stretches back to the Dryfe, and on the east the gradient is easier and the eleva¬ tion less. Across the plateau towards the south, at the base of the rising ground, there are visible the remains of a trench, also the entrance which has been from that direction ; but the whole area of the fort is under crop, and any mound which may have encircled the plateau has long since been ploughed down. The O.S. map shows the plan as an irregular oval, with almost a right angle towards the north-west, formed by the meeting of two comparatively straight facets. With its longest axis north-north-west to south-south¬ east it has measured some 200 feet by 150 feet. xliii. S.W. 2 August 1912. 115. Fort, Gallaberry, Dryfeholm.—This fort is situated on the crest of a ridge of land that lies between the Dryfe Water and the Annan about I mile above their confluence. It has been an oval enclosure, with its longest axis north-west and south-east, measuring in the interior some 320 feet in length, and, accord¬ ing to the line of its now demolished defences on the north-east, as indicated on the O.S. plan, some 210 feet in breadth. On the south-west the ground falls abruptly for 40 to 50 feet to the low haugh-land which stretches back to the Annan, and on the north-east it has a long and easy gradient towards the Dryfe ; towards the north-west and south-east it falls by a gentle inclination. Along the edge of the bank to the south-west there appears to have been no need for sub¬ stantial defences, and, as already stated. in the cultivated land on the north-east whatever lines were there have disappeared ; only at the ends of the oval do they now remain. At the north-west extremity these are very formidable, and consist of three ramparts with intervening trenches. The inner rampart rises some 4 feet above the interior, has a breadth at base of some 40 feet, with a scarp 12 feet in elevation above the inner trench, which has a width from crest to crest of 40 feet and 4 feet at bottom. The second rampart has a breadth from the bottom of the trench on either side of it of 46 feet, and rises 7 feet above the inner trench and 4 feet above the outer. The second trench is 25 feet wide at top, 5 feet at bottom, and 4 feet 6 inches deep below the crest of the outer mound, which has a breadth of 25 feet at base and an elevation of about 2 feet to the exterior. The ramparts converge towards the edge of the bank. At the south end the defences are not so well defined. The parapet mound at the end of the interior has a height of some 18 inches to 2 feet on the inside ; in front of it is a trench.some 27 feet wide and about 5 feet deep, beyond w^hich rises a rampart some 25 feet wide at base and 4 feet high to the outside. In front of this there appears to be a trench, now difficult of discernment, among the undergrowth, seemingly 40 feet wide, but the counterscarp being ill-defined this dimension must be regarded as approxi¬ mate. At this end, on the south-w’est, the bank does not fall so abruptly ; and some 10 feet below the summit level there has been a trench, now^ shallow’, about 7 feet w’ide, with a low mound on the counterscarp. A modern road in a deep cutting passes through the fort tow’ards the south-east end. The interior is overgrow’n with trees and a thick undergrowth of raspberry canes and bracken. xlii. S.E. 2 August 1912. 116. Fort, Raggiewhate.—At the upper edge of a field, some 200 yards east by south of Rag- giewiiate and close by the side of the road, arc the remains of a circular construction, measur¬ ing some 76 feet by 90 feet, surrounded b}’ a mound now 24 feet wide at base and rising but little above the level of the interior, and scarped on the exterior to a height of from 2 to 4 feet. The rampart is much rounded, 47 DRYFESDALE.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [Dumfries. and has probably been under cultivation at one time. li. N.E. 13 August 1912. Exclosures. 117. Enclosure, Castlehill, Pilmuir Common. —In a lir wood immediately in rear of the cottages at Castlehill are the very faint remains of an oblong earthwork with rounded ends. It lies with its main axis north-west and south¬ east, measuring some 230 feet by 140 feet, and has been surrounded by two ramparts of earth and stone some 18 feet broad at base with an intervening trench 21 feet wide. The ramparts are now only noticeable as slight swellings on the surface, at no place rising to a greater height than 2 feet above the bottom of the trench. The enclosure has been divided into two parts by a trench 14 to 15 feet wide, and now very shallow, across the centre. There has been a separate en¬ trance into each division. The area around is very level, and the situation is not particu¬ larly defensible ; it commands, however, a fine prospect of the dale of Annan to the northward. li. S.W. (“ Fort ”). 30 July 1912. 118. Enclosure, Castlehill.—At the edge of the Castlehill, overlooking Bengali farm and about J mile south-south-west of Castlehill cottages, is a circular enclosure. On its south¬ west side it has been considerably hollowed in the interior, so that the floor lies some 3 feet below the level of the surrounding ground, though on the north-east half it apparently remains at the natural level. It has been surrounded by a stony bank some 16 feet wide, now reduced to a trifling level on the exterior. The entrance seems to have been at the lowest point. li. S.W. (“ Fort ”). 30 July 1912. Sites. 119. Little Hutton Chapel, remains, Hall- dykes. xliii. S.W. The O.S. maps also indicate sites as under :— 120. St Michael’s Chapel, Hillside. xliii. S.W. 121. Church, Kirkton. xliii. S.W. 122. Chapel and Chapel Well, J mile west of Burnbrae. xliii. S.W. 123. Tower, Kirkton. xliii. S.W. 124. Fort, Fairholm. li. N.W. 125. Forts (2), Eskrig. li. N.W. 126. Tumulus, about 300 yards south-west of Dryfeholm farm. xlii. S.E. DUMFRIES. Municipal Structure. 127. The Midsteeple, Dumfries.—The town- house with its picturesque “ midsteeple ” (fig. 43) occupies a prominent position in the heart of the town. To the north it faces the High Street, which diverges on either side and leads to the open market square beyond, leaving the building isolated in the middle of the street. On plan the town-house is oblong and three storeys in height. At the north-east end is the square steeple. The town-house measures some 33 feet 6 inches by 19 feet within walls 2 feet 6 inches in thickness ; and the steeple about 17 feet square over walls averaging 4 feet 6 inches in thickness. The tower is carried up three additional stages to a parapet, above which it terminates in a wooden fleche covered with lead. Modern shops abut against the exterior at the northern end. The whole of the ground and first floors of the town-house, originally containing the weigh-house, town-guard house, and town-hall, are now converted into show¬ rooms, while the second floor is used as a workshop. A recently formed doorway on each floor, within the north-eastern angle, gives access from the main building to the tower at each of these levels. The wide open staircase at the south end remains and leads from the street to the main entrance at the first floor. It seems that the original intention was to enclose this staircase with a stone parapet wall, instead of which a wrought-iron balustrade was erected. The original of this decorative iron-work was forged by an Edinburgh smith.^ It consists of a series of panels, each containing two circular scrolls and separated from each Ancienl and Historical Mominiejiis — Dumfries. Fig. 43.—The Midsteeple, Dumfries (No. 127). To face 48 Vv^C: iV'V-. - ' ■ . ‘'''' ■v:-i . \ VTi*.. ' - • " 'X‘ 'c ■■. • ♦ I;., .'7, •'•'<>• . .■.■ ■?««». '■' „■?«, ’ ii, \-' ;s| '■ ■■■''"C ■•^■' '■' I'.'-i^^''' ■ ■■'■ - ■>'"- ■■:■ • ■■ i;'''''"' ■'■.1 ,- ■ j-»’V'. V • «.y.', ■' kUt‘ 1 ; V ' 'v''i^^''i''^rf'^\/'' '■ ■''xi?!f''’.'i^'''3;' t'.- ' ■..■■■■ y. '. '.'■■■ ' ■f'^v . .A % -i' /^'•♦•''' ' .>V. ^ • -f||^:- 4 :lAi ■iip4vfv ,,■' ^ . fr. ■■■r »\ ■ a £L-' IN&'- ' ' r, 'CM •• 7 .;•■> ■ ■*• ij ■; •■> ■ . ^ DUMFRIES.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [DUMFRIES. other by upright balusters representing thistles of conventional design. At the south-western angle of the tower a roughly-built outside stair gives access to the wheel-staircase near the level of the first floor, whence it leads directly to the third floor. The upper stages of the tower are of timber and are reached by ladders. The building has been refaced, where necessary, with red-sandstone, as the surface of the original masonry had decayed in course of time. The simply treated exterior shows Renais? sance characteristics. The main entrance doorway at the south end has moulded jambs flanked by pilasters with Ionic caps supporting a moulded architrave, over which is a plain frieze surmounted by a cornice and segmental pediment. Incised on the surface of the frieze, on either side of the central keystone, is the date 1707. The stone parapets are pierced at regular intervals with square openings, which contain traceried infillings of a circular form. On the surface of the western wall are two carved panels placed one above the other. The upper panel contains a central shield bearing the Royal Arms of Scotland, with supporters—two unicorns gorged and bannered. Over the shield is tlae helmet and crown with mantling and crest. The banneret on the dexter side bears the Scottish Lion within the double tressure, and the banneret on the sinister side the St Andrew Cross. Below the shield is an oval containing a representation of St Andrew, with a thistle carved in relief on either side. At the base the motto “ Nemo me impune lacesset ” is carved in raised Roman capitals. The lower panel contains the representation of a mitred figure, bearded and winged, clothed in a long robe reaching to the feet, and holding in the left hand a pastoral staff with the crook turned inwards. This figure is for St Michael. At the lower dexter angle of the panel is what appears to be the coiled tail of a serpent, and at the opposite angle the point of the staff rests on what may represent the serpent’s head. On the west side of the town-house, near the doorway to the turret staircase, two old stones are pre¬ served. The upper bears, in raised lettering, the initials H R, with a pair of shackles be¬ tween, and R MK flanking a bow and arrow. Underneath is the word “ BAillieS.” The initials are those of Herbert Raining and Robert M'Kinnell, who were “ baillies ” when the old prison was built in 1579 the east side of the High Street.^ On the lower stone (fig. 44) is a shield bearing a chevron between three fleur-de-lys, and the motto A LORBVRNE ; below is a figure like a rude crown. This stone was for a time at Knockhill with many others (No. 273). It Fig. 44. —Inscribed Stone, Midsteeple, Dumfries (No. 127). is recorded in The Statistical Account of Scotland, vol. v. p. 122 (1793), as “ still to be seen on the front of the prison,” which was “nearly adjoining” the town-house (p. 123) ; and the account states that the arms are said to have been the ancient arms of the town, and suggests that the stone which bears them may have been a part of the old prison. It seems clear, indeed, that all the stones in question were originally part of that late sixteenth-century structure. The building was erected by the town authorities, under the direction of Tobias Bachup of Alloway .3 ^ M'Dowall’s History of Dumfries, p. 461 ; - Shirley’s Growth of a Scottish Burgh, p. 31 ; 3 M‘Do wall. Iv. N.E. 6 August 1912. Defensive Constructions. 128. Castle Dykes, Dumfries.—The policies within the wall of this property contain certain mounds and ditches identified as the site of the old Castle of Dumfries. The whole ” plesaunce ” has been greatly interfered with to increase its amenity in modern times. A mansion-house has been erected on what was evidently a portion of the enclosure ; part of the great ditch has been filled up and a garden lies on the site of it ; paths and road- 4 49 DUMFRIES.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [Dumfries. ways have been excavated, and mounds are said to have been erected, so that it is very difficult to identify the original constructions. A small rocky eminence {“ Paradise ”),* ris¬ ing abrupth’ from the road, which intervenes between it and the left bank of the Nith, appears to have been converted into a mote hill by levelling up the summit, by scarping the flanks, and by cutting it off by means of a trench from a lower plateau on the south-east. Towards the river the frontage has, no doubt, been altered to some extent in the formation of the road, and the rock has been exposed and cut back on the lower part of the slope. The plan is oval in form, with its longest axis east and west, and measures some 76 feet by 44 feet. The front of the plateau on the south¬ east facing this mote appears to have been dug into in more recent times ; and the trench, which was probably continued right across this area, destroyed. It is possible that this plateau, as its situation suggests, formed a base-court ; there is, however, no indication of a trench around it. A steep slope on the west is an advantageous feature of the situation. The highest portion of the grounds to the eastward, lying north-north-west and south- south-east, is a plateau to some extent of artificial construction. The northern half, on which the house of Castledykes is built, ter¬ minates in a somewhat steep slope towards the Nith. The southern half, having a length of 190 feet or thereby from the carriage-drive which forms the division, still retains evidence of its original fortified character. On the east side, from a point adjacent to the entrance lodge, running almost parallel with the Glen- caple road till it turns sharply to form the southern extremity of the plateau, is a bold trench, measuring some 80 feet in width at the top, and varying in depth from 20 to 30 feet. A small stream runs in the bottom of this trench on the east, and is conducted in a built conduit along a line which probably follows the continuation of the trench north¬ ward, turning to the west behind the mansion- house. The original length of the plateau is now indefinite, but its breadth across the southern half varies from 74 to 97 feet. Along the west side, save for a slight hollow observ- * " Paradise ” (Gk. paradeisos, a garden), a pleasure garden, or an ornamental feature in a garden. able among the bushes to the south-west of the house, the trench has disappeared, though the scarp is distinct along the southern division. Near the centre of that side there projects a semicircular bastion, with a terrace halfway up its slope. This construction lies at a lower level than the plateau, from which it is sepa¬ rated by a carriage-drive, and it is doubtful if it forms part of the original earthwork. Historical Note.— This is the ancient royal castle of Dumfries, known already to¬ wards the close of the 12th century as the “ old castle ” {vetus castellarium) An entry in the English revenue accounts of 1335, during the time of the English occupation, relates to “ the mote of the castle and certain royal lands called Kingsholm.” As Kingholm is the name given to the river-flat in the immediate neighbourhood of Castledykes, this confirms other evidence as to the site of the castle. A description, c. 1563-66, furnished by an English agent, runs thus ; “ The oulde Castell of Dumfries, fyve miles and a half within the moutht of Nytht, standing upon the same, verye good for a forte. The platt and ground thereof in manner lyke to Roxburght Castell.” “ This again fixes Castledykes as the position. Dumfries was a royal castle, and was seized by Robert Bruce, afterwards the “ Competitor ” for the Crown, in the rising which he initiated immediately after the death of Alexander III. in 1286. During the English supremacy under Edward L, it is associated in wardenship with the other royal castles of Kirkcudbright and Wigtown. In 1298 it has a garrison of 76 all told—men with armoured horses, foot including crossbow-men, and various details, such as engineers (for the defensive engines), carpenters, and masons. Two years later the castrum is enclosed with a ” peel ” or palisade [assidendum circa castrum), material for which had been prepared by carpenters in the forest of Inglewood, Cumber¬ land. Round the peel another ditch was dug. Castrum and pelum were both fortified en¬ closures, containing dwellings, which, like the enceinte, were then mostly, if not wholly, of timber. Each enclosure possessed its own garrison. Dumfries Castle was one of those seized by Robert Bruce immediately after the assassina¬ tion of John Comyn in February 1306, but 50 Aiiiiiiit and Hislorital Monuments — Dumfries. To face /. 51. Fig. 45.— Dumfries Bridge (No. 131). DUMFRIES.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [DUMFRIES. was held on his behalf for only three weeks. It was surrendered by Sir Dugal M'Doual to Bruce on February 6, 1313, as one result of operations against the castles in the district. During the hostilities which followed the captivity of David IT, the Nithsdale castles had proved sources of serious annoyance to the English, and, before that monarch was released in 1357, he had to bind himself to destroy these places. Accordingly, Dalswinton, Dumfries, Morton, and Durisdeer were levelled to the ground, and three-quarters of a century later they were still, in great part, in a ruined condition.® 1 Registrum Epis. Glasg., p. 42 ; ® Arm¬ strong’s Liddesdale ; Bain’s Calendar, iii., iv. ; Stevenson’s Documents ; Liber Garderobce ; ® see Introd., p. xxxiii. ; Scotichronicon, xiv. 18. Iv. N.E. 30 March 1915. 129. Mote of Dumfries.—The mote-house is a large building, formerly a dwelling-house and now connected with the Dumfries Academy, situated immediately to the north of that institution, between Academy Street and the River Nith. It stands upon an eminence, apparently natural, sloping up from the north¬ wards, with a high steep bank towards the river, and a retaining wall where the flank has been cut back for the roadway on the east. Within the grounds of this mansion, though cut and carved for lawns and path- w'ays, there are still massive artificial mounds, evidently the remains of the mote and its base-court. The mote itself has probably stood on the highest point which lies in rear of the house adjoining the Academy, and has been almost entirely removed, the site being occupied by a croquet-lawn. On the north¬ west of this lawn is a broad, flat-topped ridge of soil, which seems to be a small segment of the original construction, while above the street, on the east or south-east, the scarp of the mote appears to be recognisable. If this is the case the mote has been a very large one and rather square in form. To the northward of the house a massive rampart-like mound runs parallel to the street, rising some 25 feet above it and 6 or 7 feet above the level of the lawn on its inner side, and gradually diminishing till it disappears behind the lodge. Some 50 feet from its termination another mound, less massive and sharper at the apex, as if more recently fashioned, diverges from it and curves round towards the north-west. These mounds seem to have been connected with the base-court ; but without some know¬ ledge of the transformation which must have taken place all over this ground, it is impossible to account for these earthworks with any certainty. Save for the suggestive name of the house and of the Moat Brae by the water¬ side, it would have been difficult even to infer a mote on the site. An investigation by an English agent a few years after the Reformation gives us this note : “ The towne of Drumfreiss is subjett to two lytill motes, one called the Beakin hill whiche in thre days may be (? removed) be the men in the watter of Nytht : the other at th’ east gate, where upoune the lytill chapcll standetht hard by the towne, but removi- able by less lauboures nor th’ other, onless in case of fortification, yt might for that qwarter of the towne be made a moute or bulwork.” ^ ^ Armstrong’s Liddesdale, Appendix No. Ixx., p. cx. xlix. S.E. 29 March 1915. 130. Fort, Camp Hill, Trohoughton.—This fort is situated on the crest of a ridge, at an elevation of 300 feet above sea level, some 2\ miles to the south-east of Dumfries, on the road to Bankend, and commands an extensive prospect all around. It is a circular enceinte measuring in diameter 189 feet by 198 feet, surrounded by a single trench with a breadth of some 33 feet and at most a depth of 7 feet below the scarp and 4 feet below the counterscarp, the trench being supplemented by an inner and outer rampart of earth. Somewhat less than half of the fort is in cultivated land under plough, and here the defences have been much reduced. Iv. S.E. 25 April 1913. Miscellaneous. 131. Old Bridge, Dumfries.—This important structure (fig. 45), which spans the River Nith at the western side of the town, links the shire of Dumfries with the Stewartry of Kirkcud¬ bright. It originally consisted of nine arches, but, on the reclamation of the eastern bank 51 DUMFRIES.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [Dumfries. of the river at the beginning of the 19th century, the three eastern arches became unnecessary and were removed, leaving the remaining six arches spanning a wddth of 263 feet. The arches are drop-centred, varying in width from 27 to 35 feet, and in form from the obtusely-pointed westmost arch to the segmental arch on the east ; the four inter¬ mediate arches are slightly pointed. There are no ribs on the soffits. On either side of the bridge sturdy cut-waters project from the faces of the piers and terminate in sloping The building of this bridge has been usually attributed to Dervorgilla, foundress of New Abbey in the late thirteenth century, but the following extract seems to give both an approximate date for the erection of the structure and a statement of those to whom it was due. “ 1431 [2]. Relaxation, valid during twenty years only, of a hundred days to those who during the said octaves and days visit and give alms towards the building of the bridge which has been recently begun over the river Nyth near the Burgh of Dumfries in the XiPKCUDBSlOflTSHIRE UEVEL OF RIVER NITrt' SOUTH ELEVATION DUMrRlE^.iMIRE Fig. 46.— Dumfries Bridge (No. 131). tops, save at the central pier, where the cut-water is carried up to the level of the cope of the parapet, thus forming a refuge (fig. 46). The roadway is slightly arcuated, and averages 13 feet in width. At the western extremity the parapets diverge from the alignment in a north-westerly direction, possibly for defensive reasons. The bridge is reached from the eastern bank by a flight of steps, necessitated by the removal of the three eastern arches, which are shown in a drawing of 1747 in Grose’s Antiquities of Scotland along with a “ small gate called the port,” apparently a toll-gate, removed in 1769. diocese of Glasgow by the burgesses and in¬ habitants of those parts, and also for the amplification of the Chapel of St Mary the Virgin founded near the said bridge.” ^ That there was a bridge at Dumfries before this date is clear from a charter in 1426 to the Friar Minors of Dumfries, by Margaret, Countess of Douglas and Lady of Galloway, confirming a grant of the toll dues hitherto paid to herself and her predecessors at " the bridge of Nith at Dumfries.” This early bridge was in all probability of wood. There is a later charter of confirmation of the same grant by James, ninth and last Earl of Douglas, in 1453. Later the Friars leased the “ brig custum ” to Edward Johnstone, a burgess, and on his death 52 DUMFRIES.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [DUMFRIES. to his son John (1557) and John’s heirs and assignees. The annual rent of ten marks, how¬ ever, was increased by three and fourpence or a quarter of a mark. After the fall of the Douglases in 1455, the Crown appears to have undertaken the obliga¬ tion of the upkeep of the bridge, for which purpose a master of works was appointed and royal grants made, the last of these in 1465. At the beginning of the i6th century the burgh was drawing half the tolls, and appointed annually a master of works. Not till 1623 did the burgh secure the other half from the descendant of John Johnstone. In 1620 the river in high flood wrecked about one half of the bridge. By a great effort, which not only exhausted the burgh exchequer but also impoverished the burgesses by volun¬ tary contributions, the fabric was reconstructed within a year. A corbel on the north side of the eastmost arch bears the date 1610 and the initials I. S., but this stone was taken from a house undergoing reconstruction about thirty years ago, and inserted in the bridge then under repair. The structure is in a thoroughly sound con¬ dition and is still in use for pedestrian traffic. 1 Calendar of Papal Registers, Papal Letters, vol. viii. 1427-1447, p. 347. M'Dowall’s History of Dumfries ; Shirley’s Growth of a Scottish Burgh ; The Scottish Grey friars, W. Moir Bryce. Iv. N.E. 18 July 1912. 132. Architectural Fragments from the Max¬ well House and St Christopher’s Chapel.—A collection of architectural details, formed by the late James Barbour, is built into rockeries in the garden of a villa, St Christopher’s in St Mary Street. There are four heraldic panels reputed to have come from the 16th-century town-house of the Maxwell family, which stood on the site of the modern Greyfriars Church.^ One which is placed upside down, lies beside the entrance. The shield bears a saltire with a cushion in base, all within a bordure charged with roundels. The remaining panels are in the back garden. Of these one is charged with a saltire; a second bears two coats impaled: dexter, three hunting horns stringed (Forrester) ; sinister, three crescents. two and one (PEdmonstoun) ; while a third bears: A saltire charged at the fess point with a bearded face and in base an oak tree eradicated. The undernoted fragments of ecclesiastical detail are in a rockery in the back garden, and probably came from the ruins of St Christopher’s Chapel (?), which stood on an adjacent site now occupied by St Mary’s Church. (A) A parapet corbel, hollow-moulded on the face and with a projection shaped like an ape’s head. This detail is very similar to that found in transitional work. (B) A finial with bunched foliage, of Early English type. (C) An octagonal stone basin 5 inches deep. It is I foot I inch in height by i foot 4 inches in diameter at top and tapers slightly to the base. ^ Shirley’s Growth of a Scottish Burgh, p. 49. Iv. N.E. 26 March 1915. 133. Gravestones, St Michael’s Churchyard. —Situated on a slight eminence towards the southern end of the town of Dumfries is the church of St Michael with its surrounding graveyard. In the churchyard the great majority of the monuments are table-stones, that form having been a long prevailing fashion, and very few bear emblems or aught else besides the inscription. Near the entrance gate on the right, and with its back to the street, is an upright architectural monument with fluted pilasters on either side of a central panel. Above the panel is an inscription in relief —ix memoriam viri optimi hvivs VRBIS (c)OXSVLIS JOAXXIS CORSAXI, FILIVS HOC MOXVMEXTVM (EXTRVXIT)*—QUI OBIIT 7 MAii 1629. The inscription on the panel is for the most part weathered away, and the above-quoted inscription is also rapidly decaying. A short distance further from the gate on the same side is another architectural tomb of larger dimensions. On either side of a large central panel with a semicircular panel resting on it arc two composite columns sup- * So read in M'Dowall’s Memorials of St Michael’s Churchyard. On Corsaxi see M'Dowall’s History of Dumfries, p. 209. Consul in these inscriptions is for Provost. 53 DUMFRIES.] HISTORICAL :\IONUMENTS (SCOTLAND) COMMISSION. [Dumfries. porting an entablature and pediment. The inscription in relief on the central panel has been re-cut, and reads —franxisco Irvingo COXSVLI / CHARISSIMO MARITO GRATA CONJVNX/ ET r(?p)atri PROVIDO PIA PIGNORA JVSTA HAEC / SED IVSTO LOXXE INFERIORA SACRA- RVXT OBIIT 6 NXVEMB. 1633 AETAT AN 68. ANE EPITAPHE KING JAMES AT FIRST ME BALIVE * NAMED DRVMFREIS OFT SINCE ME PROVEST CLAMED GOD HAST FOR ME ANE CROVNE RESERVED FOR KING AND COVNTRIE HAVE I SERVED. Around the arc of the semicircle runs the legend, memento homo quod cinis es et IN CINEREM REVERTERIS, With the date 1633 above in the centre, and enclosed within it are the following emblems of mortality and resurrection: a skull and cross-bones, two winged cherub heads, and an open Bible, with the initials F. 1 . and A R. on either side of the skuU. Irving’s wife was a daughter of Herbert Raining (c/. p. 49). Along the frieze of the central section of the monument is inscribed insignvm irvingorvm exegesis ILICIS EN INSTAR VIVVS VIRTVTE VIREBAM, NVNC VIREO COELO PROVIDVS ANTE SOLO. In the frieze to left and right, in panels enclosed in foliageous ornament, are respectively the initials F I and A R. Between the pillars on either side are comparatively modern inscrip¬ tions commemorating members of the family of Maxwell of Barncleuch and Glengaber. This monument much resembles the Ewart tomb in St Cuthbert’s Churchyard, Kirkcudbright.^ To the north-west of the main door of the church, and distant some 30 feet, lies a much- wxrn slab broken in two, the halves of which are wTongly placed with relation to each other. On the upper half of the stone in the centre is a skuU and cross-bones with a scroll inscribed memento mori around it, and beneath it a winged hour-glass. On the left of the central emblem is a crown above the initials H G., M. G ; on the right a thistle with the initials E W., while the date 1700 is placed half on one side and half on the other. On the lowxr portion of the stone * "We find him sitting as iMember for the Burgh in the Parliament of 1617, and high in favour at Court, receiving from James VI. bailiary jurisdiction over some Crown property in the county.” (M'Dowall’s History of Dumfries, p. 212.) has been a skull much worn away, cross-bone placed cross-ways, and, set diagonally, a winged cherub’s head. At the base, beneath a five-petalled heraldic rose on the left, and a four-petalled rose on the right, is inscribed THIS stone PERTEANETH TO H G . E W. Towards the eastern end of the older portion of the churchyard, and adjacent to a tall granite obelisk—a martyr’s monument—are two covenanters’ tombs repaired and re¬ erected in 1873. The northmost bears the inscription —here lyes william / grierson PENTLAND / MARTYR FOR HIS / ADHERING TO THE / WORD OF GOD AND / APPEARING FOR CHRI / ST’S KINGLIE GOVERME / NT IN HIS HOUSE AND / THE COVENANTED WO / RK OF REFORMATION : A / GAINST PERJURY AND / PRELACY EXECUTED / JAN. 2, 1667, REV. 12 II. UNDER THIS STONE LO HERE / DOTH LY / DUST SACRIFIC D TO TYRANY / YET PRECIOUS IN IMMENULLS / SIGHT SINCE MARTYR’D FOR HIS / KINGLIE RIGHT : / WHEN HE CONDEMNS/ THESE HELLISH DRUGES / BY SUFFRAGE SAINTS / SHALL JUDGE THEIR JUDGES. The other slab lying adjacent bears also an incised inscription —here lyes william / WELSH, PENTLAND / MARTYR, FOR HIS / AD- HEREING TO THE / WORD OF GOD AND / AP¬ PEARING FOR / CHRISTS KINGLY / GOVERN¬ MENT IN HIS / HOUSE AND THE CO- / VE- NANTED WORK / OF REFORMATION : / AGAINST PERJURY / AND PRELACIE EXE- / CUTE JANR. 2 ; / 1667. REV. 12, II. STAY, PASSENGER, READ, / HERE INTERR’d DOTH LY / A WITNES GAINST POOR / SCOT¬ LAND’S PERJURY, / WHOSE HEAD ONCE FIX’d UP / ON THE BRIDGE PORT STOOD / PRO¬ CLAIMING VENGEANCE / FOR HIS GUILTLES BLOOD. The third grave is situated about 8 yards from the others, in a northerly direction, and the 'epitaph in this case is as follows ;—here LYES lAMES / KIRKA MARTYR / SHOT DEAD UPON / THE SANDS OF DR / UMFREIS FOR HIS / ADHEREING TO THE / WORD OF GOD, CHR / ISTS KINGLY GOVE / RMENT IN HIS HO / USE : AND THE COV / ENANTED WORK OF / REFORMA¬ TION AG / AINST TIRRANIE, / PERJURIE, AND PR / ELACIE, 1685. REV. / 12, II. MAR. 54 DUMFRIES.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [duxscore. BY BLOODY BRUCE AND WRETCHED / WRIGHT / J LOST MY LIFE IN GREAT DESPIGHT/ SHOT DEAD, WITHOUT DUE TIME / TO TRY / AND FITT ME FOR ETERNITY. / A WITTNES OF PRELATICK RAGE / AS EVER WAS IN ANIE AGE. ^ Inventory of the Monuments of Galloway, vol. ii. [Stewartry of Kirkcudbright), No. 258. Iv. N.E. 28 May 1912. 134. Bell, Maxwelltown Museum. — In the Maxwelltown Museum is preserved one of the oldest outer bells in Scotland, which is said to have come from St Michael’s Church, Dumfries. It is inscribed thus in small, delicate, Lombardic letters : ; wilhelmus : DE ; CARLEIL ; DOMINUS : DE : TORTHOR- VALDE : ME : FECIT I FIERI / : IN I HONORE I SANCTI : MICHAELIS ; ANNO : DOMINI : MIL- LESSiMO : cccc : xxxx : iii, i.e. “ William of Carlyle, Lord of Torthorwald, caused me to be made. In honour of St Michael, A.D. I 443 -” At the top of the waist is a small medallion with an inscription no longer legible. A shield charged with a chevron design and a bell probably stands for the founder’s mark. Diameter 15I inches. Note G. Site. The O.S. map indicates a site as under :— 135. Tower, Place of Craigs.—Immediately south of the tower is “ Barnkin of Craigs.” Iv. S.E. DUNSCORE. Castellated and Domestic Structures. 136. Lag Tower.—This grim, broken little dwelling is situated on a knoll adjoining the farm of Lag, in hummocky ground sur¬ rounded by hills miles north-north-east of Dunscore village. The tower is oblong on plan, measuring externally 29 feet 6 inches from north to south and 25 feet from east to west, with walls 5 feet 10 inches thick on ground floor. The entrance, which is in the south wall, opens to the right on a wheel- staircase, 3 feet 3 inches wide, leading to the upper floors. The only window on the ground floor is a narrow slit in the east wall, widely flauned to the interior ; there is no fireplace. There are three storeys above the basement. The first and second floors have fireplaces in the north walls and windows with pointed scoinson arches. The building has not been vaulted. There is a garderobe recess on the first floor at the w'est end of the north wall, and a linteUed opening at ground level in the north wall communicates with a circular garderobe flue to the second floor. A court¬ yard wall runs diagonally from the north¬ west angle for a distance of 38 feet, where it returns southwards 27 feet to an arched gate¬ way 6 feet wide. Against the walls in the courtyard are traces of out-buildings. The tower appears to belong to the i6th century. The first authenticated Grierson of Lag is Gilbert ‘‘ Grerson ” in the 14th century, whose son, another Gilbert, married in 1412 one of the three heiresses of Sir Duncan of Kirkpatrick, lord of Torthorwald.^ In 1545 ” John Greorsoun of the Lag ” is one of the sureties for Robert, Master of Max¬ well, the Warden.'^ ^ Trans. Dumf. and Gall. Antiq. Soc., 1915-16, p. 38 ; ^ Reg. P.C., vol. i. p. 9. xl. N.E. Visited 23 July 1919. 137. Sundaywell Tower.—Sundaywell is a square tower, \vith very thick walls, which now forms part of a modern farm-house. Over the doorway of the modern porch, a square stone has been inserted with the initials I.K. and S.W. cut at the top, and at the foot the date 1651. Between is a shield, bearing a saltire and chief, the latter charged with three cushions. The initials are those of James Kirko, retoured heir to the seven merkland of Sundaywell in 1647, wEo probably built the tower, and of his wife, who is believed to have been a Welsh of Colliston. ” Amer Kirkauch of Soundaywell ” appears in sasines of 1599. The Kirkos ceased in the male line in the last quarter of the 17th century.^ The house is still inhabited but entirely modernised. ^ Trans. Dumf. and Gall. Antiq. Soc., 1914-15, PP- 233 ff- xxxix. S.E. \hsited 5 June 1912. 138. Bogrie Tower.—This tower, which at one time belonged to another branch of the Kirk or Kirko family, is now reduced to a two-storeyed shepherd’s house. It was partlj?^ DUNSCORE.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [dunscore. taken down in iS6o, when some of the stones from the upper store}’ were used to make repairs on the farm-steading. The tower appears to have been larger and stronger than that at Sundaywell. Over the arched doorway is a stone, partially defaced, bearing the initials I.B.W. and the date 1770. Be¬ neath is a coat of arms of the same period :— Quarterly ist and 4th : Three boars’ heads, couped, surmounted of a bend; 2nd and 3rd; Two batons in saltire with a thistle in chief, and a dagger, point upwards, in base : over all, a helmet with mantling, from which issues a man’s face, vith motto on a scroll, “ Fear God.” The first and fourth quarters in the arms are Gordon of Lochinvar, and the crest and motto are theirs, the ” man’s face ” being probably a bad representation of a demi-savage. The second and third quarters are not so certain ; but what appears to be ‘‘ crossed batons ” may really be a saltier couped, as occurs in many old stones of the Maxwell arms. Over the back door is a panel on which are cut, in relief, the initials IK • IM and the date 1660. In 1639 John Kirkoe was pro¬ prietor of Bogrie, and his wife seems to have been Jean Maxwell.^ ” Johne Kirko ” of Bogrie had his oxen stolen by Johnstones in 1599.2 The family ended in an heiress early in the i8th century, and she married a Gordon.^ ^ Trans. Dumf. and Gall.Antiq. Soc., 1914-15, p. 226 ; 2 Reg. P.C., iv. p. 844 ; 2 Trans., p. 230. xxxix. S.E. 5 June 1912. Defensive Constructions. 139. Fort, Sundaywell. — On a slightly elevated plateau, half-way between the farms of Sundaywell and Bogrie, and washed at its base on the north by a burn, are the re¬ mains of a fort. From the site the ground rises by an easy inclination to the westward, while in the opposite direction it drops with a steeper gradient. The north flank is protected by the glen of the burn ; on the south lies cultivated land, from which all traces of the defences have disappeared. The fort has been four-sided, the west side being very strongly defended. A scarp from the summit level, 4 feet in height, drops to a broad flat- bottomed ditch 33 feet across, beyond which there rises a rampart which, at greatest elevation, has a height of 4 feet. Beyond this, with another broad ditch intervening, is a second rampart with a greatest height of 5 feet 6 inches, protected in front by an outer trench some 40 feet wide. Going north the defences assume an almost rectangular form, and the inner angle in the north-west corner is somewhat sharply rounded. As the space between the plateau and the burn diminishes, the inner rampart is eliminated, and the outer converges to meet a terrace, half¬ way along the north side, some 12 feet below the summit level and considerably more above the burn. This terrace gradually disappears as it passes towards the end of the long slope down from the east end of the plateau. There is a suggestion of a wall across the east end of the plateau, but no trench or rampart is now visible in that direction. On the west a field dyke crosses the enceinte, beyond which lies the cultivated land. The total length of the interior, from the edge of the scarp at the west to the line of the supposed wall, has been some 260 feet ; the breadth is no longer obtainable. Across the second trench near the middle of the west side, is a wall foundation with a gap in the centre, to the northward of which the bottom of the trench lies at a level about I foot 8 inches above that to the south of it. Both ramparts show much stone exposed about their crests, and seem to have been surmounted by walls. Towards the south-west corner of the fort a grass-covered heap of small stones is marked ” tumulus ” on the O.S. map. xxxix. S.E. 5 June 1912. 140, Fort, Bogrie.—About \ mile north-west of Bogrie farm is a fort situated on the crest of a slight undulation which falls away in a long slope towards the Bogrie Burn on the south. In plan it is elliptical, surrounded by a stony rampart or ruined wall, overgrown and of low elevation. The interior lies at a slightly lower level than the surrounding ground and measures 123 feet by 94 feet. There appears to have been an entrance from the east about 5 feet wide, on the south of which there is a slight mound against what appears to be the inner face of the bank. The width of the rampart or wall is unobtainable without excavation. About 100 yards to the west, there is a 56 Ancient and Historical . 1 /oniinients — Dumfries. \ Fig. 47.—Lake Dwelling, Loch Urr (No. 144). To face p. 57. dunscore.] inventory OF MONUMENTS IN DUMFRIESSHIRE. [dunscore. grassy mound measuring in diameter about 20 feet, which may be a small cairn. Less than J mile due west of Bogrie on an old meadow is another similar mound meas¬ uring 15 feet by II feet superficially, which also may be a cairn. xxxix. S.E. 5 June 1912. 141. Fort, Springfield Hill.—This fort is situated at an altitude of 600 feet over sea-level on the summit of Springfield Hill, immediately to the north-north-west of the farm bear¬ ing that name, which lies on the side road leading from Holywood to Dunscore. The fort, which is oval in outline, lies with its main axis north-west and south-east and measures in the interior area some 250 feet by 160 feet. Along the south-west flank the ground falls away sharply, but from the north-west round by north to the south-east end of the fort the site has been more readily assailable. The interior area is enclosed at the south-east end by a broad rampart of earth and stone, rising at most some 5 feet above the interior level, and carried along the north-east side for about one-half of its length, where it seems to terminate abruptly, owing, possibly, to the remainder having been pillaged for stones. Except at the south-east end the outer defences, where they have occurred, have been obliterated by the plough, and stone walls confine the fort to its interior lines. At the south-east end is the entrance, well defined, and passing directly inwards through a series of ramparts. On the north side, or right on entering, there remains but a short portion of a single outer rampart which is returned to meet the inner rampart. On the south side the defences from the inner rampart outwards consist of a trench some 10 feet deep below the crest of the scarp, a broad banquette with a bold rampart at its edge, a second trench with a scarp having a vertical height of 10 feet, and in front of it a third formid¬ able rampart with a slight trench at its base. These defences pass round southw’ards from the entrance, and terminate where the slope of the hill becomes sufficiently abrupt to allow them to be omitted. xl. S.E. 6 IMay 1913. 142. Fort, Temple Wood, Friars Carse.—This fort is situated about i mile to the westw'ard of 57 the mansion-house of Friars Carse, on the top of a steep bank overlooking the Nith from the south, at a point where the river departing from a southerly course makes a sharp bend eastward. The fort is in form an irregular oval lying with its longest axis north-w'est and south-east, measuring interiorly some 200 feet in length by 142 feet in breadth at the centre, and broadening somewhat to its south¬ east extremity. From the north-north-w'est round by north to the south-east the steep bank of the river affords a formidable defence. ElsewEere the area is bounded by a rampart rising but slightly above the interior, covered in front by a broad trench some 45 feet in width and 5 feet in depth, beyond which rises a second rampart, now much w^orn dowm, with a slighter trench in front of it. The entrance has been from the east at a point where the bank on the north commences to slope steeply towards the river. The outer defences curve round the southern half of the oval from the flank of this approach to the edge of the bank on the w^est. Within the fort has been erected an ideal¬ ised stone circle, the erection of wiiich is described in a MS. journal entitled Antiquarian Researches, by the late Thomas Johnstone, under the date 1827. xl. S.E. 7 May 1913. 143. Fort, Gateside.—On a low’ plateau which rises some 8 or 10 feet above the surrounding level, and which crow’ns the top of a long steep slope rising from the road between Merkland and Low'er Halliday Hill, is the site of a fort. The scarp to the edge of the plateau may have been shaped to some extent, but otherwise there are no indications of defences. The plateau is somewiiat pear- shaped, measuring 205 feet or thereby by 170 feet. The farm, J mile to the south-east, bears the significant name of “ Moat.” xl. S.E. 7 ^lay 1913. 144. Lake Dwelling, Rough Island, Loch Urr- —About half-way along the east shore of Loch Urr, and about 60 3"ards out from it, is a low island some 67 j^ards in length b}^ 33 j^ards in breadth, of a regular oblong form, and composed of subangular boulders (fig. 47). It lies w’ith its main axis nearly east and west. All around, except at the east or shoreward end, the water DUXSCORE.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [dunscore. is deep ; but from the east end a gangway of stepping-stones leads to a much smaller grass- covered islet about 27 yards in length and some 17 3'ards distant, whence it is said a further gangway, sunk beneath the surface, gave access to the shore. The line of the sub¬ merged gangwa}^ which crosses obliquely to the smaller island is now about 3 feet beneath the water and some 3 feet above the bottom of the loch. The larger island has been encircled by a waU of dry rubble building, still standing towards the south-east angle to a height of about 4 feet, 7 feet thick at the east end, in on the Green Island, now a peninsula, at the side of Milton Loch.^ Inventory of the Monuments of Galloway, vol. ii. {Stewartry of Kirkcudbright), No. 491. xxxix. S.W. I June 1912. Cairns. 145. Small Cairns, Craigdasher.—At the south-east end of Craigdasher Hill, at an elevation of about 850 feet above sea-level, is a group of nine or ten small cairns measuring in diameter some 12 feet and of low elevation. 10 0 10 20 30 40 50 100 I I I I I i I I I I ! I I I i-l-f 150 FE ET Fig. 48.—Lake Dwelling, Loch Urr (No. 144). the centre of which there appears to have been an entrance, and diminishing to 5 feet in thick¬ ness as it passes along the flanks and round the western end. Within the enclosure thus formed are the foundations of four rectangular structures—one lying on the right of the medial line extending from the entrance inwards for a distance of 36 feet, and the other three placed at right angles across the island, as shown on the plan. The waUs of these structures are about 2 feet 6 inches in thickness and, like the surrounding walls, seemingly built without mortar. The eleva¬ tion of the surface of the island is some 3 feet above ordinary water level, and it has possibly been made up to some extent ; the smaller islet has a much lower elevation. The form of this island fortress is very similar to that To the north-east of the group is the site of a somewhat larger cairn, which has been almost entirely removed, while around it are about half a dozen small cairns similar to those just described. xlvii. N.E. 31 May 1912. 146. Small Cairns, Sunday well Moor.—On a plateau on Sundaywell Moor, the crest of the watershed, between the summit of Knockoure and an eminence about J mile to the east, lie a group of small cairns. One which lies to the south-west of the group and measures some 14 feet in diameter, is the east termina¬ tion of an ancient stony bank or wall which passes away westward in an irregular line for a distance of 80 yards or so, before it gradually dies away. Both cairn and dyke are equally DUNSCORE.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [durisdeer. overgrown and present similar features of antiquity. The cairns of this group vary from 12 to 19 feet in diameter, several measuring 16 feet across. They mimber at least a dozen, and are at an elevation of 800 feet over sea- level. At the north edge of the plateau is a distinct hut foundation with a circular chamber at one end, measuring ii feet over all, with apparently an oblong compartment attached to it, across the outer end of which a short length of walling for a butt or shelter has recently been built. xxxix. S.E. 31 May 1912. 147. Small Cairns, Bogrie Moor.—About I mile to the north, at the head of a secluded valley which opens out towards the farm of Bogrie, is another group of small cairns. The region is very desolate, and the heather-clad ground on which they are situated bristles with rocks and boulders. Yet near to the cairns are the remains of large enclosures, in the centre of one of which is traceable the foundations of a rectangular oblong structure, much over¬ grown, measuring over all some 29 feet by 19 feet, probably the ruins of a turf cabin on a stone base. The largest of the cairns, which rises as a grey mass from the heather at the south side of the group, measures some 17 feet in diameter. Adjacent, around a slight knoll, run the ruins of an old enclosing fence, the stones of which are very small, and which shows within the foundations of an oblong structure. The elevation of the group is also about 800 feet. About I mile to the north-east is another large group of at least a dozen small cairns, in which the stones are much exposed, and which measure in diameter about 16 feet and in elevation not above 2 feet. xxxix. S.E. 31 May 1912. Miscellaneous. 148. Entrenchment, Loch Urr.—The land around the southern end of Loch Urr, towards the west, is heathery moorland and rather wet; but to the westward of a southerly arm of the loch there rises, well above the water, a peninsula with a grassy surface. The neck of this peninsula is traversed by a broad hollow of rather natural appearance, damp and marshy in the bottom, above which, with a steepish scarp, rises the peninsula. This scarp is rough, irregular, and not altogether artificial looking. Above it, at the west end and extending a considerable distance east¬ ward, is a rampart-like mound of clay rising some 2 to 3 feet above the level in rear of it. Near the middle of its course, across the neck, it entirely disappears without any apparent reason, to come into evidence again towards the east side, where it is quite distinct, returning northwards for a short distance along the top of the bank facing the water, and terminating in a bulky flat-topped mound. xxxix. S.W. I June 1912. 149. Laird of Lag’s Tomb, near Farthingwell. —Sir Robert Grierson of Lag (1650-1736) is buried in the family burial-place in the old churchyard near Farthingwell. At the back of the enclosure are two armorial panels—the upper contains a shield bearing arms : A star or mullet between three square locks (for Grierson), and below the shield the initials I G ; the lower panel, the date 1656, with the initials S D on the upper line, and three locks, flanked by the initials W G and \V M, on the lower line. xli. S.W. 10 June 1912. Sites. The O.S. map indicates sites as under : 150. Monastery, Friars Carse. xli. S.W. 151. Church near Farthingwell. xli. S.W. DURISDEER. Ecclesiastical Structures. 152. Durisdeer Church.—The church and village of Durisdeer are situated at the Niths- dale end of the pass formerly called the Wall- path, and about 2 miles north-north-east of Carronbridge Railway Station. The present church was built in 1699 in the Renaissance style of the period but on a larger and more elaborate scale than was usual in a Scottish country church. It is cruciform on plan with a tower set at the west end, around which is grouped a range of semi-domestic buildings. These buildings are entered by a west doorway, from which the 59 DURISDEER.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [durisdeer. ground-floor apartments are approached to the right and left and the upper floor by a stone staircase on the left side. The ground floor of the tower is entered opposite the main doorway, and a stone wheel-staircase leads to the two next storeys. This encircling build¬ ing is said to have been built for a school. Throughout the whole structure there is the same treatment ; a moulded basement course, simple but effective mouldings to doors and windows, and a heavy moulded cornice. The belfry stage of the tow'er has a circular-headed window in each wall, divided into tw^o pointed lights by a plain mullion w'hich branches at the top. A sharply projecting moulded cornice encircles the parapet. The church is entered by doorways in the south and east w'alls and also from the vestry, which is on the north side of the tower at the west end. There is a flat ceiling, and the interior is devoid of ornament except in the north transept, which is entered by a low segmental arch filled with a fine WTOught-iron screen, obscured by the placing of the pulpit immediately in front of it so as to face the south transept, which is open to the church and has neither arch nor screen. The north transept was evidently built as a burial-place for the Queensberry family. It has a window in the w^est wall and a doorway in the east wall. Mural Monument. — The north wall is almost entirely covered by a large 18th-century mural monument of black and white marble, with effigies representing James, 2nd Duke of Queensberry, and Mary his wife. Wrought-iron Stand and Bracket. —In this transept are preserved a 17th-century hour-glass stand and a baptismal basin bracket of about the same date, both of wrought iron with scrolled supports. The former is cylindri¬ cal and 5 inches in diameter, with a total height of II inches, including the support. The latter is a basin-shaped cage io| inches in diameter at the top and with a total height of 2 feet ij inches, including the supporting rod, wLich is elbcwved in the middle. 153. Martyr’s Tomb. — On a recumbent table-stone on the south side of the church : HERE LYES DANIEL M'=-/MICHEL, MARTYR, SHOT / DEAD AT DALVEEN BY / SIR lOHN DALYEL, FOR HIS / ADHEREING TO THE / WORD OF GOD, CHRISTS / KINGLY GOVERNENT IN / HIS HOUSE : AND THE / COVENANTED WORK OF / REFORMA¬ TION AGAINST / TYRANNY PERJURY AND / PRE¬ LACY : 1685 • REV ; 12 • II AS DANIEL CAST WAS IN / LYONS-DEN FOR PRAYING UNTO GOD / AND NOT TO MEN / SO LYONS THUS CRUELY / DEVOURED ME / FOR BEAIRING WITNES TO / TRUTHS TESTIMONY / I REST IN PEACE TILL / JESUS REND THE CLOUD / AND JUDGE TWIXT ME AND / THOSE WHO SHED MY BLOOD 154. Grave-slab, Durisdeer Churchyard.— Within the churchyard to the west of the church is an upright slab 3 feet 7 inches high and 2 feet 6 inches broad, on which is the effigy of a mason clad in a long coat with cuffs reversed, with an apron round his waist, flowing hair and a round feathered cap on his head, bearing in his right hand a meU and in the left a chisel—the figure being set against a portico surmounted by a cherub’s head and bordered at the sides with a scroll. On the reverse side is an index hand and the inscription :— HEIR • LYES • ANDRO • PATRIK GEORG • AND • REBECKA • CHILD REN • TO • WILLIAME • LUKUP • M'^ • OF WORK • IN • DRUMLANGRIG • 1685. xiv. S.E. 5 June 1912. 155. Kirkbride Church.—The church of the ancient parish of Kirkbride, now merged in Durisdeer parish, is situated 5| miles south¬ east of Sanquhar, on high ground | mile to the east of the Sanquhar to Carronbridge high road. The building is orientated and oblong on plan, measuring 49 feet 9 inches by 19 feet 9 inches over the walls, with a sacristy 13 feet by 12 feet projecting from the north wall. The walls of the church are 2 feet 9 inches to 3 feet II inches thick. The east wall is crowned with a belfry and stands almost entire. From the ground-level to the under side of the belfry is 17 feet 6 inches and from the ground to the waU-head 6 feet 6 inches. In the north-east angle of the east wall is an aumbry i foot 3 inches wide and i foot 4 inches deep. The entrance doorway is in the south wall. It has a pointed arched head in two stones. In the reveal is a bar hole. In this south wall are 60 DURiSDEER.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [durisdeer. three small windows, only one of which is entire, having an obtusely pointed head slightly recessed from the wall face. The north and west walls are fragmentary. The sacristy opens into the church by a semi¬ circular arch 7 feet wide. This arch is well built and is unmoulded but has a chamfered edge. Kirkbride greatly resembles Little Dalton Church (No. 97), and was probably erected about the same time—in the early i6th century. It has two uncommon features, the belfry being on the east gable and the doorway near the east end. In the north-east angle of the church lies a fragment of a tombstone, on the border of which, in Gothic lettering 2| inches high, is inscribed :— ORATE PROPICIET The building is in a bad state of repair, and yearly grows more fragmentary. xiv. N.W. 23 May 1912. Castellated and Domestic Structures. 156. Drumlanrig Castle.—Drumlanrig Castle, 4 miles by road north-west of Thornhill, is a 17th-century mansion of Renaissance design. Pleasantly situated on high ground, amidst the finely wooded policies of the estate, it overlooks the windings of the River Nith on the east and commands an extensive prospect to the south. The main approach is from the north, and the vista through the straight half mile of lime-tree bordered avenue towards the massive pile of the castle in the distance is most impressive. On the south are the gardens which once ranked with the finest in Scotland. The castle, built of pinkish sandstone, is on the courtyard plan with a square tower at each of the four external angles and a round staircase turret at each corner of the court¬ yard. The plan (fig. 49) is very similar to that of Heriot’s Hospital, Edinburgh. The towers are carried one storey higher than the wings, which contain four storeys ; the central portion of the north fa9ade has only three storeys, an arrangement which graduates the sky-line and renders the design of this, the principal facade, still more effective. This facade is built of ashlar, and opens to the forecourt ; the kitchen premises and cellarage of the basement are screened by a quadri¬ partite-vaulted arcade, which projects from the face of the wing, providing a covered promenade at ground and supporting a terrace on the upper level. The principal entrance to the castle (fig. 50) is at this latter level, opening off the terrace, which is reached from the forecourt by a double staircase. The central portion of the main wing is carried forward to the extent of the arcade, forming a porch. The arcade consists of a group of three semi¬ circular archways set on either side of the staircase, which encloses a similar archway. Between the arches are fluted pilasters with moulded bases and caps supporting an archi¬ trave, frieze and cornice surmounted by a balustrade. Each arch springs from moulded imposts and has a moulded keystone of the not unusual console type. On the frieze a bestial head is centred over every pilaster, and between each are five triglyphs. On the inter¬ vening spaces are carved the Douglas heart surmounted by a coronet alternating with three cross-crosslets fitchy for Mar. The towers and the extremities of the wing terminate in cornices surmounted by balus¬ trades, which are received at the angles of the towers by circular turrets with ogival roofs. The main portion of the front has fluted pilasters on the piers between the windows and at the angles of the porch, terminating in moulded bases and in Corinthian capitals under an entablature, which is surmounted by a balustrade. The porch is elaborately treated but is in harmony with the plainer portions of the facade. Between the pilasters on its front is a blind arch, within which lie the upper- floor window and a doorway at the terrace level. Heavily carved festoons depend from the window-sill and enclose a cartouche. The cornice forms a semicircular pediment containing a heraldic achievement. The whole is surmounted by an octagonal clock turret with angle pilasters supporting an entablature. The turret roof is ogival, covered apparently with lead and adorned with a cresting. The clock has a square wooden dial dated 1686. The windows on this la9ade have raking cornices enclosing sculptured pediments ; the architraves are shouldered. The windows on the first floor have triglyphs on their friezes, Fig. 49.—Drumlanrig Castle (No. 156). 62 Ancient and Historical Monuments—Dump ies. To face p. 62. Fig. 50.—Drumlanrig Castle : Principal Entrance (No. 156). A Ancient and Historical Monuments — Dumfries. Fig. 52.— Sundials, Drumlanrig- Castle (No. 156). To face p. 63. DURISDEER.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [durisdeer. which are plane ; the friezes of the upper-floor windows are cushioned. The other elevations are less ornate ; the walls are built of irregularly-coursed rubble, with dressed backset margins to the voids and to the angles of the square towers, indicat¬ ing that these walls were roughcast or harled —a common practice in Scotland. A con¬ tinuous corbel-table, surmounted by a balus¬ trade, is carried along the wings and at a higher level along the towers. A basement course of slight projection returns along the walls. On the south, projecting stairs lead from the first-floor apartments to the garden (fig. 51). These exits are unpretentious yet altogether charming in their arrangement. In the centre there is a fine doorway flanked by Roman Doric columns, which support an entablature and segmental pediment broken by a central panel bearing a replica of the achievement on the north fagade. At either extremity of the wing are doors of similar detail to the windows. In front of these three openings are plats, whence the steps descend. The central plat is borne on a pavilion, and the others on porches, all of Renaissance design. The grace¬ ful wrought-iron balustrade is worthy of note. On the central plat are two sundials (fig. 52), the larger dated 1692, the smaller, which is incomplete, bearing the legend henry \vynne LONDINI FECIT. On the north fa9ade the porch is vaulted to support the super-structure, and the entrance is secured by a massive iron gate, commonly reported to have been removed from Tibbers Castle (No. 157), but there is no reason to suppose that this was the case. Similar gates are fitted to the basement doors. The entrance opens into a vestibule, which occupies the central portion of the wing and communicated directly with the quadrangle by an arcade similar to that screening the basement on the north. This apartment is now utilised as a lounge ; the arcade is glazed, and a fireplace is inserted in the west wall. The living rooms intercommunicate and are placed in the north and south wings. The bedrooms are contained within the lateral wings and the two square towers on the west. The provision of a corridor in each of these lateral wings shows a marked development in planning and is unusual at this period. The dining-room—a handsome apartment occupy¬ ing the central portion of the south wing—has a gracefully panelled plaster ceiling; oak panelling clothes the full height of the walls and is enriched, not altogether happily, by the later addition of carving by Grinling Gibbons or one of his school, which has been removed from another apartment. A doorway in the north wall, now built up, gave access to the quadrangle. East of the dining-room in the south wing is a spacious staircase, built of oak, superseding the wheel-stairs in the circidar towers. The drawing-room lies immediately over the dining¬ room and is treated in a similar manner ; its panelled walls are faced with old tapestry. A modern chapel built against the south wall greatly diminishes the beauty and import¬ ance of the quadrangle. The circular towers have doorways flanked by fluted pilasters bearing an entablature. The various storeys of these towers bear on the window lintels the dates of their erection. The north-east tower is dated 1679 • 1679 • 1689, the north¬ west tower 1684 ■ 1687 • 1687. The windows looking into the quadrangle have pediments, triangular and segmental, infilled with heraldic carving, and the tow'ers terminate in the ogival roofs covered with lead. The rain-water heads, embossed with the crowned heart of the Douglases winged, are noteworthy. Dimensions. —The building measures 120 feet across the north and south tow^ers b}^ 148 feet across those on east and w^est. It was erected by William, first Duke of Oueens- berry, in the last quarter of the 17th cen¬ tury. The architect is unknown ; the princi¬ pal facade has been ascribed to Sir William Bruce, from its resemblance to his work at Holyrood Palace, Edinburgh. Within Durisdeer churchyard (No. 154) is a headstone recording the death of the children of Williame Lukup M^ (Master) of Work(s) in Drumlanrig 1685. The principal floor plan and the north fagade are illustrated in Vitruvius Britannicus, vol. i. pis. 37 and 38, Colin Campbell, 1767. xxii. N.W. 8 June 1912. 157. Tibbers Castle. — The ruins of this castle are wdthin the policies of Drumlanrig Castle, at the northern extremity of a bold headland which rises abruptly from the level DURISDEER.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [durisdeer. haiigh on the south bank of the Barn Burn, near its confluence with the river Nith. The actual site is separated from the body of the headland by a partly artificial ravine and so Fig. 53.—Tibbers Castle (No. 157). forms an isolated mound, which in part also seems to be artificial (fig. 53). This mound was no doubt the “ mote ” of the historical references below, while the portion of the plateau to the south, measuring 300 by 200 feet, which is cut off by a conspicuous ditch and rampart, would have formed the “ bailey.” The summit of the mound is roughly oblong on plan (fig. 54), measuring 140 feet from north to south and 100 feet from east to west. It is enclosed by walls of enceinte, 7 to 9 feet thick, with circular towers at the salient angles. These walls are ruinous and less than 12 feet high. The area enclosed measures 125 by 86 feet. The entrance gateway is situated at the east end of the south curtain and has been flanked by circular towers. A considerable portion of the south-east angle tower remains. but the foundations of the corresponding tower on the west side are now in a very fragmentary state. The gateway has evidently been defended by (i) an outer drawbridge, (2) a portcullis, and (3) an iron gate. It seems probable that the approach from the south across the deep ravine would be effected originally by means of a timber staging which stopped short of the gateway, leaving a space to be spanned by the lowered draw¬ bridge. The sill of the gateway and the level of the interior courtyard are fully 4 feet above the ground-level outside the main walls. Abutting against the exterior of the flanking towers, and centering with the gateway, are the foundations of two parallel walls some 8 feet 10 inches apart and 3 feet g inches in thickness, projecting southwards for a dis¬ tance of about ig feet, which are evidently Fig. 54.—Tibbers Castle (No. 157). of later date than the double towers. On the level ground to the south of the ravine, and facing the entrance to the castle, there is a low mound of earth, measuring some 25 feet DURiSDEER.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [durisdeer. by 15 feet, which seems to suggest the position of the gangway on that side. Within the ward are remains of the internal constructions ; one, of rectangular form, at the south-west angle, measures some 37 feet 6 inches by 21 feet 6 inches, within walls averaging 4 feet in thickness ; and other two buildings at the north-west angle, 53 feet by 25 feet 3 inches and 27 feet 4 inches by 12 feet 6 inches respectively, within walls varying from 2 feet 6 inches to 4 feet 5 inches in thickness, and only a few feet in height. The north-east angle tower is now in a very ruinous state, but there seems to have been a small postern adjoining it in the east wall, giving access to the ward by a zig-zag passage. The east curtain is only 4 feet in thickness, and parallel to it there is an inner wall, now very much ruined, enclosing a passage 3 feet 6 inches in width, which appears to have con¬ tinued southwards in the direction of the south-eastern angle tower. At the north¬ eastern angle of the interior courtyard there is a fine example of a built well, with a present depth of some 46 feet. Considerable portions of the west angle towers and the enclosing walls have fallen outwards, forming a mass of debris at the top of the sloping bank. From its plan the building apparently dates from the end of the 13th century. In August 1298 Sir Richard Siward, a Scot serving Edward I., was just beginning the erection of “ his house at Tibbers.” ^ This “ house ” was erected upon what is on record afterwards as “ le Mote de Tibris ” and the “ Castcll mote of Tibberis,” ^ descrip¬ tions which point to an earlier type of castle. On 3rd September of the same year, after the battle of Falkirk, Edward I. was at Tibbers. In June 1300 the place had a garrison of thirteen “ valletz d’armes ” and eight archers.® Two years later is record of a royal grant to Siward of ;^ioo for the repair of “ his castle of Tybres.” ^ King Edward can be judged from the context to have appre¬ ciated its strategical value.® Siward was “ Sheriff of the county of Dumfries and Con¬ stable of the castle” when it was-captured, and he in it, by Robert Bruce on his way northwards after the murder of Comyn. It was held for Bruce by John de Seton, who in his turn was captured with it by the English, upon which he was drawn and hanged.® There¬ after it was garrisoned for the English, from 22nd February to 25th September 1306, by Thomas Bell, with eleven squires and fifty foot archers.^ In its later history it passes to the Dunbar Earls of March, thence to the Maitlands of Auchencastle, and from them to the Douglases of Drumlanrig.® A charter to Robert Maitland in 1489 grants ” locum castri et montem nuncup. le Mote de Tibbris” ; and, in a confirmation of 1592 to Sir James Douglas of Drumlanrig, the lands and barony are associated ” cum castro et lie castell-moie.” ® 1 Bain’s Calendar of Documents, etc., vol. ii. No. 1005 ; ® Hist. MSS. Commission XV., App., part viii. pp. 18, 19 ; ® Bain’s Calendar of Documents, etc., vol. ii. No. 1141 ; Ibid., 1307; ® Cf. Introd., p. xxxii.; ® Calendar, ii. p. 486 ; ’ Calendar, iv. pp. 390-391 ; ® Hist. MSS. (as cited), pp. 32-36; ® Reg. Mag. Sig., s.d.. Nos. 1885, 2034. xxii. S.W. 9 June 1912. Defensive Constructions. 158. Fort, Drumlanrig.—This fort is situated at the north end of a large grassy park, about ij mile north-north-east of Drumlanrig Castle, on the east side of the Nith, occupying the edge of the high bank some 60 feet above the river. On its north flank the glen of a small rivulet, deepening riverwards, affords a strong natural defence. The fort appears to have been circular with a diameter of 74 feet, but all traces of a rampart towards the river and glen have disappeared. Around the landward side from bank to bank there has been formed a stony rampart, now almost obliterated on the east but well-preserved towards the south, with a trench to the outside, now most observable where it dips towards the river. The greatest height of the rampart above the interior is 3 feet ; its breadth at base 14 feet. To the exterior it has a height of about 4 feet at its highest point farthest from the river, but gradually increases as the ground in front falls away towards the bank. xxii. N.W. 10 June 1912. 159. Fort, Kirkleys, Langknowe.—This fort has been situated on the summit of a grassy ridge some 300 yards to the west of the cottage of Langknowe and adjacent to the tunnel DURISDEER.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [durisdeer. of the Glasgow and South-Western Railway. The site has been much under cultivation, and the lines of the fort, except at the south end, are no longer clearly expressed. As shown on the O.S. 25-inch scale, the plan of the fort has been quadrilateral with rounded angles, but with a considerable curvature on the south front. The lines of this plan are still trace¬ able, though the hollow which marks the line of the trench on the north is not very distinct. On the west flank there is a very slightly defined terrace, while on the east a hollow, along the length of which rushes are sprouting, clearly indicates a ditch on that side. From north to south the interior diameter is about 150 feet. The defence at the north end con¬ sists of a trench some 33 feet in width and 4 feet in depth. The elevation above sea- level is 644 feet. xiv. S.W. 10 June 1912. 160. Mote Hill, Ballaggan.—This mote hill is situated in cultivated land to the east of Ballaggan farm-house, from which it is distant some 200 yards. It is a flat-topped artificial hillock erected on the end of a slight natural terrace, which has probably here projected to a point, above which it rises about 8 feet, while at the opposite side its elevation is 18 feet. Whatever defences have existed to the southward in the direction of the terrace have completely disappeared ; but around the re¬ mainder of the mound, at 10 feet below the summit, is a trench or terrace, now for the most part the latter, with a width of from 9 to 10 feet. The summit measures some 27 feet in diameter ; its surface has been con¬ siderably interfered with, probably by weeds and rubbish being deposited on it. Ballaggan is a lairdship in the sixteenth century [Scottish Papers, ii. p. 627), apparently belonging to Hunters [Bucc. MSS., p. 51). xxii. N.W. II June 1912. 161. Fort, Cleuchhead.—On the edge of a steep wooded bank falling sharply to the bottom of the glen which marks the division between the parishesof Durisdeer and Penpont, and about J mile south-east of the cottages at Cleuchhead, is a small fort of an irregular circular form. From the north the ground declines towards the fort ; to the south is the glen, and on the east a small ravine, now to some extent choked with rubbish, gradually deepening as it debouches on the glen, and forming the channel of a small stream. Along the side of this ravine, even where it is shallow at its commencement, there no longer remains any trace of defences ; but from its edge at the north-east corner of the fort a broad rampart of earth, with a ditch to the outside, curves round in a somewhat distended semi¬ circle to the face of the bank overlooking the glen on the south-west. The interior measures about 100 feet by 76 feet. On the north the trench has a width of about 38 feet and a depth of about 7 feet, diminishing on the west to 32 feet and 5 feet respectively, the counter¬ scarp gradually becoming less pronounced as the trench opens on the bank of the glen. From the interior on the north the rampart rises to a height of about 6 feet. The sur¬ rounding land on the west seems to be full of springs, and a strong run of water drains into the trench from that direction. xxi. N.E. II June 1912. 162. Earthwork, Durisdeer.—In low-lying land, much of which in former times was probably under water, just to the north of Lochside, and about | mile west of Durisdeer, there rises an earthen mound, foursided and approximately a square, the actual measure¬ ment at base being 97 feet on the north, 114 feet on the east, 103 feet on the south, and 102 feet on the west. The surface of the mound has long been under cultivation, and the edges of the scarps have been somewhat broken down; but where best preserved these rise to a height of from 6 to 9 feet above the floor of the enclosing ditch. The ditch shows a width of about 60 feet across the top, but has been much destroyed by ploughing, and, except along the south and to a less extent on the east side, the counterscarp has almost disappeared. On the north and east the counterscarp appears to have been banked up to some extent. The area of the mound above the ditch is not now level, but rises towards the north-east : the gradient may to some extent be due to the action of the plough. Where a section of the mound has been exposed by the falling away of soil on the south side, a number of large water-worn 66 DURiSDEER.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. lDURISDEER. boulders are exposed, laid to some extent in courses. xiv. S.E. 21 June 1912. 163. Fort, Kirk Burn, Durisdeer.—About i mile to the north-east of Durisdeer Church, up the glen of the Kirk Burn, is situated a small, oblong, rectangular, entrenched earthwork (fig. 55). The glen is narrow, and its sides steep ; and from Penbane Hill on its eastern flank a spur projects into it, on the crest of which the fort is situated, looking out beyond it to Tynron Doon (No. 609), some 9 miles away. The ground falls away sharply on both flanks and also to the south-west, the main axis of the fort being from north-east by north to south-west by west. The enceinte measures some 82 feet by 72 feet, and has been surrounded by a parapet mound still some 3 feet in height on the flanks, some 6 feet at either side of the entrance, which has been through the centre of the north¬ east end, and still mas¬ sive at the opposite extremity. The trench, which defends the con¬ struction on all sides, measures some 15 feet in width, increasing to 25 feet on either side of the entrance, where, be¬ tween it and the base of the rampart, a slight berm intervenes. The scarp of the trench, which is steep, has a height above the floor of from 12 to 15 feet, while the counterscarp has an elevation of 4 or 5 feet. The entrance which, as already stated, is in the centre of the north¬ east end, has a width of 5 feet where it passes through the rampart and 14 feet as it crosses the trench. Some 24 feet in front of the passage over the trench, covering the entrance but extending chiefly to the east, is an outer trench, visible for a length of some 36 feet, with a width of 17 feet. The angles of the fort are rounded off, and the trench at the south-west end is for some distance cut through rock. An ancient track or bridle path known as the Well Path leads northward along the east side of the glen through the hills into the Dalveen Pass ; and a section of the Deil’s Dyke, which curves round the base of the Black Hill, thence up the right bank of the Kirk Burn, trending straight for this fort, terminates at the edge of the enclosed bank about I mile distant from it. xiv. S.E. (“ Earthwork ”). 21 June 1912. Miscellaneous. 164. Castle Doorway, Coshogle.—In a white¬ washed cottage by the roadside on the way to Coshogle farm steading, the doorway, with segmental arch and moulded jambs, is evi¬ dently a relic of the old castle of the Douglases of Coshogle. The jambs have quirked rose mouldings, which are continued round the head, and above the latter there has appar¬ ently been a hood-moulding of some sort terminating in stops which are now much decayed and all that remains. The jamb mouldings are received on splayed stops at base. 165. Heraldic Stone, Coshogle.—Another relic of the old castle is a stone panel bearing arms, inserted into the gable of a cottage behind the farm steading. It shows two shields bearing arms thus : Dexter, Quarterly, 1st and 4th, A bend between six cross-crosslets fitchy (for Mar) ; 2nd and 3rd, A heart, on a chief a cushion between two stars (for Douglas). Sinister, A saltire and chief, the latter charged with three cushions (Johnstone). Above are the initials R. D. N I. for Robert Douglas and Nicholas Johnstone his wife ; and below, the date 1576. Robert Douglas of “ Coschogill ” is twice on record in 1581.* He was related to the Drumlanrig family, as the arms show. Both achievements have been heavily coated with whitewash. ^ Buccleuch MSS., pp. 27, 29. 166. Heraldic Stone, Dalveen House.—On the west gable of Dalveen House is a sunk square panel within a moulded border bear¬ ing arms: ist and 4th, A bend between si.x cross-crosslets fitchy (Mar); 2nd and 3rd, A heart, on a chief a cushion between two stars (Douglas). On the north gable of a back wing is a slab bearing date 1622 and initials 1 H N D, the three latter being com- 67 Fig. 55.—Fort, Kirk Burn (No. 163). DURISDEER.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [eskdalemuir. billed in a monogram. The line of Douglas of Dalveen was an offshoot from the Drumlan- rig house. Hugh Douglas of Dalvene is on record in 1559.^ ^ Biiccleiich MSS., p. 24 ; cf. also Introd., p. xxix. Sites. 167. Enoch Castle.—The site of this castle is a four-sided plateau, strongly defended on three sides by the steep natural banks over¬ looking the glen of the Carron Burn and a small feeder which flows by two sides of the site. On the fourth side a hollow way, no doubt to some extent cut through the rock, leads down to the Carron. Nothing beyond the lines of the foundations beneath the turf remains \’isible ; a circular depression possibly indi¬ cates the well. For historical references see Introd., p. xxvii. xxii. N.E. 8 May 1913. The O.S. maps also indicate sites as under :— 168. Chapel near Moorcleugh. xiv. S.E. 169. Coshogle Castle, Coshogle. xiv. S.W. 170. Durisdeer Castle, about 500 yards east of Castlehill, Durisdeer. xiv. S.E. ESKDALEMUIR. Defensive Constructions. 171. Fort, Over Cassock.—This fort is situ¬ ated on along pear-shaped promontory running southward from the front of Over Cassock farm-house, and contained between the wooded glen of the Barr Burn on the west and a deep natural ravine, bearing the name of “ The Lake ” on the east, which opens on to the bed of the burn at the lower end of the promontory. The site, by nature strongly defended, has been formed into a fort with two enclosures, an upper and larger one, sub-oval in form, wflth a slight concavity on one side, measuring some 260 feet by 220 feet, and a smaller one, somewhat harp-shaped, subtending it, measur¬ ing 115 feet by 80 feet. The upper area has been partially encircled by a bold rampart of earth and stone, except on the south-east arc overlooking the ravine, where the grassy mound probably conceals a wall, as here only has it been set back from the edge, so as to leave between a well-defined berm 8 feet in width. It is doubtful, also, if the rampart extended above the high precipitous bank of the burn. The mutual rampart on the south¬ east between the two enclosures has a height of about 5 feet on the upper side, where also it is covered by a trench some 26 feet wide ; while on the lower side it has a height of 4 feet, and slopes away gradually into the falling level of the enclosure. The mound elsewhere around this enclosure is slight. The trench on the upper side of the mutual rampart, and the convexity of the latter towards the main en¬ closure, are probably due to the desire to render the smaller enceinte defensible in the event of the upper being captured. Across the north and upper end of the promontory, towards its constricted neck between the ravine and the glen, there have been drawn two ramparts beyond the inner encircling mound, with a space varying in breadth from 30 to 36 feet between the inner and intermediate, and a trench 30 feet wide and 5 feet deep between the latter and the outer mound. The inter¬ mediate rampart has a breadth at base of 28 feet, and a height of 7 feet. There has been an entrance about 8 feet wide into the upper enclosure from the ravine near the centre of the east arc at a low level, above which the rampart rises on either side to a height of from 6 to 7 feet. , The ground immediately before this entrance is wet, and there appears to be a spring in it. Beyond the lower enclosure the promontory projects with considerable declination to the side of the burn. To the northward there stretches into the moorland, for a distance of a hundred yards or so, a deep hollow some 36 feet wide and from 7 to 8 feet deep, which terminates abruptly at its upper end. xviii. S.W. 8 July 1912. 172. Roman Camp, Raeburnfoot.—The camp at Raeburnfoot (fig. 56) lies in a remote pastoral valley some 14 miles from the town of Langholm, with grassy, undulating hills swell¬ ing up on all sides. Through the valley the road winds onward to Ettrick. Its situation is a plateau, which lies in the angle formed by the junction of the Rae Burn with the White Esk ; the ground has a gradual inclination from south ESKDALEMUIR.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [eskdalemuir. and east, rises sharply to a height of some 30 feet on the west from the low meadow-land that reaches to the margin of the Esk, and is cut off by a hollow on the north from the rolling moorland beyond. Occupying all the higher ground of the plateau, and projecting beyond the base of the plateau on to the low-lying ground on the east, is an oblong entrenchment with its longest axis north-north-west to south-south-east. On the west it rests on the edge of the steep bank with now no apparent artificial defence ; but enclosing the enceinte on the three other sides is a broad Fig. 56.—Roman Camp, Raeburnfoot (No. 172). rampart, now considerably levelled by plough¬ ing but some 30 feet in breadth at base, with a clearly defined trench to the outside. The trench and rampart are best preserved on the north face, where the former meas¬ ures some 30 feet across from crest to crest, 8 feet in depth below the scarp and 2 feet below the counterscarp, while the latter has a breadth at base of about 30 feet. The area enclosed measures through the centre some 540 feet by 360 feet and contains about 4I acres. Somewliat west of the centre at the north and south ends, there are entrances through tlie vallum ; that on the south meas¬ uring 12 feet wide, that on the north, though the original dimension has been interfered with by the intrusion of a turf dyke, apparently similar. With its western flank resting on the edge of the steep bank a smaller enclosure has been contained within the larger area at 86 feet within the rampart at the north end and 106 feet at the south, apparently surrounded, except on the west, by a rampart and two trenches, the rampart being preserved for a short distance only on the north and south sides, where it abuts on the western bank, and the trenches recognisable by the darker colour of the vegetation as well as by the slight de¬ pressions on the surface. The rampart has a breadth at base of some 38 feet. These lines are traceable from the west flank east¬ wards for a distance of 212 feet on the south, and, owing to the irregular line of the base from which they start, rather more on the north, and are lost beyond the rounded angles, where they pass along the face of the plateau above the low-lying ground on the east. The area thus enclosed has measured approxi¬ mately 230 by 210 feet. Near the centre of the north and south ends are breaks in the vallum which represents entrances ; that on the north side 38 feet wide, and that on the south 36 feet. The trenches are made to form a con¬ nection on either side, though on the upper or west side of the south entrance this arrange¬ ment is not now very distinct. At the en¬ trance to the larger enceinte from the south the trench coming from the west takes a curve outwards for some 12 feet across the entrance, as evidenced by the darker colour of the grass and the growth of rushes, thus forming the outer half of a clavicula. The excavations conducted by the Dumiries and Galloway Antiquarian Society in Novem¬ ber i8g8 provided the material upon which the present plan is based. The outer rampart, about 30 feet wide at base, was found to be composed of earth ; the rampart of the inner enclosure, about 35 feet broad at base, of layers of earth and clay. At the south gateway the road had a gravel surface, and where it entered and where it passed out on the north from the inner enclosure gave signs of cobble surface and larger stones, vEich possibly were kerbs. This was the only roadway found. Accumulations of stones, some resembling walling, were present at certain parts, but nothing that could be definitely assigned to a building. {Transactions, Dumfries and Galloway Antiq. Soc., 1897-98, pp. 17-27.) A fort identical in plan, though slightly 69 ESKDALEMi-iR.j HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [eskdalemuir. smaller in area, is found at Castleshaw on the west side of the Pennines, half-waj^ between York and Chester, at an elevation of about 900 feet above sea-level. But while the area of the inner enclosure at Raeburnfoot is about one-third of the whole enclosed space, that of the same enclosure at Castleshaw is only one-sixth of the whole. Most of the pottery found at Castleshaw seems to date from the first centur}'. The inner enclosure contained a bath. There are in possession of the Dumfries and Galloway Natural History and Antiquarian Society at Dumfries, and preserved in the Ewart Public Library there, the remains of at least two large dolia or amphorce, of un¬ doubted Roman origin, found in the excava¬ tions at Raeburnfoot. xx\d. N.W. (“ Earthwork ”). g July 1912. 173. Fort, Bessie’s Hill.—This enclosure is situated on the north-east end of an eminence which rises from the eastern flank of the Castle Hill and lies at an elevation of some 850 feet above sea-level. The area it occupies slopes considerably towards the east and is uneven on the surface, being to some extent hollowed out by art. On plan the construction is pear-shaped, with its longest axis north and south, and measures interiorly some 200 feet by 175 feet. It is surrounded by a bold rampart of earth and stone wfith an average breadth of 51 feet at base and 22 feet at the entrance, having a trench to the outside, which, except on the north, measures some 15 feet in width and 6 feet in depth below the crest of the scarp. The principal entrance, 7 feet in \Hdth, has been from the lowest point, and opens into a hollow, which is carried inwards between two long parallel mounds seemingly formed by excavating the surface around them. There has also been an entrance from the northward above a steep downward slope, between which and the rampart lies a natural terrace, some 15 feet wide, occupying the position of the trench. The rampart is stony, and the trench has in places been cut through rock. xxvi. S.W. 9 July 1912. 174. Fort, Bessie’s Hill.—Some 200 yards to the south-east of No. 173 is another fort. It is formed on an eminence at an elevation of 800 feet above sea-level overlooking the valley of the White Esk, about 7} mile west of the farm of Holm Mill. Its.main axis lies north-east and south-west. Towards the north is a hollow cutting it off from the higher slope of the hill beyond ; and on the south, with an abrupt, and in places precipitous rocky face, the ground declines rapidly to the base of the hill. On plan the fort is semi-oval with the edge of the steep southern escarpment forming the chord. So strongly is it protected by nature on the south that artificial earthwork has been dispensed with ; but around the periphery towards the lower ground, below a bold scarp from the summit level, there runs a double trench with an intervening rampart 22 feet broad at base. The trenches from crest to crest measure, the inner 33 feet across and the outer 16 feet, while the former has a depth of 12 feet below the crest of the scarp and 3 feet 6 inches below that of the counterscarp, and the latter similar measurements of 8 feet and 3 feet 6 inches respectively. The rampart has been of earth and stone, and there has been considerable rock cutting in the formation of the trenches. The entrance has been on the summit level at the north-east angle, some 8 to 9 feet wide, passing between the end of the defences and the steep escarpment on the east, the road towards it being traceable for some distance up the hillside. It opens on a slight excavated hollow in the interior. There is no well visible in the area of the interior, but in the inner trench there is a considerable clump of rushes at one point, which may indicate the presence of water. xxvi. S.W. 9 July 1912. 175. “Fort,” Over Rig.—Somewhat less than I mile north-north-west of Castle O’er House, at the bottom of a natural hollow forming, as it were, a hemicycle, over the upper side of which passes the road to Eskdalemuir, is a semicircular entrenchment resting on the edge of the right bank of the White Esk, which flows by some 20 feet below. Tire enceinte is a semicircular plat of very low elevation, overlooked from all points except the direction of the river, with a chord of some 200 feet and a radius of 85 feet or thereby, with no indication of a parapet mound. It 70 ESKDALEMUiR.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [eskdalemuir. is surrounded by an inner trench some 3 to 4 feet in depth and 13 feet in width, a con¬ centric rampart some 5 feet in height and 18 feet in breadth at base, with an outer trench beyond 15 feet broad and 3 feet deep, having in places a slight mound on the counter¬ scarp. Towards the north for some 60 feet the outer trench appears to have been filled up. At an elevation somewhat higher, but under the 600 feet level, there passes around the amphitheatre a terrace some 10 or 12 feet broad, changing to a trench where it makes a return at either side towards the river. Part of the central area has been broken away at the edge of the bank by and at deepest 6 feet in depth. The rampart has been surmounted with a wall some 4 feet in thickness, the base of which remains visible on the south side of the entrance. The entrance has been from the south-west some 4 feet wide, and is clearly defined ; the road¬ way leading over the trench and through the rampart. The interior has been considerably hollowed out, the centre being as much as 10 feet below the level of the crest of the rampart on the west, which at this point is almost of equal elevation to the ground out¬ side. Against the rampart on the south-east side, and partly under the sheepfold, are the foundations of a small oblong structure, the the erosion of the river, and in the section is exposed in places what appears to be a clay floor immixed with charcoal and frag¬ ments of burnt bone; the number of flat water-worn pebbles displaced and in situ at the top of the section gives a distinct sugges¬ tion of paving. XXXV. N.W. 5 July 1912. 176. Fort, Bank Head Hill.—This fort is situ¬ ated on the west side of the broad table-land which forms the summit of Bank Head Hill and just where the ground begins to slope down¬ wards to the valley of the White Esk. From Castle O’er House, across the river, the con¬ struction is distant about \ mile to the north¬ east. In form the enceinte is circular, measur¬ ing some 165 feet in diameter, and is sur¬ rounded by a broad rampart of earth and stone, 20 feet broad at base, increased to 25 feet on either side of the entrance, with a trench in front measuring from 20 to 34 feet in width back wall of which has been recessed into the rampart significantly broadened behind it. A low^ mound, apparently of earth, some 5 feet broad, with a shallow trench some 8 feet wide on the north side, meets the outer edge of the enclosure on the north, and leaves it on the south-east, passing southward along the crest of the moorland. Under the name of the “ Deil’s Jingle” this bank is traceable, as shown on the O.S. map, for about f mile to the north of the enclosure and J mile to the south of it. XXXV. N.W. 12 July igi 2 . 177. Fort, Castle O’er.*—The fort of Castle O’er (figs. 57, 58) occupies a prominent rock * "O’er ’’ is for " Over,” and Castle O’er is the equivalent of Overby or Overbie, an alternative name =" Overtown,” which thus links up the position with jMiddlebie and Netherby. At the latter two there are Roman camps (for iMiddlebie, see Birrens, No. 462). It is thus possible that the original Overbie or Castle O’er was really Raeburnfoot Camp, No. 172. (Cf. Stat. Acct., xii. p. 614; and New Stat. Acci., iv. pp. 401-2.) ESKDALEMUIR.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [eskdalemuir. eminence which mounts to a height of some SS4 feet above sea-level and 80 feet above the lowest point of the moorland at its base. It is situated about | mile north-west by west of the mansion-house of Castle O’er, on the crest of a long ridge forming the watershed between the Black Burn on the west and the White Esk on the east. The main axis of the hill, around which lie the entrenchments of the fort, is from east-north-east to west-south-west ; the hill rises by a steep grassy slope from north and south, has a bolder inclination on the west, and presents an abrupt and rocky aspect towards of the projecting rampart which forms the south side of the outer ward, a trench (fig. 59), for the greater part of its length hewn through rock with a mound to the outside formed from the upcast, is carried along the south-east flank at the base of the steep scarp to the east- north-east end, where, turning sharply up hill, it terminates adjacent to the commencement of the second defences that pass along the west side. There are two entrances into the central area, one from east-north-east and the other at the opposite extremity. At the base of the eminence, from south to east, there Fig. 58.—Fort, Castle O’er (No. 177). 300 FEET the east and south-east. The main enclosure on the summit is in form rather more than a half oval, with its south-east side formed of the almost straight edge of the steepest slope, and it crosses the summit some 100 feet from its west-south-west end. Around it there ap¬ pears to have been erected a wall of dry-stone masonry, now entirely dilapidated and for the most part overgrown. Outside this a second line of defence, consisting of a second wall or a stony mound with a trench beyond and a mound on the counterscarp, runs from the centre of the east-north-east end by the west side round to the west-south-west end, where it is met by an arm projecting from the line of the inner defences, thus containing against the west-south-west end of the main enclosure an outer court or bailey. From the outer end is a large enclosure formed by a trench some 23 feet wide and from 5 to 6'feet deep, with a mound to the outside, which runs south from a point adjacent to the west-south-west entrance for a distance of 180 feet, thence with a sharp return continues approximately parallel with the inner defences till it meets the steep upward slope opposite the east-north¬ east entrance, thence, with a break, extends for a short distance along the higher ground and takes a right-angled return towards the inner defences. From westward of the west-south¬ west entrance a spur of rock runs outward, from the base of which, on the south, a trench some 17 feet wide and 4 to 5 feet deep, with a mound on the counterscarp, takes a bold sweep southward past the entrance to the south angle of the large enclosure along 72 A>icie>il and Historical Monuments—Dumfries. Fig. 59. — Fort. Castle ( 3 'er : Trench (No. 177). Fig. 60. — Fort, Castle O’er: Ramparts No. 177}. Fig. 61.—Fort, Castle O'er: Entrance No. 177'. ESKDALEMUiR.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [eskdalemuir. the base of the hillock, stopping short of the outer mound, so as to leave a passage to the area which it encloses. A roadway is trace¬ able approaching the west-south-west en¬ trance from the north, up a slight hollow between the rocky spur and the defences. The entrance to which it leads is about 5 feet in width. Within it the outer ward has pro¬ bably been hollowed by excavation. As the roadway passes through the inner wall it enters an excavated level area faced by an elevated oval plateau some 4 feet above it on the right, while on the left there remains an arc of an oval enclosure, indicating a diameter of 59 feet and measuring some 30 feet along the curve, with a breadth of 2 feet 4 inches. This segment has been formed of living rock left in relief on excavation. Between the plateau and the enclosure the roadway evidently passed with a width of 6 feet, rising to the higher level beyond. The approach at the east-north-east extremity, after passing the outer trench, is carried up a steep slope through the inner defences and is faced directly in the interior with an unexcavated mass of rock, with a slight oval depression on the top, around which the roadway seems to have been taken at a lower level to right and left. The width of this entrance is not accurately obtainable but appears to have been about 8 feet. Some 65 feet in from it is a well-defined circular enclosure, with a diameter over all of 35 feet. Some 20 feet beyond it a stony segment equal to about f of a circle is visible, having a diameter of 22 feet, an elevation of a few inches above the general level and a breadth of 2 feet ; while just beyond, and at the highest point of the interior, there appears to have been a third, excavated to a depth of about 3 feet on the upper side and measuring some 28 feet in diameter. On the east there is a break of some 20 feet in the continuity of the ditch which contains the large enclosure at the base of the hillock, and immediately in rear of it is a well. As the scarp of the mound to the outside of the trench is carried across the break it has probably not been an entrance, but the ground has possibly been left unexcavated on account of the well. The entrance to this enclosure has evidently been to the south of the west-south-west entrance, where an un¬ excavated space is left between the base of the inner rampart and the commencement of the trench which surrounds it. From the north¬ east, where the outer defences turn on the crest of the ridge, a broad trench some 3 feet in depth passes down the hillside to the head of a ravine, into which flows a rivulet. XXXV. N.W. 5 July 1912. 178. Fort, The Knowe.—This fort (fig. 62) is situated on the slope of the hill facing east¬ wards, just above the south entrance to Castle SECTION AB 100 50 O 100 200 [I H I IIIII I- \ -1 ^ Fig. 62.—Fort, The Knowe (No. 178). O’er House. It is an oval enclosure, with its longest axis north - north - east and south- south-west, measuring in the interior some 150 feet by 130 feet, and with a marked gradient .to the east. Horse-shoe-wise, on the flanks and the upper or western side, rises a bold rampart which, on the interior, imperceptibly merges into the general level. Outside it lies a very deep V-shaped trench, some 36 feet wide, ii feet deep below the crest of the scarp, where deepest, and 6 feet from that of the counterscarp. At 10 feet back from the crest of the counterscarp rises an outer mound, 16 feet broad at base and 4 to 5 feet high, covered by a shallow outer 73 ESKDALEMUiR.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [eskdalemuir. trench beyond with a width of i6 feet. Pass¬ ing down the slope on the south the outer defences entirely cease, leaving the inner ram¬ part alone remaining but much diminished in height ; while on the north side the outer defences at the lower end converge, and the inner trench is carried forward to the face of the lower slope. The entrance is from the lowest level, and there are the usual indica¬ tions of excavation in the interior. Near the centre is a somewhat level plat some 25 feet in diameter. The ramparts are formed of earth and the splintered rock excavated from the trenches. XXXV. N.W. 9 July 1912. Enclosures and Stone Rings. 179. Enclosure, Fingland. — At the very edge of the haugh-land that lies between the main road from Eakdalemuir to Ettrick and the White Esk, about 150 yards to the south-south¬ east of Fingland and the same distance back from the river, are the remains of an entrenched earthwork. The construction has been much interfered with by the formation of a road along each side of it, and, except for a small segment on the north arc, its lines are not very noticeable. The segment consists of a double rampart with a trench between, some 26 feet wide and 5 to 6 feet deep. The ram¬ parts measure at base 34 feet and 22 feet respectively. The plan, as indicated by the remains, has probably been oval. The interior is very uneven and rather damp. xviii. S.W. (" Fort ”). 8 July 1912. 180. Enclosures, Rough Castle Hill, Garwald, —On the top of Rough Castle Hill, some 500 yards to the west of Garwald farm-house, and at an altitude of 993 feet over sea-level, are two contiguous enclosures surrounded by slight earth and stone banks. They are situ¬ ated the one above and the other at the base of an outcrop of rock slightly quarried on the face at the back of the lower enclosure, and rise to a height of about 9 feet. The lower enclosure is oval in form, measuring some 160 feet by 120 feet, and has been hollowed out to some extent, so that the level of the interior is lower than that of the ground outside. The entrance is from the south-east, and, as usual, opens on the lowest part of the enclosure. In the south-west angle is a hollow with a stony bank around it, which seems to be the site of a hut. The second enclosure, at the higher level, appears to have been more oblong in form, but the bank around it is slight and difficult to determine. There is no trench, and the construction has little of the appearance of a fort. xxvi. N.W. (“ Fort ”). 8 July 1912. 181. Enclosure, Longknowe, Garwald.—This enclosure is situated at an elevation of over 1000 feet above sea-level on the haunch of the hill overlooking the valley of the Monk- ingshaw Burn more than 100 feet below. It is oval in form, measuring 180 feet by 140 feet, and has been surrounded by a stony bank some 7 feet broad at base, and now reduced to a height of about 2 feet. The interior is very uneven, and on the north the rock is near the surface. There are two entrances ; one at the south end, about 8 feet wide, opening on the lowest level, and the other on the east, also opening on a hollow but faced with higher ground in front of it. The site is a plateau at the edge of a steep slope, and quite unnoticeable from below. xxvi. N.W. (“ Fort ’'). 8 July 1912. 182. Enclosure, Johnstone House.—On the south-east slope of the hillside, some 300 yards west by north of Johnstone House, partly on a plateau and partly on the steep ground above it is an oval enclosure having its longest axis north-east and south-west and measuring some 280 feet by 170 feet. It is surrounded by a broad rampart of earth and stone. Com¬ mencing at the lower edge of the plateau on the north-east, curving round the north and gradually mounting the steep slope of the hill, it passes along the north-west side some 30 feet above the lowest level of the interior and returns down the slope at the south-west. It contains on the upper or north-west side a natural terrace, which lies some 18 feet above the lower part of the enclosure. There has been an entrance into the lowest side of the area from the south, near its west end,, and seemingly another from the east. From the right side of the former a wall passes inward across the interior, stopping some 12 feet short of the steep scarp from the terrace 74 ESKDALEMUiR.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [eskdalemuir. on the upper side. In continuation of the edge of the scarp from the terrace an inner bank curves round towards the east and meets the outer bank on the south-east, which, however, appears to have been carried on¬ wards, with a sharp curve to the south. The south entrance was reached by a covered way from the north-east, which at its commence¬ ment is concealed by a sharp right-angled return. In the interior there are indications of stonework and artificial hollowing. On the upper side of the enceinte, outside the sur¬ rounding bank, is a trench some i6 feet wide, from which the material for the mound has evidently been obtained. The position, not observable from below, is more sheltered than defensive. xxvi. N.W. (“ Fort ”). 8 July 1912. 183. Enclosure, Moodlaw. — A circular en¬ closure measuring about 100 feet in diameter is situated on the left bank of the Moodlaw Burn, some 20 feet above the level of the stream and about J mile to the south of Moodlaw farm, being surrounded by a bank of earth and stone some 10 feet thick at base and from 2 to 3 feet high. There has been an entrance at the lower side from the direction of the burn. The interior, which is rather wet, has been hollowed out to some extent, so that on the upper or eastern side the ground level to the outside of the bank is about 3 feet higher than that within it. xxvi. N.W. (“ Fort ”). 8 July 1912. 184. Enclosure, Mid Raeburn.—This enclos- lu'e is situated on a point of land at the junction of a small stream with the Rae Burn, about 11 miles above the confluence of the latter with the White Esk. It lies on falling ground, facing straight down the valley to the south-west. In form it is circular, with a diameter of about 115 feet, resting on the edge of the bank on the south-east; and has been surrounded by a bank of earth and stone some 15 feet broad at base, with, on the north arc, a trench some 25 feet in widtli having a sliglit mound on the counterscarp. In front of the trench the bank is more formidable than else¬ where on the periphery. The entrance has been from the lower side. Some 50 feet to the eastward, beyond the enclosure, is the site of a hut—an oblong with rounded ends— measuring 18 feet by 14 feet interiorly. xxvi. N.E. (“ Fort ”). 8 July 1912. 185. “ Stone Rings,” Hartmanor.—At the upper end of a meadow on the east side of the road to Eskdalemuir via the Shaw Rig, and about J mile north of the farm of Hartmanor, is a large tripartife enclosure consisting of an oval, with its longest axis north and south, measuring 180 feet by 130 feet, divided un¬ equally at about one-third of its length from the north end, and having annexed to it on the south-west arc a semi-lunar annexe. Sur¬ rounding the whole is a stony bank, some 14 feet broad at base and from 2 to 3 feet in height, entirely overgrowm, with entrances into the oval enclosure at either side of the dividing bank, on the upper or east side, and appar¬ ently an entrance into the annexe from the interior on the south. The construction is situated on ground sloping towards the south¬ west near the bottom of the valley, shows no particular defensive aspect, and is consider¬ ably scooped out in the interior. xxvi. S.W. 12 July 1912. 186. Enclosure, Watcarrick.—On the slope of the hillside, just overlooking the narrow' glen of a burn, about J mile to the south of Watcarrick farm, is a pear-shaped enclosure surrounded by a stony bank much reduced in height, and spread. At the broader end, from the w'est side, the bank curves inw’ard across the interior dividing it into tw'O portions. At the upper end there is a deep semicircular recess against the bank, w'hich appears to have been excavated. The entrance has been into the lower end. This construction show's no defensive character whatever, and has probably been a cattle enclosure. xxvi. S.W. (‘‘ Stone Ring ”). 9 July 1912. 187. Enclosure, Eskdalemuir.—At the base of the Castle Hill on the east and adjacent to the farm of that name, which lies a short distance to the south, is a double circular en¬ closure, in form of a figure of eight, one circle, the northmost, being of less diameter than the other and slightly recessed into it. The surrounding bank is of earth and stone, and the ground immediately behind it is distinctly 75 ESKDALEMUIR.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [eskdalemuir. hollowed out, while the whole interior area of the smaller circle has been excavated to a depth of 3 to 4 feet at the centre. On the east side of the smaller circle are the foundations of a small oblong turf hut, measuring interiorly some 12 feet by 9 feet, with a surrounding bank some 6 feet wide. The entrance into the main enclosure is into the larger circle at its lowest point. A small stream flows b}^ on the south. xxvi. S.W. (“ Tuan Rings ”). g July 1912. 188. “ Stone Ring,” Saugh Hill Plantation. —In the meadow on the lower side of the Saugh Hill Plantation near its north end, the O.S. map marks a ” Stone Ring.” It is the faintly discernible outlines of a circular en¬ closure on level ground presenting no defensive characteristics, nor is it excavated in the interior. The remains are probably those of an old pen. Between them and the end of the wood are one or two small hollows excavated on the south face of a knoll, which may possibly be the sites of dwellings. XXXV. N.W. 12 July 1912. 189. Stone Enclosure, Castle O’er Hill.—On the south side of the ravine into which the trench opens which runs down from Castle O’er fort, and on the east side of an old track, is a small four-sided enclosure, roughly a square of 50 feet, surrounded by a bank of stone and turf some 5 feet broad and from i foot to 18 inches high. It is marked on the O.S. map, but has probably been merely a pen. XXXV. N.W. 5 July 1912. 190. Enclosure, Bank Head Hill.—At the base of Yards Rig, Bank Head Hill on the south-west, and about J mile south-south-east of Castle O’er House, is an enclosure situated on a bluff formed by the debouchement of a ravine on a bank that margins the low-lying ground stretching to the White Esk, distant some 100 yards to the west. The enceinte is for the most part occupied by a sheep pen, and about ^ of the periphery has gone. The enclosure has been oval, measuring apparently about 170 feet by 130 feet, and has been surrounded by a bank formed of earth and stone some 26 feet wide at base and from 5 to 6 feet high. The entrance, which has been from the north, has been some 6 feet wide. XXXV. N.W. (” Fort ”). 12 July 1912. 191. Enclosure, Tanlawhill.—This enclosure is situated at an elevation of 800 feet above sea-level, on the summit of the hill which rises to the west-north-west of Tanlawhill farm. It is circular in form; the south arc, however, takes a slightly irregular line as it follows the contour of the summit at the edge of a steep downward slope to the bottom of the valley of the Tanlawhill Burn. The enceinte measures from north to south ig6 feet and from east to west 176 feet, and is encircled by a massive stony rampart rising on an average some 3 feet 6 inches to 4 feet on the interior, and showing evidence of the former existence of a stone parapet along its crest, with a concentric trench beyond, 30 feet in width on the north arc, where it is best defined, 10 feet in depth below the crest of the scarp, and 5 feet below that of the counterscarp, with an outer mound on the counterscarp. Overlooking the steep slope to the southward the rampart is much less prominent than elsewhere on the periphery, and the trench has been dispensed with. The entrance, which is from the east, is some 15 feet in width as it passes over the trench, and 5 feet through the rampart. The roadway is sunk below the ground level on the outside, and is so carried forward into the interior. A cross wall, curving from east to west, cuts off about a quarter of the interior contained by the south arc, and elsewhere there are indications of walls forming enclosures against the rampart. There has been much hollowing out of the interior, and for almost its whole length the trench is cut through rock. XXXV. S.W. (" Fort ”). 15 July 1912. 192, Enclosure, Westside.—On the top of the hill overlooking the farm of Westside, at an elevation of 852 feet over sea-level, is another large circular enclosure of the prevailing type. It is surrounded by a stony rampart some 22 feet broad at base with a concentric trench beyond, cut through rock and some 36 feet wide and 10 feet and 7 feet deep respectively below scarp and counterscarp with a mound above the latter. The interior is oval in form. ESKDALEMUIR.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [eskdalemuir. with its longest diameter north and south, and measures 200 feet by 175 feet interiorly. It has been considerably scooped out, especially in front of the entrance, which is from the direction of the valley and at the lowest point. Facing the entrance is a prominent rock, which, curving round to the north, de¬ marcates the outline of the hollow and forms an elevated platform beyond. On this plat, to the north of the entrance, is a pear-shaped mound of rock, overgrown with tnrf, a segment of a circular enclosure with a diameter of 40 feet, the ground having been hollowed to a depth of 2 feet within it. This rock founda¬ tion has been fashioned by the cutting away of the rock on either face, and is analogous to a similar foundation in Castle O’er fort (No. 177). XXXV. N.W. (“ Law Birren (Fort) ”). 15 July 1912. 193. Enclosure, Hamlin Knowe, Westside.— This enclosure is situated on the summit of an eminence which rises from the wide expanse of white hill-pasture lying to the westward of the dale of the Black Esk and about | mile due west of Westside. It lies at an elevation of 935 feet above sea-level and some 330 feet above the bottom of the valley, whence there is an approach up a long hollow. The enclosure is of the type prevalent in this region, and is oval with its longest axis east and west ; it shows, however, a departure from the general plan in a semi-lunar projection from the true line of the oval, which forms a bay or fore¬ court at the east end. The interior measure¬ ments tlirough the centre are 227 feet and 133 feet, while the bay has a chord of 60 feet and a bisectional diameter of 36 feet. The enceinte has been surrounded by a stony rampart, now for the most part demolished, with a steep scarp to the outside some 8 feet in height, covered by a shallow trench. The entrance is from the east and is 8 feet wide, sunk beneath the general level, the roadway being carried on into the interior up a hollow between two masses of outcropping rock. Within the main area of the interior, to the right of the en¬ trance, is a circular hollow 27 feet in diameter, cut out of rock, a segment of which is left around to form a foundation, as in the last example. The interior is uneven on the surface, but there has been less hollowing of it than usual. XXXV. N.W. (“ Fort ”). 15 July 1912. 194. Enclosures near Westside.-—About J mile to the south of Westside, on the top of a hill which rises up from the south side of Henrie’s Burn near its confluence vith the Black Esk, is a large pear-shaped enclosure which has been surrounded by a stony rampart now almost reduced to ground level. The in¬ terior has been hollowed, so that the enclosure resembles the bed of a large pond. On the north the bank has been kept back from the edge of the plateau, so that no part of the construction is observable on the skyline from the valley below. The elevation of the summit is 794 feet above sea-level. On a plateau on the north slope of the hill, less than halfway down to Henrie’s Burn, is a small circular enclosure, measuring interiorly some 42 feet, surrounded by a turf bank 12 feet thick at base and about 2 feet high, hollowed to some extent in the interior and seemingly entered by a narrow opening 2 feet wide from the west-south-west. Some 30 feet to the east lie the foundations of two conjoined parallel enclosures, quadrangular, and measuring interiorly 32 feet by 12 feet, which are surrounded by banks. On the opposite side of the burn, some 200 yards up the hillside, is another small enclosure or hut-circle, oval in form, measur¬ ing 23 feet by 20 feet, surrounded by a turf bank 9 feet thick at base and 2 feet high, and with a narrow entrance from the south¬ east not clearly defined. Within 15 feet to the westward, occupying the angle of the feal dyke of an old field, is another small enclosure of similar character. XXXV. N.W. (unnoted). 15 July 1912. Hut-Circles. 195. Hut - Circle, Carterton Knowes.—On the moorland which seems to pass under the general name of Carterton Knowes, lying on the west side of the Black Esk and near the source of the syke that flows into that stream, about J mile to the south of Henrie’s Burn and facing south under shelter of a slight rocky eminence on the north, is a small circular enclosure or hut-circle, measuring 77 ESKDALEMUIR.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [eskdalemuir. 26 feet in diameter interiorl}’ and surrounded by a turf bank about 4 feet wide. The entrance has evidently been from the south¬ east. XXXV. N.W. (unnoted). 15 July 1912. 196. Hut-Circles near Westside.—About 100 yards to the eastward of No. 192, down the hillside, lies a hut-circle oval in form and measuring interiorly 19 feet by 16 feet, with its longest axis north-east to south-west and surrounded by a turf bank some 3 feet 4 inches thick. As in the larger enclosure on the hill top adjacent, there is a semi-lunar projection at the south-west extremity cover¬ ing the entrance, vdth a chord of 10 feet and a bisectional diameter of 6 feet, into which is an entrance from the south. Farther east and about J mile down from the larger enclosure, is another hut-circle, measuring interiorly 18 feet by 19 feet. The enclosing bank is some 8 feet in thickness at base and has a height all round of about 18 inches. The position of the entrance is not at all clear, though it has possibly been from the east. xxxv. N.W. (unnoted). 15 July 1912. 197. Hut-Circles, Raeburn.—Several of these hut-circles are to be seen between Mid- Raeburn and Raeburnfoot on the right bank of the burn, others on a slight eminence on the side of the burn opposite to the Roman camp, and one within the outer precinct near the north end. There is also one on Bankhead Hill, a short distance to the west of the higher of the two enclosures near the Bankhead Burn. xxvi. N.W., N.E. (unnoted). 15 July 1912. Stone Circles. 198. Stone Circle, “ Girdle Stanes.”—On a narrow level haugh on the left bank of the White Esk, 300 yards north-west of Cot, at an elevation of 585 feet above sea-level, are the remains of the “ Girdle Stanes ” stone circle (fig. 63), a portion of which on the west seems to have been carried away by the river. What remains is a crescentic arrangement of twenty-six stones, forming about three-fifths of a circle, with an approximate diameter of 78 about 130 feet. In the bed of the stream are six or eight boulders of no great size, which possibly were part of the missing arc. The stones are placed very irregularly both as regards alignment and distance apart. Of the twenty-six Stones which remain ten are erect and are more or less pillar-like, one < being a double stone, four are prostrate pillars, one is a fiat slab, and the remainder are squat. The standing-stones vary from i foot 6 inches to 5 feet 4 inches in height and the prostrate pillars from 5 feet 6 inches to 7 feet in length. Probably a considerable number of stones has been removed. See Antiquaries, xxxi. p. 285. xxvi. S.W. 5 July 1912. 199. Stone Circle, “ Loupin’ Stanes,” near Hartmanor.—The stone circle known as the “Loupin Stanes”-(fig. 64) is situated at an elevation of 600 feet above sea-level, on an undulating meadow some 80 yards from the left bank of the White Esk and 600 yards north-east of the “ Girdle Stanes ” circle (No. 198). It consists of twelve stones placed at irregular intervals, of which, on the west side, only two, standing 8 feet apart (A and B on plan, fig. 64), are true pillars, all the others being simply boulders. ESKDALEMUiR.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [eskdalemuir. The setting does not form a true circle but is flattened on the western arc ; it measures some 38 feet in diameter from north to south and 31 feet from east to west. The pillar stones, which are flat-topped, measure about 5 feet 4 inches in height, and the two highest of the smaller stones do not exceed 2 feet. while few of them reach i foot. The site has been banked up nearly all round from I foot to I foot 6 inches, perhaps with the object of making the interior level. There are numerous large stones adjacent, some of which may be the remains of other circles, but others are probably merely boulders naturally deposited or outcrop of rock. The field was in a hay crop at the date of visit, and the fallen stones were not easily seen. See Antiquaries, xxxi. p. 281. XX vi. S.W. 5 July 1912. Miscellaneous. 200. Entrenchment, Old Graveyard, Bank- head, Eskdalemuir. — This entrenchment, now occupied by a burial ground with a modern wall along the scarp of its ditch, lies on the right bank of the White Esk, some 25 feet in elevation above the river, and at the south end of a hillock traversed by the road from Lang¬ holm to Eskdalemuir. In plan it is rect¬ angular, measuring about no feet by 100 feet, and is surrounded by a trench some 25 feet broad and 6 feet in depth below the scarp. A mound forms the counterscarp, sharply defined on the south and east, rising some 5 feet above the bottom of the trench on the south and 3 feet on the east. On the west the defences have been destroyed by the formation of the road. Almost in the centre of the east side, facing the river, there has been an en¬ trance, seemingly some 7 to 8 feet wide where it passes through the outer mound. This construction is marked on the O.S. map as the site of a chapel. xxvi. S.W. 9 July 1912. 201. Entrenchment, Bank Head Hill.—On the west side of Bank Head Hill, where its lowest slope terminates on a steep bank over¬ looking the White Esk, and about | mile north of Castle O’er House, on the opposite side of the river, is a short segment of an entrench¬ ment, some 6 feet deep and 24 feet wide, con¬ tinuing for a distance of about 100 feet. It is not carried forward to the bank at either end, but terminates fully 100 feet distant from it, nor is there any indication that the fosse ever continued further. XXXV. N.W. {“ Fort ”). 12 July 1912. 202. “Tumulus,” Allangillfoot. — About ^ mile above Allangillfoot two burns meet, and on the left bank of that flowing from the north¬ east, a little less than \ mile above the junc¬ tion and just beyond where a rivulet empties itself into the stream on the right bank, is an artificial mound of earth and stone measur¬ ing over all some 15 feet by 18 feet and in height about 3 feet. It lies at the foot of the steep bank of the gill, close by the stream, and water from a spring at a higher level keeps the ground wet in front of it. Some excavation has been done on the top of the mound, but nothing has been revealed to show its character. xxvi. S.W. (unnoted). 12 July 1912. 79 ESKDALEMUIR.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [ewes. 203. Covenanter’s Tomb, “ Hislop’s Grave,” Craighaugh.—In a field adjacent to the high road, and some 300 yards to the north-north¬ west of the hamlet of Craighaugh Holm, is a “ through ” stone, with an inscription on the upper slab commemorating Andrew Hislop, a covenanter, shot on the spot by Sir James Johnston of Westerhall and John Graham of Claverhouse on 12th May 1685 and buried here. The inscription, which is of 18th-century character, and appears to have been re-cut, reads as follows : HERE LYES ANDR. HISLOP / MARTYR SHOT DEAD UPON / THIS PLACE BY SIR JAMES / STE JOHNSTON OF WERHALL / AND JOHN GRAHAM OF C / LAVERHOUSE FOR ADHERI / NG TO THE WORD OF GOD / CHRIST’S KINGLY GOVERN- / MENT IN HIS HOUSE AND / YE COVENANTED WORK OF / RE¬ FORMATION AGNST TYRAN / NY, PERGURY AND PRELACY / MAY 12 1685 REV 12. II. HALT P / ASSENGER, ONE WORD WI- / TH E THEE OR TWO WHY I LY / HERE WOULDST THOU TRU- / LY KNOW BY WICKED HAN- / DS, HANDS CRUEL AND UNJ- / UST WITH- o OUT ALL LAW / MY LIFE FRM ME THEY / THRUST & BEING DEAD / THEY LEFT ME ON THIS S / POT, & FOR BURIAL THIS / SAME PLACE I GOT. TR- / UTHS FRIENDS IN ES / KDALE NOW TRIUMPH / THEN LET {sic), VIZ THE FAITH- / FUL FOR MY SEAL [that] / GOT 1702. xxvi. N.W. 9 July 1912. Sites. 204. Enclosure, Eskdalemuir.—This en¬ closure lies on the west side of the glen, adjacent to the church of Eskdalemuir, on cultivated land under crop at the date of visit. The O.S. map shows it as a triangular figure surrounded by a rampart, with the north and west sides fairly straight and almost at right angles to one another, subtended by a curve. It lies on a slope, has been hollowed out in the interior, and measures apparently about 105 feet along each of the straight sides. xxvi. S.W. (” Fort ”). 9 July 1912. 205. Enclosure, Birren Knowes, Todshaw- hill,—About J mile to the east of Todshawhill farm-house, on the east side of the summit of Todshawhill and between the Birren Knowes, are the fragmentary remains of a hollowed enclosure. A segment of a stony rampart remains on the south, but elsewhere is only faintly and intermittently traceable. There are the usual indications of excavated hollows in the interior. XXXV. N.W. (” Fort ”). 15 July 1912. The O.S. maps indicate sites as under :— 206. Chapel, Bankhead, xxvi. S.W. 207. Fort near Davington. xviii. S.W. 208. Tumulus near Raeburnfoot. xxvi. N.W. EWES. Defensive Constructions. 209. Fort, Tarrona.—On the lower slope of Old Hill, not more than 100 feet above the low meadow which intervenes between its base and the Ewes Water and some 120 yards to the south-east of Tarrona farm-house, is a defensive enclosure or fort elliptical in form, lying with its main axis north-north-west and south-south-east, and resting at the latter end on the steep bank of a small ravine, down the bed of which flows the Foots Burn. The area of the fort has been encircled by a broad rampart, now most prominent on the north¬ east or upper side, where, at highest, it has an elevation from the outside of some 3 feet 6 inches, and, owing to the levelling of the interior, an inner face some 9 feet high. On the opposite or lower side of the enceinte the rampart is low, but the ground outside it has been steeply scarped to as much as 15 feet of vertical height. All round it appears to have been surmounted by a stone parapet, the stones of which in some parts, and the bed from which they have been removed in others, are clearly discernible. The enceinte has measured some 230 feet along the longest axis by 193 feet bisecting it. Beyond the field dyke, which cuts across the enclosure at the side of the ravine, the rampart is traceable, curving inwards from either side to pass along the edge of the bank 80 EWES.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [ewes. where it is steepest. To the eastward of where it impinges on the bank at the upper side there is a gap in the periphery opposite a hollow which leads down to the burn, and which may have served for watering cattle. In the middle of the west side there is an opening, evidently enlarged, measuring some 15 feet across the bottom, which may have been the original entrance. A narrow gap towards the north is probably secondary. There is no indication of any trench outside the rampart. xlv. N.E. 29 June 1912. 210. Fort, Brieryshaw Hill.—On the sum¬ mit of Brieryshaw Hill is a fort. The hill has an elevation above sea-level of some 700 feet ; north-west of it rises the lofty Stake Hill to an altitude of 1348 feet, while to the south-east the ground falls with a con¬ siderable declivity to the Ewes Water in the valley below. The fort, placed at the east edge of the plateau that forms the summit, is circular in form, measuring interiorly some 192 feet in diameter, and is surrounded with a rampart of earth and stone some 23 feet broad at base, having around the greater part of the periphery an outer concentric mound some 33 feet broad at base, with an inter¬ vening trench measuring from 28 to 35 feet in width from crest to crest and some 7 feet in depth. Towards the east, overlooking the downward slope, the trench and outer mound gradually give place to a terrace some 18 feet broad with an outer ramp. Beyond these defences, on the west, there is an outer concentric trench some ii feet in width, deepening as it passes southward. The en¬ trance, 8 feet wide, has been from the east or lowest side of the fort, and opens at a low level into an excavated hollow some 6 feet deep, which is carried back with a rising gradient to the rampart on the north and north-west and is flanked by a ridge of higher unexcavated ground on the south ; this ridge, however, has also been excavated in places. On the highest area are the foundations of an oblong hut with its longest axis east and west, measuring in¬ teriorly some 22 feet by 9 feet, with a sur¬ rounding bank apparently of turf or soil some 6 feet thick and having a doorway 2 feet 6 inches wide'in the south wall at 8 feet from its east end. It is noticeable that the 81 rampart on the half of the periphery that encloses this site is higher and more massive than on the north half. The inner rampart appears to have been surmounted by a stony crest, traces of which are visible on the south arc. xxxvi. N.E. 13 July 1912. 211. Fort, Loch Hill.—The summit of the Loch Hill, which rises to a height of 824 feet above sea-level in the angle between the Glen- divan Burn and the Ewes Water, has been surrounded by a broad, stony rampart with a scarp cut on the natural surface below it— of which, for the most part, the scarp alone remains. A portion of the rampart, however, remains on the north, and measures some 22 feet broad at base with a scarp on the exterior some 6 feet high. The enceinte has been circular, following the outline of the summit of the hill, with an approximate interior diameter of 244 feet. Except along the sum¬ mit to the north-west, whence the ascent is gradual, and towards the greater height on the south-east, the flanks of the hill have a steep declivity. The entrance has been from the north. In a hollow to the north-west lies a large pool or pond. xxxvi. S.E. (unnoted). 13 July 1912. Enclosures. 212. Enclosures, Fiddleton.—An enclosure is situated about 700 feet over sea-level on a plateau some 100 yards to the north-west of a cottage at Fiddleton, round the base of which, on the north and west, flows the Glen Burn, while on the north-east it is demarcated by a ravine opening on to the bed of the burn. The enceinte has been bean¬ shaped, controlled by the situation, and. except on the south-west arc, towards the higher slope of the Pen Craig, its outline is indistinct. In that direction, at the highest point, there is a scarp some 6 feet high, formed by excavation, which gradually merges into the slope of the interior, while, where the ground falls away to the westward, there is a swelling indicating a rampart. Across the Glen Burn, about 100 yards to the north-north-west, and at a slighth- higher elevation is another enclosure, also occupying a plateau, with a grassy hill rising to the EWES.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [ewes. north-west to a height of some 1714 feet above sea-level. On the east, from the front of the plateau, the ground falls sharply to the valley of the Mosspaul Burn, while an equally steep gradient carries it to the bed of the Glen Burn on the south. The construction has been four-sided with slightly curving sides and rounded angles, and, set with its longest axis north-west and south-east, has measured 160 feet by 143 feet. The upper or north-west end has been formed by excavation leaving a scarp some 6 feet in vertical height, above which is traceable a slight stony mound, which elsewhere marks the periphery, with a breadth at base of about ii feet. The entrance has been in the east angle from the edge of the bank overlooking the burn, and has been some 7 feet in width. In the centre of the interior there is a stony artihcial mound, with an elevation of about 3 feet above the lowest level, and against the west side are discernible the foundations of a small oblong structure. In the north angle there is a circular or pear- shaped foundation measuring some 33 feet by 28 feet, with its longest axis south-west and north-east. xxvii. S.E. (unnoted). 3 July 1912. 213. Enclosure, Blackball.—This enclosure is situated about 80 yards to the south-east of Blackhall at the foot of the Blackball Hill and at the edge of a steep bank some 25 feet high above the Carewoodrig Burn, which flows by it on the west. It has been in shape an elongated oval with its longest axis parallel to the edge of the bank, measuring interiorly some 150 feet by 87 feet, but has suffered so much from the plough that the mound which probably surrounded it is now only recognis¬ able at the edge of the bank, where it has a breadth at base of from 8 to 14 feet. The interior has, as usual, been hollowed out. xxvii. S.E. (unnoted). 3 July 1912. 214. Enclosure, Upper Hill, Unthank.-—This enclosure is situated on the hip of the Upper Hill, on the west side of the Ewes Water, opposite Unthank and to the north of the Sikefoot Burn, which flows down in a series of cascades past a cottage at the roadside. It is at an elevation of some 700 feet, the ground declining sharply from it to east and west and rising above it. In form it is sub¬ oval, measuring interiorly some 173 feet by 142 feet, and is surrounded by a stony bank some 14 feet broad at base and rising to a height of about 3 feet. The entrance has been on the lower side from the south ; and some 25 feet in from it the interior has been crossed by a stony bank cutting off about one-fourth of the area to the south, with a gap through it directly opposite the entrance. The interior is less hollowed out than is generally the case in these structures. xxvii. S.E. (unnoted). 3 July 1912. 215. Enclosure, Mosspeeble.—This enclosure is situated at an elevation of some 530 feet over sea-level to the south-east of Mosspeeble farm, occupying the greater part of a plateau, into which the north slope of Hartrith Rig breaks before dropping to the Ewes valley. It has been oval in form with a longest diameter of some 250 feet, and is surrounded by a bold, stony mound some 26 feet wide at base, rising on the upper or south-east side to a height of 6 feet above the ground immediately behind it. Outside the mound, on the north¬ east arc, there runs a flat-bottomed trench some 26 feet wide, the floor of which lies some 5 feet below the crest of the scarp and at most 3 feet 6 inches below that of the counter¬ scarp. The north-west quadrant of the per¬ iphery has been almost entirely eradicated, and on its site stands a shepherd’s cottage. The interior lies at two levels. An irregular and apparently excavated hollow runs across the inner side of the north segment, leaving the ground on the south at a general higher level of some 3 to 4 feet. There is a distinct trace of a stony bank crossing the hollow and bisecting the higher platform. The entrance seems to have been from the west into the lowest part of the hollow. xxxvi. N.E. (unnoted). 3 July 1912. 216. Enclosure, Meikledale.—Situated some 200 yards to the north-east of Meikledale House, on a plateau at the base of a steepish slope which entirely overlooks it, are the remains of another enclosure ; it has been fre¬ quently ploughed over and its features greatly altered. The interior has been hollowed to some extent. Some 50 feet east-north-east is EWES.J INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [ewes. a small triangular enclosure measuring some 50 feet on each face, also hollowed out. XXXvi. N.E. (unnoted). 3 July 1912. 217. Enclosure, Meikledale.—This enclosure is situated at the edge of the steep bank on the north side of the Meikledale Burn, in an angle formed to the west of where that stream is joined by a smaller burn flowing down from Stibbiegill. The construction is oblong in form with its longest axis north-west and south-east and measures 200 feet in length, 50 feet in breadth at the north-west end, and 104 feet at the south-east, which is curved. On the upper side the ground has been con¬ siderably excavated, and there is a barely discernible stony mound along the crest of the scarp, which, however, is clearly defined, though of low elevation, on the other three sides. It has been composed of flatfish stones, apparently piled up, and has measured some II feet in width at base. The entrance, some 6 feet wide, has been near the south angle adjacent to the edge of the bank, giving access, as usual, to the lowest part of the enceinte. The enclosure lies at an eleva¬ tion of about 500 feet over sea-level. xxxvi. N.E. (unnoted). 3 July 1912. 218. Enclosure, Rigfoot.—The shepherd’s cottage at Rigfoot, about 500 yards west- north-west of Meikledale farm, is situated in another of these excavated enclosures. xxxvi. N.E. (unnoted). 3 July 1912. 219. Enclosure, Brieryshaw Hill.—About 100 yards to the southward of No. 217 and at a slightly lower level, is another circular or oval enclosure, with its longest axis north¬ east and south-west and measuring along it some 193 feet. It has been surrounded by a bank of earth and stone much worn down in places and highest towards the south, where it has an elevation of some 3 feet 6 inches. The interior has been considerably hollowed in parts, and is divided into two by a stony bank which crosses it from east to west a little to the south of the centre. Against the rampart on the south is a circular plat, which appears to be the site of a hut. There has been an entrance, 3 to 6 feet wide, into the north division by the side of the dividing wall, which is marked by a stone standing upright on one side to a height of 2 feet 5 inches, the ground in front of which is. very wet. The entrance into the south division appears to have been fi'om the south-east, adjacent to the hut site, and is 2 feet wide. xxxvi. N.E. (unnoted). 13 July 1912. 220. Enclosure, Arkleton. — This enclosure is situated, at an elevation of 600 feet above sea- level, about I mile east-north-east of Arkleton, on the crest of a hog-backed grassy ridge bearing the name of the Birren Rig, which lies between the Birren Sike on the east and a burn which flows down to Arkleton Water on the west. In form it is irregularly cir¬ cular, measuring about 190 feet in diameter, and is surrounded by a broad rampart of earth and stone some 30 feet broad at base and 6 feet high. There have been two entrances— one from the south-east into the lowest part of the interior, opening into an excavated hollow- measuring approximately some 70 feet by 60 feet in diameter, which is flanked by higher ground along the northern half of the interior and diminishes gradually in depth on the south; the other from the south-west, also opening on a hollow from 60 to 70 feet in diameter, which is faced by an arc of higher and unexcavated ground towards the north. As is usually the case in these enclosures, there are various depressions both in the higher and lower areas demarcated by banks of soil or of unexcavated rock, varying in size and form but suggestive of sheep or cattle-pens. Against the rampart on the north, at the highest point, a stony foundation projects into the interior for a length of 42 feet with a return to the east at right angles for 33 feet. xxxvi. N.E. 13 July 1912. 221. Enclosures, Upper Glendivan. — Some 300 yards above the shepherd’s cottage at Upper Glendivan is an oval excavated en¬ closure, and the remains of another are to be seen on the top of the bank of the burn, immediately to the south of the cottage. xxxvi. S.E. (unnoted). 13 July 1912. C.MRX. 222. Cairn (remains of), Sorbie Bridge. — Some 50 yards up tlie wooded glen above Sorbie road bridge and on the north side of tlie EWES.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [glencairn. path, are the remains of a large circular cairn— a segment consisting of about a quarter of the original, measuring some 7 feet in height. It is said that a stone cist was found when the cairn was being removed for road metal many years ago. xxxvi. S.E. (unnoted). 13 July igi2. Miscellaneous. 223. Standing-Stone, Meikledale.—A stand¬ ing-stone, known as the “ Grey Wether,” is situated on the haugh-land, some 250 yards south-south-east of Meikledale. It is a large whinstone slab, measuring in greatest height 4 feet 8 inches, in breadth 3 feet 5 inches, and in thickness i foot, and faces west-south-west and east-north-east. xxxvi. N.E. (unnoted). 3 July 1912. 224. Church, Foundations of. Unthank.— Within the old churchyard of Unthank the foundations of a church are indicated by grassy banks, but measurements are unobtainable. 225. Table-stone, Churchyard.—Lying to the south of the west end of the foundations of the church within the graveyard at Unthank is a ” through ” table-stone, measuring 5 feet 2 inches b}^ 3 feet io| inches, inscribed at the top : HEIR LYES ADAM ELLIOT OF MEIKLEDALE WHO DIED AUG l6 1682 AGED 83. In the centre of the slab is a shield, bear¬ ing arms : A lion sejant full-faced, holding in his dexter paw a thistle slipped, in the sinister a shield bearing four pales within a bordure, on a chief a saltire. xxvii. S.E. 3 July igi2. 226. Carved Stones, Arkleton.—Built into the wall of the entrance hall at Arkleton is an oblong slab inscribed in four lines as follows — WALTER ELIOT AL VERTEV TOIL AL EARTHLY VERTEV IS EOT TOIL & PAIN SO AL IS LOS ; VNLES THAT CHRYST WEE GAIN. The second line forms an anagram of the names given in the line above. Embedded in the wall on the outside of the house is another stone commemorating the marriage of this Walter Eliot with Katharine, daughter of Forster of Stonegarthside, and inscribed W E • K F 1671. xxxvi. N.E. 3 July 1912. 227. Bell, Ewes Kirk.—Hanging on a tree to the south of the church of Ewes at Kirk- style is the old bell, i foot 6 inches in height and I foot 5 inches in diameter at the mouth, inscribed, between a double leaf border, JACOBUS MONTEITH ANNO 1652. xxxvi. S.E. 13 July 1912. Sites. The O.S. map indicates a site as under :— 228. St Cuthbert’s Church, Kirkstyle. xxxvi. S.E. GLENCAIRN. Ecclesiastical Structure. 229. Old Church.—The church of Glencairn is among a number of churches confirmed to the Bishop of Glasgow in a bull of Pope Alex¬ ander in 1178. In the 15th century Bishop Turnbull granted it to the chapter as a common church. The east and west gables of the old church remain with a few feet of ruinous fragments of the side walls close to the east end. The latter contains two narrow lancet windows, plainly chamfered on the outside, of which that to the south is 4 feet 8 inches in height and 7 inches wide internally, with splays on the inside 4 feet wide, the edge of the splay being 3 feet 9 inches from the ground. The other seems to be at a higher level, but is impenetrably cased in ivy. The east wall is 3 feet 10 inches thick, the side walls 3 feet 2 inches, the west wall 3 feet 10 inches. The lower part of the east wall, with the windows, probably dates from the 13th century, but the upper part has been altered at a later date and finished with crow-steps. Both gables are thickly covered with a growth of ivy. XXX. S.E. 19 Oct. 1915- 84 GLENCAIRN.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [glencairn. Sepulchral Monuments. 230. Mural Monument. —Outside the east end of the old church, and built against the northern part of the east wall, is a hne archi¬ tectural monument, consisting of a central recessed panel, flanked by fluted pilasters, with pyramidal finials. Between the finials on a cornice is a moulded panel, on which is carved a skull and cross-bones, an hour-glass, a spade, a hand holding a bell, and an object resem¬ bling an axe. The central panel has been inscribed, but the inscription has almost entirely weathered off. Above it five lines of inscription read as follows :— VIRO VIRTUTIBUS UT FORTUNIS / BEATO PATRI SUO BENIGNISSIMO STEPHANO / LAURI A MAXWELTOUN MNEMOSYNON HOC / LAPIDE-IS EXPOSUIT / OBIIT I4 -BRIS MDCXXXVH 231. Mural Monument. —Built into the east wall of the churchyard is a panel with the incised inscription :— THERE IS NO OBTAINING THE PRIZE OF HAPPINESS WITHOUT RUNNING THE RACE OF HOLINESS. The date is probably late 17th or early i8th century. 232. Martyrs’ Tombs. —Near the centre of the old church is a table-stone, measuring 5 feet 61 inches by 2 feet, inscribed ;— HERE LYES JAMES / BENNOCH SHOT DE / AD BY COL. DUGLAS / AND LIVING¬ STONS / DRAGOONS AT ENG / LSTON FOR ADHE / REING TO THE WORD / OF GOD CHRIST’S KI / NGLY GOVERMENT / IN HIS HOUSE AND / THE COVENTED (sfc) WO /RK OF REFORMATION / AGAINST TYRANNY / PERJURY AND PRELA / CY APR. 28 • 1685 REV / . 12 . II. The inscription at right angles to the fore¬ going is now imperfect, owing to the stone having been broken, but the whole has been re-cut on a new stone placed alongside two others within an enclosure in the churchyard. These two latter commemorate Robert Edgar with Robert Mitchell, and John Gibson, all covenanting martyrs. The main inscription is almost identical in the three cases, that given above varies very slightly in one or two words from its companions ; the rhymed additions at right angles are as follows for each of the stones respectively :— I. Bennoch. HERE LYES A MONUMENT / OF POPISH WRATH / BECAUSE i’M NOY PERJUR / D i’m SHOT TO DEATH / BY CRUEL HAND’S MEN / GODLES AND UNJUST / DID SACRIFICE MY BLOOD / TO BABELS LUST. 2. Edgar and Mitchell. HALT PASSENGER TELL IF / THOU EVER SAW / MEN SHOT TO DEATH / WITHOUT PROCESS OF LAW / WE TWO OF FOUR WHO IN / THIS CHURCHYARD LY / THUS FELT THE RAGE / OF POPISH TYRRANY. 3. Gibson. MY soul’s in heaven / HERE’S MY DUST / BY WICKED SENTANCE / AND UNJUST / SHOT DEAD CONVICTED / OF NO CRIME / BUT NON-COMPLYANCE / WITH THE TIME / WHEN babel’s bastard / HAD COMMAND / AND MONSTEROUS TYRRA / NTS RULD THE LAND. XXX. S.E. 19 Oct. 1915. Castellated and Domestic Structures. 233. Old Crawfordton.—This small ruin is situated on rising ground to the south-west of the Cairn Water, about 3 miles by road to the south-east of iMoniaive. A range of farm buildings abuts against the north¬ eastern angle of the old building, of which the basement, some 8 feet in height above the ground level, is the only portion now remaining. On plan (fig. 65) it is a simple oblong, measuring about 34 feet by 19 feet 6 Fig. 65.—Old Craw¬ fordton (No. 233). inches, with walls varying from 2 feet 6 inches to 4 feet 6 inches in thickness. The entrance is in the centre of the north wall, giving access to a passage which communi¬ cates witli two vaulted cellars and with a wheel-stair at the nortli-western angle. The eastern cellar is lighted by a narrow window GLEXCAiRX.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [glencairn. in the north and east walls, and the western cellar has a single light and a small aumbry in the south wall. The building is built of rough rubble throughout. It is now in a very ruinous state, and is densely covered with ivy. John Crichton of Crawfordton is on record in 1656 {Reg. P.C., s.d.). xxxix. N.E. 7 June 1912. 234. Breckonside Tower.—The ruins of this tower lie | of a mile north-north-east of Crossford Station on the Cairn Valley Light Railway. The}’ consist of a tower to the west of the site, measuring exteriorly 29 feet 6 inches by 21 feet, with walls 4 to 5 feet in thickness, uith an adjoining outbuilding to the north and another 5 feet distant to the east. The walls of the tower are only 3 feet above ground, and those of the other buildings are reduced to turf-grown mounds. John Maxwell of “ Brakensyd ” is among the Borderers sworn to Edward VI. in 1552-3. {Border Papers, i.. No. 396.) xl. N.W. 6 June 1912. 235. Maxwelton House.—This mansion is situated on the high left bank of the Cairn Water, some 3J miles east-south-east of Moniaive. It has been considerably modern¬ ised, but the harled walls, steep slated roofs and crow-stepped gables maintain an appear¬ ance in keeping wdth a 17th-century origin. To-day the most interesting features are ; {a) The lintel on the south wing, inscribed :— NI CCEPTA DOMINUS IVVERIT FRUSTRA STRUIS MOLES SUPERBAS AEDIUM (Unless the Lord assist your undertakings, in vain you erect proud buildings). The lintel is protected by a contemporary label and is surmounted by a moulded panel, framed by a panel mould similar in section to the label, which bears a shield charged per pale a cup beneath two laurel branches—for Lawrie ; and a muUet between three cushions—for Grierson (cushions being an older Grierson charge than the locks). Above the shield are the initials 1 . L. and A. G. for John Lawrie, eldest son and successor of Stephen Lawrie, who had pur¬ chased Maxwelton from the Earl of Glencairn, and his wife Agnes Grierson, elder daughter of Sir Mdlliam Grierson of Lag ; below, between two sex-foliated flowers the date 1641 : {b) On the gable of the same wing a second armorial panel, but at too great a height to be deciphered ; {c) Above the entrance within the conservatory a later panel also moulded, and bearing beneath a helmet and mantling a shield charged per pale a cup with a garland between two laurel branches for Lawrie, and a chevron between three ears of rye for Riddell. The achieve¬ ment is surmounted by a label inscribed viRTUS SEMPER viRiDis (Virtue ever flour¬ ishing). Beneath the shield is a cartouche inscribed thus : “ Sir Robert LavTie. Dom. Jean Riddell.” XXX. S.E. 6 June 1912. Defensfv^e Constructions. 236. Fort, Castle Hill, Dalwhat Glen.—This fort is situated on the Castle Hill, a spur of Glenskelly Hill, projecting into the Dalwhat Glen some 3I miles above Moniaive, with an elevation of 1000 feet or thereby over sea-level. On the south-west it rises by steep grassy slopes for some 400 feet from the bottom of the glen ; on the south-east the foreground undulates in hillocks and hollows for some 300 yards before it assumes a steeper gradient on the side of the valley ; on the east is a comparatively easy inclination, and on the north-west the ground dips to a hollow a short distance in front and then rises in tiering heights to Glenskelly and the higher hills beyond. The site of the fort is on a saddle-backed ridge. It is sub-oval in outline, having its main axis north-north-west and south-south-east. The defences, except at the north-north-west end, are faint and indistinct, but at that point are well pre¬ served. A slight stony mound encircling the oval has surmounted a scarp with an average height of about 5 feet, which rises from a terrace some 17 feet in width at the north-north-west end but diminishing in breadth along the flanks. At this end there is an indication of a slight mound at the edge of the terrace, which drops by a scarp some 5 feet in height to a narrow trench some 6 feet wide, the upcast from which has formed a low mound to the outside. Passing along the flanks this trench gradually merges into a lower terrace, which, some two-thirds distant towards the south- GLENCAIRN.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [glencairx. south-east end along the west side, merges in the steeper slope of the hillside. Along the east side the remains of the lower terrace are very slight, and disappear entirely at a hollow which leads up into the interior from the east-north-east near the centre, and which has probably been the entrance. The length of the interior is some 272 feet and its breadth at the centre 158 feet. The highest point in the centre of the fort is some 18 feet above the upper terrace on the west side and II feet on the east. The lines of this fort recall those of the fort near Laggan Loch, Glasserton, in Wigtownshire.^ 1 Inventory of the Monuments of Galloway, vol. i. [County of Wigtown), No. 5. xxix. N.E. 3 June 1912. 237. “ Mote,” The Orchard, Snade.—Close beside the railway, on its west side, about | mile east by south of Boreland, is a large circular entrenched earthwork (fig. 66). It is situated on low-lying ground opposite a loop Fig. 66.—"Mote,” The Orchard, Snade (No. 237). of the Cairn Water, which flows by at its nearest point some 170 yards distant. The ground has a natural fall from the west towards the centre of the construction, which has been raised a little above the dead level as it e.xists to north and south. A central circular plateau. measuring 103 feet by 116 feet, around the edge of which there appears to have been stonework, has been surrounded by a rather flat-bottomed trench, soft under foot, measur¬ ing from 23 to 26 feet in breadth. Beyond this rises a massive rampart some 5 feet 6 inches in height towards the higher ground on the west, but 5 feet on the east, and at the highest point rather flat on the crest. Outside of this is a second encircling ditch, some 40 feet in width across the top and 7 feet and 10 feet deep below scarp and counterscarp respectively on the west and east. This ditch is for the most part wet ; and, possibly for the purpose of flooding it, there is a break in the continuity of the outer scarp towards the south; while the inner mound, in the same direction, is reduced to a level of about i foot above the present bottom of the ditch and is lower than the level of the interior plateau. Towards the east there is a slight hollow in the scarp of the outer ditch, and, extending southward from it, a rampart surmounts the counterscarp. On the south a modern ditch has been cut to drain away the water from the outer ditch. xl. S.W. 5 June 1912. 238. Lower Mote, Ingleston.—About i mile east-south-east of Moniaive there rises out of the flat and somewhat marshy land that lies in the bottom of the Cairn Valley a prominent elongated hillock of sand and gravel, having its main axis east-north-east and west-south¬ west (fig. 3 of Introduction). This hil¬ lock has been formed into a mote-hill and base-court by the erection of an entrenched mound at its west-south-west extremity, and by levelling the summit and steeply scarp¬ ing the sides towards the east-north-east, where a second trench across from side to side forms the termination of the base-court in that direction, 43 feet distant from the actual end of the hillock. On the east the Jarbruck Burn flows by near the base ; and all around, in former times, the ground was, no doubt, marshy. The mote rises to a height of 36 feet from the ground level and slopes sharply, except towards the foot, where the slope is easier. It is circular in form, the plat on the summit measuring some 31 feet in diameter, and has been cut off from the base- court by a trench some 25 feet in width, 12 feet in depth from the summit, and 5 feet below GLEXCAiRX.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [glencairn. the highest point of the counterscarp, which curves around it, opening on the natural scarp of the hillock at either side. There is no indi¬ cation of a parapet mound on the summit. The base-court stretches to the eastward for a distance of about 213 feet from the edge of the trench, with a width of 95 feet at the west-south-west end and of 36 feet at the opposite extremity. The ditch which termi¬ nates it, cut athwart the hillock, is V-shaped, 36 feet wide, ii feet in depth below the scarp and 6 feet below the counterscarp. The Fig. 66a.— Mote, Ingleston (No. 238). height of the summit of the base-court is some 30 feet above ground level, and its sides are sloped at an angle of about 38°. From the base-court to the south of the mote a road¬ way 10 feet wide, between parallel mounds, leads downwards to a squarish platform at the foot of the slope with a slight mound at its south-w^est side and a dip in the level to the north, as if the road had been deflected at right angles to pass along the end of the hillock beneath the mote itself. “ Uchtre Edzare of the Inglistoun ” served on an assize in 1482 [Bucdeuch MSS., No. 96), and Edgars of Ingleston are on record at various dates down to the early part of the 17th century. The mote is thus probably the early feudal seat of this ancient family. (Cf. Introd., p. xxv). XXX. S.E. (“ Mote and Bow Butts ”). 4 June 1912. 239. Mote, Moatland.—At the edge of the low-lying ground which marks the bottom of the Cairn Valley on the north side are a series of undulating hillocks, gradually rolling back on the higher ground to the northwards. Situated about 400 yards south by east of Moatland, now incorporated in the farm of Birkshaw, is one of these hillocks, rising to a somewhat pi'ominent head at one end, with a lower shoulder to the southward. The adja¬ cent land on the east has been low and wet, while on the west it has stretched away, with a gradual declination from the level of the shoulder. This hillock has been converted into a motehill by the excavation of a trench around three sides—the east flank alone being left untrenched, the soft nature of the ground in this direction having probably afforded sufficient protection. The scarp of the trench on the north is about 10 feet high ; on the west, at highest, about ii feet and on the south about 6 feet. Along the west, where the ground on both sides does not differ greatly in elevation, the trench is V-shaped, very boldl}' cut and 34 feet in width, and has a slight mound on the counterscarp, from the top of which it measures 8 feet in depth. At the south end the trench is only 22 feet in width. The prominent end of the hillock rises at the north to a height of 24 feet and measures 34 feet across its flat summit. From the base of its upper and steeper slope on the south the shoulder extends southwards for a distance of 105 feet, with a breadth of 95 feet or 105 feet, according to whether the edge of the enceinte was the side of the plateau or the line of a dyke that runs along the flank. xl. S.W. 5 June 1912. 240. “ Shancastle Doon.”—This is a promi¬ nent hill on the north side of the Cairn Valley, about 2 miles to the east of Moniaive, rising to an elevation of 815 feet over sea-level. The name suggests the existence of a fort, of which there are indications, though these are by no means definite. Along the edge of the plateau, in certain places, there are visible large laid blocks of stone, while the scarp beneath is strewn with debris. XXX. S.E. 6 June 1912. 241, Mote, Maxwelton.—This mote has been formed out of a natural gravel ridge, rising above a low-lying meadow that stretches back from the left bank of the Cairn water, some 200 yards distant, and is to be seen by the 88 GLENCAIRN.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [glencairn. roadside some miles to the east of Moniaive. The mote rises from the centre of the ridge, and has an oval summit, with its longest axis north-north-west and south-south-east, measuring in diameter 70 feet by 60 feet, and is very level, showing no depression in the centre nor mound around the sides. Some 13 feet below the summit level on the south- south-east a trench 26 feet in width from crest to crest has been cut across the hillock, and a mound, some 3 feet in height, formed of the upcast, has been raised to the lower side. As the trench passes round to the westward it flattens to a terrace some 25 to 30 feet above the meadow below, and runs out at a slight hollow, running down the face of the hillock on the west. A line of stones across the face of this hillock suggests that it has been crossed by a wall. Beyond it the north-north-west end of the ridge is cut through some 10 feet below the summit in a manner similar to the opposite end, by a trench 16 feet wide, and now of slight depth, while 30 feet beyond is a second and slighter trench, whence the hillock gradually declines. Along the east side, where cultivation has encroached, there seems to have been a terrace, but it has almost dis¬ appeared. The counterscarps of the trenches are very stony, as if they had been faced or, more probably, surmounted with a wall, and at the west end of the defences, on the north-north-west, there is exposed what appears to be a stone base. There is no sign of a base-court. XXX. S.E. 6 June 1912. 242. Fort (supposed), Horse Park, Maxwelton. —On the north side of the road leading from Crossford to Maxwelton Mains, and some 250 yards from where it leaves the main road to Moniaive, is what appears to be a fort situated on a slight rise, covered with a young planta¬ tion, and having a small pond at either end. Around it runs the ruins of a rude wall, with a distinctly defensive character where best preserved, which is for a very short dis¬ tance on the west side. From the line of this wall eastward, the bank below has been regularly scarped to a height of some 5 feet. The thick growth of young trees makes it now impossible to discern accurately the true character of this enclosure. On the highest point there are foundations of an oblong building, probably a cottage, overgrown with thick grass. xl. N.W. (unnoted). 6 June 1912. Cairns and Hut-Circles. 243. Small Cairns, Craes Hill.—On a western spur of Craes Hill, at an elevation of some 950 feet over sea-level and about | mile north by east of Lochur farm-house, in a slight hollow to the south of a prominent outcropping rock, lie seven small cairns marked “ tumuli ” on the O.S. map. With one exception, a cairn situated towards the east side of the group, they are larger than the usual cairns of this class : two, oval in form, measure respectively 15 feet by 12 feet and 16 feet by 12 feet ; while a third of the same form, but disturbed to some extent, measures 26 feet by 19 feet. An oblong one measures 20 feet by 12 feet, and two circular cairns 16 feet and 25 feet in diameter. Those that are not circular show no preference for any particular orientation. None of them is above 2 feet in elevation. Enclosing the upper end of the hollow at the base of the rock, in a semicircle, are the ruins of an ancient wall, along the chord of which, or within it, most of the cairns lie. Close to the wall, near some clumps of rushes at the north end, there is just traceable a foundation running parallel, some 20 feet within it, for a distance of some 20 feet. About I mile due east of the last, just above the 900 feet elevation and near the source of a burn, is another group of small cairn-like mounds, most of which show that hollow in the centre which suggests the ruin of a small hut. One to the west side of the group is oblong in form, measuring interiorly 14 feet by 7 feet, and evidently the site of a shieling bothy : some 40 feet to the east of it is a small circular construction, evidently a hut ruin, with an interior diameter, as far as ascertainable, of about 5 feet. There appears to have been an entrance from the west, on either side of which the wall has projected somewhat, forming a passage 6 feet in length—the regular breadth of the wall in its ruined state, and as far as ascertainable without excavation, being about 4 feet. The circular mounds measure about 14 feet in diameter and 2 feet in elevation. A sliort distance to the south, against a rock, 8q GLENXAIRN.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [glencairn. is a large enclosure for sheep or cattle, suggest¬ ing that here have been shielings. On the moor, about | mile north-east of Lochur farm-house, and J mile south-east of the small cairns first described on Craes Hill, are the remains of a cairn, much dilapi¬ dated, with a diameter of 37 feet and an elevation of about 2 feet. There is no indica¬ tion that the interment has been disturbed. Within 30 yards to the south-west is a group of four small grass-grown cairns with diameters each of about 12 feet. About 150 yards south by west of the large cairn is another group of about half a dozen, of similar size to the last. xxxix. S.W. 31 May 1912. 244. Small Cairns, Holmhead Hill.—On the south-west flank of Holmhead Hill and about I mile east by north of Dalwhat, is a group of four or five small cairns situated, like so many in this region, at an elevation of 800 feet above sea-level, and near the source of a small stream. Two or three of them have been dug into in the centre, but no cist is exposed in any one. The largest is stony all over and measures 20 feet in diameter, with an elevation from the lowest side of about 2 feet. XXX. N.W. 3 June 1912. 245. Small Cairn, Heathery Plantation.—Just to the south of the extreme south-east point of the Heathery Plantation, and about 30 yards back from the dyke, is a small cairn-like mound overgrown with turf, which measures about 12 feet in diameter. xxxix. N.W. (unnoted). 4 June 1912. 246. Small Cairns, Girharrow.—On the north slope of the Glen of the Girharrow Burn, and some 250 yards south of the most south¬ westerly point of the Ellrig Plantation, is a group of small cairns. They lie in a slight sheltering hollow near the source of a stream at an elevation of about 900 feet over sea- level. In number they extend to about half a dozen, but formerly there have been a few others, from which the stones have been en¬ tirely carted away. The average diameter is about 16 feet. • About 100 yards down from the low side of the Ellrig Plantation, near its west side, is another small group. Over a small hillock to the east there may be seen in a hollow the ruins of an ancient sheepfold rectangular in form. xxxix. N.W. (unnoted). 4 June 1912. 247. Small Cairns and Hut-Circles, Gir¬ harrow.—On the south side of the glen, about J mile south-south-east of the east point of the Ellrig Wood and on the end of one of the many round-backed undulating ridges which trend down the hillside, is a large group of small cairns with a circular sheepfold occupying the highest point. An examination of the wall of the fold shows that it has been erected on an older stone foundation, 5 feet in breadth, probably belonging to a hut-circle which has measured interiorly some 29 feet by 25 feet. Traceable down the hillside to the eastward is the line of an ancient wall, now greatly dilapidated, against which lie the ruins of several small circular or sub-oval enclosures ; while some 40 feet to the east of the sheepfold is an oblong enclosure, measuring interiorly some 14 feet by 5 feet 6 inches, surrounded by a mere outline of stones. Beyond the wall to the eastward a rivulet flows down a hollow, and ruins of walls crossing its course may indicate the former existence of a mill of some sort, though the water-power is now very slight. The small cairns in this group number about a score and are situated at an elevation between 800 and 900 feet above sea-level. At about the same elevation, from 200 to 300 yards to the eastward of the last group, in a slight hollow and south-east of the east point of the Ellrig Plantation, is a large group of small cairns, some of which are of considerable size. The most prominent towards the east end of the group is overgrown with heather and measures in diameter some 22 feet by 18 feet and in elevation 3 feet. Near it a confused mass of stones showing some struc¬ tural work is evidently the ruins of an oblong hut, measuring over all 20 feet by 15 feet. Towards the lower end of the hollow is a cairn measuring some 17 feet across, the centre of which has been partially excavated to the extent of exposing a slab at least 10 inches thick and about 2\ feet square, which may be the cover of a cist. 90 GLENCAIRN.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [glencairn. In a sheltering hollow, about mile east of Girharrow, and at an elevation of about 800 feet above sea-level, is another group of about a dozen small cairns of less diameter than the previous lot, having dimensions of from 12 to 14 feet. xxxix. N.E. (unnoted). 4 June 1912. 248. Small Cairns, Crossford Hill. — In a slight hollow on very rough stony ground on Crossford Hill, about | mile north-north-east of Crossford, is a group of about a dozen small cairns, which measure in diameter from 10 to 12 feet and in elevation about i foot. xl. N.W. 6 June 1912. 249. Long Cairn, “ White Cairn,” Fleuch- larg.—In arable land some 200 yards west by south of Fleuchlarg farm-house are the remains of a long cairn known as the " White Cairn.” It lies with its longest axis north- north-east and south-south-west and has ex¬ tended to a length of 140 feet, with a width of some 30 feet at the north end and 80 feet at the south. It has been greatly demolished from the north end, but for the last 50 feet southward it is still a massive construction, with an elevation of 12 feet. xl. N.W. 5 June 1912. ^Miscellaneous. 250. Cross-Shaft (portion of), Hastings Hall, Moniaive.—Hastings Hall is situated at the west end of the village of Moniaive, and there, set up in a rockery, is a portion of a cross, consisting of the shaft and lower arm, believed to have been brought from near Stroan- freggan, in the Stewartry of Kirkcudbright, many years since. The stone, which is of sandstone, rises 5 feet 9 inches in height above ground, is 6| inches in thickness, 18 inches in breadth at the base, 17 inches at the head of the shaft, and i2| inches at the top where fractured. At the upper end of the shaft is sunk a panel, measuring 22 inches by 12 inches, on which are rudely carved two figures facing each other and probably intended to represent the Annuncia¬ tion. The arm of the cross is carved in low relief, and extends from the fractured top of the stone for 15 inches, and expands from 4| to ii inches, the edges of the stone running parallel to the sides. At each edge of the stone, opposite the lower extremity of the arm, there is a fracture showing a slight shoulder, indicating that some curved projection has been broken off, while a similar curved point at the top, on the right side of the stone, evidently marks the springing of the side arm. XXX. S.W. 6 June 1912. 251. Martyrs’ Monument, Ingleston.—In the northern portion of the garden at Ingleston farm-house, which has been allowed to revert to grass, there stands against the wall a slab, inscribed in incised lettering :— IN THIS YARD WERE SHOT JOHN GIBSON, JAMES BENNOCH, ROBERT EDGAR, ROBERT MITCHELL, AND ROBERT GRIERSON APRIL 28 1685 BY COLORELL (sfc) DOUGLAS AND LIVINGSTOUNS DRAGOONS FOR ADHEREING TO CHRISTS KINGLY GOVERMNENT (stc) IN HIS CHURCH AGAINST TYRANNIE, PERJURIE, AND PRELACIE. XXX. S.E. 4 June 1912. .252. “ Mercat ” Cross, Moniaive.—The cross is a slender column, ii inches square and 8 feet long, with chamfered edges and is set on a drum of masonry, 6 feet in diameter, and 4 feet 6 inches high. It supports a cap 12 inches square, with chamfered corners, dated on one face 1638, and surmounted by a large stone ball, from which issues a formidable spike, the remains of a weather vane. XXX. S.E. 30 ]May 1912. 253. Cross-Slab, Woodlea.—Here set up in the rock garden is a whinstone slab, brought from Auchencheyne, whereon is incised a plain Latin cross i foot 9 inches in height, the side arms of which slightly expand outwards, and which has an incised boss in the centre. The shaft has no horizontal groove to close it at the foot, in this respect resembling one of the crosses at Laggangarn.* There is an 18th- century inscription at the base of the stone, indicating that it has been used as a boundary mark. It is fully described and illustrated in * See Inventory of the Monuments of Galloway, vol. i. (County of Wigtown), No. 282. GLEN-CAIRN.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [GRETNA. the Proceedings of the Society of Antiquaries, vol. xlvi. p. 262 (illus.). xxxix. N.W. 31 May 1912. 254. Mound, Birkshaw.—Some 60 yards north-north-west of Birkshaw is an oblong gravel mound surmounted with trees and rising some ii feet at the south-south-east end and 15 feet or thereby at the north-north-west, in which direction the ground declines. The summit is ellipsoidal in form and has been surrounded at the edge by a stone wall, the foundations of which are clearly discernible, enclosing an area measuring 53 feet by 14 feet. There is no sign of a trench at the base, and it is difficult to say what this has been and to what period it belongs. xl. S.W. (“ Tumulus ”). 5 June 1912. Sites. 255. Cairn, Moniaive. —The cairn about J mile south of Moniaive may be noted as a site, for it has been almost entirely cleared away. A cinerary urn was found within {Annals of Glencairn, p. i). XXX. S.E. 4 June 1912. 256. Cairn, Old Crawfordton.—Some 200 yards to the north-east of the farm-house of Old Crawfordton is the site of a large circular cairn, of which little more than a surrounding fringe of stones remains. xxxix. N.E. (unnoted). 4 June 1912. The O.S. maps also indicate sites as under :— 257. St Cuthbert’s Chapel, Nether Kirk¬ cudbright. XXX. S.W. 258. Chapel, about 300 yards north-west of Glenesslin School, xl. S.W. 259. Peel Tower, Crawfordton. xxx. S.E. 260. Castle, Snade. xl. S.W. GRETNA. Defensive Constructions. 261. Fort, Westhills Moss. —On the north side of the Moss rises a grassy hillock to a height of 8 to 10 feet, which is marked on the O.S. map as a fort. There are no remains of fortifications around it, but at the north-west end there lies across it a trench some 104 feet in length, 30 feet in width, and 6 feet in depth, not carried forward to the edge of the hillock on either side. Ixiii. S.E. I October 1912. 262, Fort, Douglas Farm.—This fort lies about J mile to the south-west of Douglas farm, and is situated at the edge of a bank which falls to a hollow on the south and south-west by an easy gradient for some 20 feet. It is an oval entrenchment, which lies with its longest axis north-east and south-west, measures interiorly 176 feet by 151 feet, and is sur¬ rounded by a flat-bottomed trench some 28 feet wide and 5 to 6 feet deep at greatest depth, with an earthen mound forming the counterscarp. On the north-east are the remains of a parapet mound, which elsewhere around the periphery has disappeared. Somewhat to the north of the centre the interior area is crossed by a slight hollow, possibly secondary, opening into the trench at the east end. At its west end the trench is partially filled in, and to the north of the filling its breadth is reduced to 10 feet ; but there is no break in the counter¬ scarp mound opposite to indicate an entrance. A break through the counterscarp at its east end appears to be undoubtedly secondary. Ixiii. N.E. I October 1912. Miscellaneous. 263. “ Lochmaben Stane.”^—Standing on slightly rising ground within 300 yards of the Solway, and about | of a mile from the farm buildings of Old Gretna, is a huge granite ice- borne boulder (fig. 67), measuring 18 feet 2 inches in circumference and 7 feet 6 inches in height. At 76 feet distant to the north-north- east lies a second and much smaller stone, measuring only some 3 feet in height. Pre¬ sumably these stones are the remains of a stone circle, from which all the other stones have long since disappeared. The larger boulder is a historical relic of considerable interest. From very early times it has been a meeting place for gatherings of peace or war. In 1398 Commissioners representing England and Scotland met at “ Clochmabanestane ” to 92 Ancient ami Historical Monuments — Dumfries. Fig. 67,—Lochniaben Stane (No. 263). To face p. 9 : GRETNA.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [hoddom. treat on the details of a truce.^ In this citation we have the earlier form of the name, involving the Irish, “ Cloch,” Gaelic “ Clach,” a stone. “ Stane ” at the end is therefore tautologous, added in an age when the original tongue was no longer familiar. “ Maben ” has been identified with a warrior (Mabon) named in the early Welsh poems,- but there was a Celtic deity of this name, in Welsh Mabon, Gaulish Maponos (=the great youth), some¬ times identified with Apollo. The same place was found convenient for the meetings of the English and Scottish wardens on the west side, as in the provision of 1398, “ The men of Galloway, Nithsdale, Annandale, and Crawford Muir, shall meet the Wardens of the West March for redress of claims at Clochmabanestane ” ; ^ and Wharton’s refer¬ ence to it in the i6th century ; “ Loughmaben Stone standyng in Scotland, wher we have beyn accustomyd to keipe days of marches.” * Indeed, as marking the northern termination of the most useful ford on the Esk—the Sulwath or ” muddy ford ” ®—it was a re¬ cognised place of public assembly on the western border. 1 Foedera, hi. (4), p. 152 ; ^ see Excursus on ” The Lochmaben Stone,” in Sir Herbert Maxwell’s Dumfries and Galloway, pp. 132- 135 ; ^ Bain’s Calendar of Documents, iv. p. 109 ; 4 Hamilton Papers, ii. p. 281 ; ® see Neilson’s Annals of the Solway. Ixiv. N.W. I October 1912. 264, Grave-slab, Gretna Green Churchyard. —Some 80 feet to the south of the east end of the church lies a slab, dated 1650, bearing Johnstone arms and an inscription for the most part illegible, but which commences :— SAD OUR OF DEATH REWARD OF SIN BOTH GOOD AND BAD DRINKETH THEREIN WITNES JOHNE JOHNSTONE HEREIN UNDER SCRIPTED WHOSE BONES ETC. 265. Coped Stone, Gretna Green Church¬ yard. —Adjacent lies an early coped stone, on the flattened ridge of which are incised two parallel lines, but no sign of a cross appears at the end of them. Ixiv. N.W. I October 1912. 266. Roman Altar, Westhills.—Standing in a field directly to the north of Westhills farm buildings, and some 80 feet distant from the back wall of the farm house, is a Roman altar. Its extreme height above ground is 3 feet 9 inches, 3 feet 2 inches being the portion applic¬ able to the “ die,” and 7 inches to the cornice or capital : its breadth 2 feet 3 inches across the capital, and i foot ii inches across the die, while the thickness of the latter is i foot 8 inches. The focus, which is regularly formed with a transverse hollow on either side, measures i foot in diameter. The altar bears no inscription, but on one face is an incised figure, with three tongues, somewhat resem¬ bling a bench mark beneath what appear to be the letters L L placed back to back, and separated by a square compartment contain¬ ing a cross. These markings are probably modern. Ixiii. S.E. (“Stone”), i October 1912. Sites. The O.S. maps indicate sites as under :— 267. Red Kirk, near Red Kirk Point. Ixiii. S.E. 268. Tower, Westhills. Ixiii. S.E. 269. Stonehouse Tower, ^ of a mile north¬ west of Stonehouse. Ixiii. N.E. 270. Fort, Gretna Hill. Ixiv. N.W. HODDOM. Ecclesiastical Structure. 271. Church Foundations, Hoddom Bridge. —The present parish of Hoddom was con¬ stituted in 1609 out of the three older parishes of Hoddom, Luce, and Ecclefechan, when a church was erected in a central position to serve the new parochial unit. At Hoddom a fragment of wall of the old church remained in the graveyard, and in the course of 1915 excavation brought to light the greater part of the foundations and some three or four courses of the masonry at the east end of the nave. The building thus disclosed shows rectangular nave and chancel with some indications of a structure leading out from the north side at the junction of nave and 93 HODDOM.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [hoddom. chancel. The chancel is apparently square ended both externally and internally and measures 15 feet 2 inches in length by ii feet 8 inches within walls varying from 2 feet 8 inches to 2 feet 4 inches above the scarcement 4 inches wide, which is continued along the side walls of both apartments. The chancel opening measures 8 feet 2 inches, but there is no sign of shafting or other decorative feature. The foundations of the nave have been followed for about 38 feet in walls 3 feet 5 inches thick above the scarcement, within Fig. 68. —Plan of Churchyard and Church Foundations, Hoddom Bridge (No. 271). which the chamber is 18 feet 6 inches broad. On the line of the north wall the foundations of three buttresses have been tmcovered, all with a projection of 18 inches. One on the chancel wall, 10 feet 6 inches from the east corner, is i foot g inches wide; another, on the nave wall, some 7 feet from its eastern corner, is I foot 6 inches wide ; while the last, 24 feet from the same corner, has a width of 3 feet. The presence of two adjoining rows of square stones, the outer being of four blocks about 2 feet square, at the western extremity of the north wall cannot yet be accounted for, the outer row being off the line of the main wall. The remarkable feature of this church is that, to judge from what remains, it must have been built out of the ruins of the build¬ ings at Birrens Station (No. 462), some three miles distant. The foundations of the side walls rest upon two longitudinal rows of chan¬ nelled stones of square section, which are from the gutters of the camp, while other stones show diagonal tooling and Roman dressing. More particularly at the east corner of the south wall, where it joins the cross - wall of the chancel, a stone is bonded into the latter which bears, within a cartouche, the inscription (upside down) LEG-Ui-vi, obviously a relic of Birrens (see No. 462). Amongst the debris was found a voussoir from the archivolt of a post-Reformation window; the ingoing is splayed exteriorly and checked internally to receive a stout wooden window case. Another important discovery was a few fragments of much corroded glass, some of which, however, were of sufficient size to show the brown pigment and cross-hatching on greenish grisaille similar to the larger fragments found at Coldingham Priory, Ber¬ wickshire, which are illustrated and described in the volume on that county (Frontispiece and No. 74). Such glass is of the end of the 13th or beginning of the 14th century, and the construction of the building at Hoddom cannot, therefore, be later than that period. A “ piece of painted glass ” is recorded to have been found on the same site in or shortly before the year 1789 [Riddell MSS., vol. vi. fo. 269). Hoddom (Hodelme) was a Mensal Church of the bishopric of Glasgow from the latter half of the 12th century [Reg. Epis. Glasg., i. No. 26). Ivii. N.E. 15 September 1915. Defensive Constructions. 272, Fortifications, Birrenswark or Burns- work,*—The table-like top of Burnswork,rising as it does to a height of nearly 940 feet above sea-level, is a conspicuous object in the land- * This is the best authenticated form of the name. The current spelling, “ Birrenswark,” is a modification of “Birrenswork,” which again seems to have arisen from some confusion with “ Birrens,” and to have been an invention of William Maitland {Hist, of Scotland, p. 192). But it is used, independently^of Maitland, by General Roy. 94 HODDOM.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [HODDOM. scape as viewed from many points both north and south of the Solway. The remarkable group of fortifications of which it is the centre was first noticed by Alexander Gordon in his Itinerarium Septentrionale, published in 1727. Gordon writes only of the two large enclosures which cling to the' northern and southern slopes respectively. These he regards as con¬ nected by a rampart which ran round the eastern side of the hill, and so as forming a unit, which he describes as the “ most entire and best preserved ” Roman encampment he has ever seen. The less conspicuous remains on the summit would appear to have escaped his eye, although it was not long before they caught the attention of other antiquaries who had been attracted to the spot by the warmth of his eulogium. A convenient sum¬ mary—virtually a bibliography—of subse¬ quent references to the site was given by Dr Christison on pp. 201-210 of the Proceedings of the Society of Antiquaries of Scotland, vol. xxxiii., which also contains (pp. 219-243) a report by Mr James Barbour on the excava¬ tions carried out on the spot for the Society in 1898, as well as a notice (pp. 243-249) by Dr Joseph Anderson of the various objects then found. More recently Professor A. Schulten of Erlangen, in an article entitled “ Birrenswark : Ein britannisches Numantia ” {Neue fahr- bucher fiir das klassische Altertum, 1914, pp. 607-617) has recorded the impressions left on him by a visit to the hill in 1913. Gordon, on the strength of a supposed resemblance to entrenchments of the Flavian era, as described by Josephus, believed that the “ incampment ” which he discovered had been constructed by Agricola. Roy, on the other hand, while not doubting the Roman origin of the fortifications, was convinced that Agricola could not possibly have been their author. Towards the end of the i8th century a new and interesting hypothesis was put forward by the writer who deals with the parish of Tundergarth in Sir John Sinclair’s Statistical Account of Scotland (vol. xix. p. 446). He considered that the two large enclosures “ were at first formed by the Romans besieging a body of the ancient Britons, who had occupied the summit of the hill.” This view was maintained at greater length in an anonymous paper published in 1792 in vol. i. of Archceologia Scotica (pp. 124 ff.). It was whole-heartedly adopted by Dr Christison in the Proceedings of the Society of Antiquaries of Scotland, vol. xxxiii. (pp. 210-218), where a practically complete system of circumvallation appears upon the plan (plate hi.), while it has since found an enthusiastic advocate in Professor Schulten, as is indeed plain from the title of his paper. The siege theory has unquestionably a certain plausibility. On closer examination, however, it proves to be beset with difficulties of a practical kind. To begin with, the surface appearances make it very doubtful whether the so-called lines of circumvallation, with their subsidiary works, are lines of circum¬ vallation at all. At the best, it has never been claimed that the hill was completely sur¬ rounded. But, even if we accept the ex¬ planation that the section on the west, which is open, was effectually cut off from the out¬ side world by a marsh, there is no such organic connection between the " lines of circumvallation ” and the large enclosures as is suggested by the plan on which the argu¬ ments of Christison and Schulten rest [Pro¬ ceedings, vol. xxxiii. plate III.). Thus, at the north-eastern corner of the more northerly enclosure the low rampart, which is still distinctly visible approaching from the east, does not join up with the main defence in the manner they suppose ; it pursues its course down the hill, leaving a clear interval between itself and the camp, as shown at ” F ” on the amended plan which is here reproduced (fig. 70). Further, as a system of circumvallation, the lines and works would present some curious features. It will be noted that, while the forts and redoubts would, for the most part, lie on the inner side of the lines as if their main purpose had been defence against attack from without, the most formidable of all, the great enclosure on the south, as well as what is called on the Proceedings plan the ” East Enclosure, D,” would lie half on the inner and half on the outer side, as if in doubt from which of the two quarters most danger was to be appre¬ hended. Finally, except on the assumption that, man for man, the defenders of Burns- work were far better armed and far more formidable at close quarters than their op¬ ponents, it is not easy to believe that the \ory restricted area afforded by the hill-top 95 HODDOM.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [hoddom. 96 Fig. 70.—Birrenswark (No. 272). HODDOM.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [HODDOM. could have harboured a sufficiently large garrison to render such elaborate siege-works necessary. It is, of course, possible that the difficulties to which attention has been drawn may admit of an explanation that is not incompatible with the theory of a siege. Further, if that theory be rejected, we are at once confronted with a different set of difficulties that seem equally insoluble, unless we regard the large enclosures on north and south as belonging to different periods. If they were con¬ temporaneous, what motive could their de¬ signer have had for thus dividing his forces, except the desire to hem an enemy in ? And why, if the hill-top were friendly ground, should it have been necessary to make the entrances that faced towards it specially strong ? The truth is that without further excavation it would be unwise to express a definite opinion. In the meantime there is only one thing that can be said with safety. Though the grounds on which his view is based are fanci¬ ful, Alexander Gordon was in all probability right in associating the large enclosures with the first Roman invasion of Scotland. Scanty as were the relics recovered in i8g8, their testimony was fairly conclusive. Besides a much-worn fragment of the tell-tale “ Samian,” they included ” sixty-seven sling bullets of lead, varying from | inch to inches in length, and from | oz. to oz. avoirdupois.” The acorn-shape which many of them showed (see fig. 71) justifies their recognition as Roman glandes. But glandes ceased to be used in the Roman army about the end of the ist century of our era, being replaced by bullets of stone. The connection with Agricola seems, therefore, pretty well established. It should be added that ex¬ amples were found not only in the large southern enclosure, but also near the summit of the hill, some of them bearing unmistakable traces of impact. The earthworks that are scattered along the flanks of the hill include remains of one or two circular structures, which were per¬ haps originally intended to shelter sheep or cattle. The most complete is that which lies within the western half of the great southern enclosure, immediately to the south of the west gate (fig. 70, B). It has a diameter of about 80 feet, and the manner in which its circumference impinges upon the rampart, which it adjoins, furnishes clear evidence that it is intrusive, a feature which an examina¬ tion of the ground renders even more obvious than the plan might suggest. If these struc¬ tures (to which not a few parallels could be cited from other parts of Dumfriesshire) be set aside as being in all likelihood mediaeval at the earliest, there remain at least three distinct types of fortification, each of which is well represented in the Burnswork group. The first and simplest (Type I.) consists of a low rampart with a more or less rudimentary ditch in front of it, and with no very special defences where gaps have been left for en¬ trance. It is not marked by characteristics pointing to any particular period or to any particular race of builders. In Type II. the rampart is much higher and the ditch corre¬ spondingly deeper, while the entrances are protected by devices like the titulns and the claviciila, which we are accustomed to associate with the art of military engineering as prac¬ tised by the Romans. Tj^pe III. is strikingly different. It shows a double rampart with a ditch intervening, and for some distance beyond the entrance the approach to the gate is flanked by a single rampart on either hand, leaving a passage some 4 feet wide. Ex¬ amples of Type III. are noted elsewhere in this volume (Nos. 393, 451, 487, 596). The hill is precipitous on the east side (“ The Fairy Craig ”), but more or less easily approachable on each of the three others, the slope being gentlest on the south. Its summit, which presents a surface by no means Fig. 71. —Roman Glandes from Birrenswark (No. 272). 97 HODDOM.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [hoddom. so uniforml}' flat as distant views might lead one to anticipate, is surrounded by a rampart (T^’pe 1.) which follows the sinuosities of the edge. The extreme length of the enclosed “ plateau ” is about 1700 feet from east to west, while its extreme breadth from north to south is about 700 feet in the western portion and about 400 feet in the eastern. The superficial area is about 17 acres. This rampart would appear to have had no ditch. Three entrances * are more or less distinctly visible—one on the north and two on the south, the more westerly of the latter being obviously the most important of the three. The examination made in 1898 seems to have revealed the existence of a second rampart on certain parts of the northern front. But it cannot be said that either this feature or the system by which the gates are supposed to have been specially protected is obvious to the ordinary eye to¬ day, though both are here reproduced on the plan. Possibly the condition of the vegeta¬ tion at the date of visit (April 1915) was un¬ favourable for observation. It should be added that Mr Barbour was of opinion that, on the south side at least, the gates had been stone-built structures. The evidence is fully set forth in his report {Proc., xxxiii. p. 240). The broad or western end of the hill-fort is separated from the remainder by a low mound running almost due north and south. Within the space thus cut off, and hard by the supposed western gateway of the fort, a small heart-shaped enclosure (fig. 70, D) is clearly discernible. The plan, in Mr Barbour’s words, “ when carefully surveyed, is found to be symmetrical, and resembles a horse¬ shoe, pointed at the north, and with a base at the south composed of two straight lines, which retire towards the centre and form a re-entering angle.” The extreme length over all from north to south is about 130 feet, and the extreme breadth is not substantially different. Excavation showed that the en¬ closing mound, which is about 12 feet broad and stands 3J feet high, had been much more elaborately constructed than the rampart round the plateau. There were facing-walls * A fourth entrance, shown on the west on the plan in the Proc. Soc. Ant. Scot., vol. xxxiii. p. 232, is very doubtfully ancient. at each side, and at least one of the stones that were uncovered bore marks of the mason’s pick. The entrance to the enclosure was on the west, and it had been about 4 feet wide. Some 400 feet east of the dividing mound, and 80 or 90 feet south of the north rampart, are the remains of what is presumably a bronze-age cairn (fig. 70, E), having a dia¬ meter of about 70 feet. The top stands fully 10 feet above the level of the circum¬ ference. This height is, however, partly due to the configuration of the ground ; at the centre the artificial work is only 5^ feet above the natural level. The cairn was opened by Mr Barbour in 1898, but proved to have been rifled. It contained the remains of a’ cist. About 300 feet farther east the plateau is again crossed by a low mound running from north to south. It was sectioned in 1898, when it was found to consist of soft earth and to mark the line of a deep drain. The great southern enclosure is the most impressive unit in the group. In outline it is approximately rectangular, with rounded corners, the interior dimensions averaging 840 feet by 580 feet. A glance at the plan will reveal certain irregularities of shape which it is not necessary to describe in detail. The defences are of Type IT On the north, looking towards the hill, rampart and ditch, taken together, vary in breadth from 42 to 58 feet. Clearly this was the side from which there was most apprehension of a sudden attack. Elsewhere the correspond¬ ing limits of variation are 30 feet and 35 feet. The rampart, when sectioned, was found to have been constructed of earth, with a bonding of branches or brushwood, and to have been faced with a pitching of stones on its outer side. It rested on a bed of finer earth, 9 inches thick, above which was a layer of clay from i J to 3I inches deep. At a point where, judged by the condition of the pitching, it appeared to retain something like its original height and form, the highest part is ii feet above the bottom of the ditch immediately in front of it, and ii| feet above the level of a pavement, some 5 feet wide, which ran round the inner margin of the whole enclosure. The practical identity of level is accounted for by the fact that the section which yielded these measure¬ ments was on the north side, where the upward slope of the ground is fairly steep. The ditch 98 Ancient and Historical Monuments — Dumfries. Photograph by Mr J. Crais Annan, Glasgow, P'lG. 72.—Birrenswark : South Camp (No. 272). To face p. 99. HODDOM.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [hoddom. (it should be added), when opened, was seen to be V-shaped, and to be cut 2 feet into the solid rock. The entrances are six in number, three of them being on the north and one on each of the other sides. The three on the north are respectively 35, 50, and 35 feet wide, and each of them is covered by a circular mound, some 60 feet in diameter (fig. 72). The central of these " Three Brethren,” to give them their local name, was carefully examined in 1898. It had been piled up on a roughly laid founda¬ tion of stones, and had also been faced all round with a pitching of stone. Apparently it is still of the full original height, the top being loj feet above the bottom of the V-shaped ditch that surrounds it, and 12 feet above the level of the ground at the entrance which it defends. The apparent anomaly involved in these figures is again to be ac¬ counted for by the natural slope of the ground. The ” Brethren ” on east and west of it are very much the same in character. The three remaining gates were protected in somewhat similar fashion, although here, too, the de¬ fensive mounds represent a departure from the normal form of titidus. They are oval in shape, and about 50 feet long, and have been defended in front and at the sides by a well- cut V-shaped ditch. They had been stone- pitched all round like the " Brethren,” The greater portion of the enclosure to¬ wards the south is divided into two roughly equal parts by a stream which has its source in a spring in the interior. Although the bed of this is now fairly deep, the sides of the cutting reveal no definite traces, structural or other, of human inhabitation. The ex¬ cavations of 1898, however, are said to have disclosed indications of ” walling, pavement, and debris, evidently of large and important buildings.” The small circular earthwork close to the west gate (D) has already been noticed as being in all probability comparatively recent. On the other hand, the rectangular ‘‘ redoubt ” which occupies the north-east corner (C) is unquestionably ancient, more ancient indeed than the large enclosure itself. It is rect¬ angular, with rounded corners, and measures about 100 feet from north to south and about 70 feet from east to west (fig. 73). The de¬ fences have been of Type III. The double rampart with intervening ditch is still well marked for the greater part of the circuit, but on the east as well as on the north the original design has been seriously interfered /NTE^/OH' or 'RLDOU5T. Fig. 73.—Birrenswark (No. 272). with by the construction of the later and much larger enclosure. On the north side, in particular, that part of the intervening ditch which is adjacent to the north-east 99 HODDOM.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [HODDOM. corner has been completely filled up. As the outer rampart is at this point merged in the rampart of the main enclosure, the effect is to produce a broad platform immediately to the rear of the latter. Incidentally the result has been to bur}^ completely the old entrance to the “ redoubt.” The fact that in 1898 considerable remains of two low stone walls that had flanked this entrance were found in the heart of the stone-pitched rampart, is clear proof that the ‘‘ redoubt ” is an earlier structure than the main enclosure. The passage between the walls had been 4 feet 8 inches wide. Within the “redoubt ” itself the traces of a still smaller enclosure are distinctly visible. This and the various features, including post-holes, revealed by the excavation are shown on the plan in the Proc. Soc. Ant. Scot., xxxiii. p. 229, reproduced in fig- 73- The great northern enclosure, as shown on the plan, which is based solely on the present surface-appearances, is much more irregular in form than its southern counter¬ part. The south, east, and west sides are approximately straight, while ” the line of the north side, starting from the north-east comer, runs westwards nearly halfway, where there is a southward set-off, diminishing the vddth of the westward area, and giving the appearance of two camps of ordinary pro¬ portions but different widths conjoined.” The comers are rounded, and the length of the enclosure over all, from east to west, is about 1000 feet, while the breadth from north to south is about 400 feet at the eastern end and about 300 feet at the western. It may be that the peculiar shape is due to the marsh that lies opposite the western portion of the northern face, coupled with the fact that there is a natural ridge which the defences here appear to follow. But, pending further excavation, it would be unwise to exclude the possibility that the remains now visible may really represent a camp of quite regular outline and measuring 500 feet by 400 feet, flanked by an annexe on the west. The defences are fairly complete on the southern and eastern sides, and there they conform to Type 11 . Although they are largely obliter¬ ated on the western side and on the western half of the northern side, the same type would seem to have been used for these portions also. On the other hand, on the eastern half of the northern side it is curious to encounter Type III.—the double rampart with inter¬ vening ditch, which we have seen to be probably earlier than Type IT Viewing the whole as a single enclosure, Mr Barbour has given it six gates. T\vo of these, how¬ ever, are conjectural. Of the remainder the three lying nearest to the east are protected by oval mounds which may be regarded as tituli, while at the fourth, which looks towards that part of the northern face of the hill where ascent or descent would be easiest, the rampart turns outwards on the right hand of the entrance and projects sufficiently far to cover the gate completely, thus re¬ minding one forcibly of the characteristically Roman device of the clavicula which, how¬ ever, was more usually on the left-hand side of an attacking party. It should be added that in 1898 “ the excavation in the interior was very limited, and so far little evidence of occupation was discovered.” No relics were found except a stone ball and two pieces of iron. It is difficult to give any clear description of the “ lines of circumvallation.” They are most distinctly traceable towards the base of the hill on the north-east where a low mound is visible for a considerable distance. As already indicated, however, this mound does not join the north-eastern corner of the northern enclosure, but passes it by without making connection at all. There is a small ditch in front of it. The earthwork is, therefore, of Type L, and, if intended as a fortification, must have been constructed, not to hem in an enemy on the hill-top, but as a defence against attack from outside. To the west of the great southern enclosure there are again indications of the former existence of a line of earthworks. A section cut in 1898 suggested that here the line had been of Type III., and the probability that this was so is strengthened by the proximity of a remarkable enclosure, marked “ A ” on plan, which is de¬ fended by a double rampart, with intervening ditch, and approached on the west by a narrow passage, 60 feet long and 4 feet wide, flanked on either side by a prolongation of the outer rampart of the enclosure. In shape it " is triangular, the base being at the south and rectilinear, but with a knee towards the west 100 .Air ic il ait.i His/ 07 -ical Monunicuts — Dumfries. Tojacep. loi Fig. 75.—Fragments of Cellic Cross from Knockhill (No. 273). HODDOM.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [HODDOM. end, while the sides are unequal and show an outward curve.” There is an opening on the east side, but it is more than doubtful whether this was ever an entrance. The interior area is hollowed out so that the bottom is below the level of the ground outside. The only relics found here in 1898 were a broken quern and a portion of a glass armlet. In addition to the objects discovered during the excavations of 1898 and described in the Proc. Soc. Ant. Scot., vol. xxxiii. pp. 243-249, mention may be made of the coins of Nero (54-68 A.D.), Vespasian (69-79), Trajan (98-117), picked up about 1727 (see Proc. Soc. Ant. Scot., lii. pp. 216-7), and also of a bridle- bit of native origin (fig. 74), from a moss in the pieces had formed part of an elaborate Celtic cross, as in fig. 75. This was no doubt the " carved stone cross ” which, in 1789, Mr Sharpe, the proprietor of Hoddom, was pro¬ posing to transfer to the castle from the site of the old church.* The pieces had been split lengthways so as to give the summer¬ house wall the advantage of both scupl- tured faces. That one of the upper slabs was also split across in being so treated was of course an unfortunate accident. As readjusted they show the central portion and left arm of the cross. On each face is a central medallion, about 9 inches in diameter within the sunk surface, enclosing a figure of “ Christ in Majesty,” but on different scales. Fig. 74.—Bridle-bit from Birrenswark (No. 272). neighbourhood, which was found before 1785 and is now in the National Museum. The bit is a most skilful single casting in bronze and bears an enamel ornamentation in red and yellow, the article having been designed and finished with all the care propei to a piece of jewellery. It has been much used and in some parts repaired. li. S.E. 15 April 1915. Miscellaneous. 273. Cross Fragments and Inscribed Stones, Knockhill.—Built into the wall of a summer¬ house in the grounds of Knockhill House were, until lately, several fragments of the Celtic cross type and two inscribed stones apparently from the Roman station at Birrens. They have now been released from this position by E. J. Brook, Esq., of Hoddom, and, what¬ ever their final destination, are meantime carefully preserved. On removal it was found that one set of The open book in the left hand is more clearly indicated on the obverse, and the index finger of the other hand points to its pages ; on the reverse the right hand is raised in the sign of benediction. Both figures are in tunic and heavy mantle (Greek pallium), and each bears a nimbus on the head, on which the hair is long and seems plaited or curled, but is carried to the back. Unlike the Ruthwell Cross, the nimbus is not distinguished by a cross within the circle. Comparison with figure details on the Ruthwell Cross will show how this treatment of the hair is a male characteristic, the long hair of the women hanging down in front in two pleats. The margin of the medallion on the obverse is marked with a pearl ornament; that on the reverse is plain ; both are enclosed in the plain bead moulding that frames all the panels. On the reverse side the lower part of the figure passes be5-ond the margin. * Riddell MSS., vol. vi., p. 270. Cf. No. 271. lOI HODDOM.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [HODDOM. The inner panels of the arms contain designs of animals placed back to back. On the reverse these are the winged ox, the S5'inbol of St Luke, distorted somewhat so as to occupy the full space. It is not clear what the corresponding pair on the other face stand for. They are not winged, have their heads turned inwards, and are adjusted to a simple knot framework. In the outer panels there are, on the obverse, the upper parts of the bodies of two figures, each wear¬ ing a nimbus, of which one holds an open book, while the other, looking towards it, holds what seems to be a musical instrument ; and on the reverse the upper part of a winged figure wearing long hair and a nimbus and carrying a sceptre or floriated rod to its right. Of this figure the hair curls upwards from the shoulder. In the former case we have possibly an illustration of a citole, an instru¬ ment of the guitar type. A similar identifi¬ cation has been offered in an illumination on the title-page of a copy of the Gospels in Latin made in the 8th century for the Emperor Charlemagne, on which adoring angels and elders are equipped with instruments ap¬ parently played with the fingers or a plectrum. The hair of both figures is cut short above the ears, which are thus exposed clearly to view, while the lie of the hair and the shape of the upper surface suggest the tonsure. The figure on the reverse, apparently, as in the case of a similar figure on the coffin of St Cuthbert at Durham, stands for the archangel Gabriel or Michael. \\Tiat we may take to be the surviving section of the shaft displays on both faces the corresponding portions of two panels within a marginal roll and separated by small plain horizontal panels. On the obverse what is left of the upper panel is occupied by a foliageous pattern in circles branching from a central stem, as in the Jedburgh slab; * the pattern on one side passing behind the ring of a roundel with boss, of which only the upper part can be shown. On this side, too, are the bare feet and legs of a human figure en¬ tangled in the foliage. In the lower panel are two heads, each within a nimbus. Here again that to the left of the picture shows the long hair; that on the right is apparently short-haired. On the reverse the upper panel * Early Christian Monuments, p. 434. exhibits only the lower parts of the legs, in one case bare, and bare feet in both cases resting on roundels with a central boss, while the lower shows but one complete head and nimbus of a pair. The sides of the shaft display a complicated and highly conventionalised scroll work, with interlacing, based on the vine motive, in which the grape bunches are reduced to a mere pattern. That the stone of the shaft is of a lighter shade of colour and coarser in substance need not militate against it belonging to the above cross-head, as the Ruthwell Cross (Appendix) shows the same characteristic. The whole cross can be calculated to have been about 15 feet high, the arms extending to 3 feet, and the thickness of the remaining portion of the shaft being 6| inches. It was thus not far short of the Ruthwell Cross, though probably of more slender proportions. The shaft apparently tapered from the base to 5j inches at neck. The neck is hollow'ed in a curve which is repeated in a shallower form at the outer ends of the arms, giving a similar double curve to that on the Ruthwell Cross. An edge-roll and inner bead-mould¬ ings run round the whole cross, which is carved in relief in the Permian sandstone of the district. Of the other fragments (fig. 76), No. i is the pedestal of an altar to Fortune, and is 11 inches high by about 14 inches broad. The inscription is partly obliterated, but can be restored to read :—“ fortvnae pro salvte p. campa(ni) italici, prae(fecti) cohortis SECVNDAE TVN(GRORVM) CELER LIBERTVS, with the usual terminal l(ibens) l(vbens) m(erito) : To Fortune, Celer a freedman, for the safety of Publius Campanus, an, Italian, prefect of the second cohort of the Tungri, gladly, willingly, deservedly [dedi¬ cated this altar].” This stone came from Birrens, as also did the memorial slab to Afutianus referred to in the article on that station (No. 462). The rest of the pieces of sculptured work (fig. 77) suggest portions of perhaps three crosses, of which No. 3 is different in material —being a coarser sandstone, paler in colour— in design and workmanship. It is about 13 inches wide by 10 inches high. The design is of spirals ranged round a central boss. Nos. 102 Ancient and Historical Monuments — Dumfries. Fic. 76.—Pedestal of Roman Altar from Knockhill (No. 273). Fio. 77.— Other Sculptured Fragments from Knockhill (No. 273). Toi i. r p. 102 . HODDOM.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [HODDOM. 2 (i2 by 6 inches) and 5 (ii by 5^ inches) are the sides of cross-shafts, and show a conventionalised foliageous pattern, that on No. 5 being not unlike the pattern on the sides of the shaft in fig. 75, but with the grape pattern replaced by a rosette. On Nos. 4, 6, 8 , are various knotwork designs. On No. 7 is an Agnus Dei (10 by 8| inches), a lamb bearing a banner in relief, while No. 9 is again a portion of the horizontal arm of a cross with double curves. The outer end contains a beaded medallion within which is the figure of a lamb. With the exception of No. 3, all these pieces were of the local Permian sandstone. It is desirable to col¬ late with these frag¬ ments a portion of a cross-shaft (fig. 78) from the walls of the old church at Hoddom, which was demolished in 1815. The back has been tooled away. On the front a full- length figure clothed in the usual mantle, which leaves the right arm free, holding an open book towards the left arm, and with the usual long curled hair and nimbus, stands within an architectural canopy. The sides are formed of square pillars with square bases and capitals, on which three lines of moulding are incised. Resting on the capitals of the columns is a gable, surmounting a semicircular arch. In the front of the gable is a small round-arched opening, and the roof terminates in what seems to be a cross-patty or at least a small cross. On either side of this finial is placed a head, that to the spec¬ tator’s left showing the characteristics of the tonsure. On the right a similar figure, but only half-length, stands within a similar canopy, which, however, in this case is surmounted by a cupola with a round-arched shallow opening and finials, the character of which is not clear. On the left again the same subject is treated in much the same way. In contrast with the central figure, those on the sides are clearly tonsured, this being particularly clear in the case of that on the left. That on the right seems to wear some sort of coronal or band of ornament across the forehead. This section of a shaft is i foot 10 inches high, measures 8J inches wide at the top and 10 inches at the bottom and from 5| to 6| inches deep. The corner mouldings are slightly different from those on the other cross fragments. The Byzantinesque cupolas and square piers of the architectural elements, if contemporary, probablj^ indicate an early 12th-century date. Ivii. N.E. 15 September 1915. 274. Cross Slabs, Hoddom Churchyard. — In the course of the excavations noted on p. 93, several grave-slabs were uncovered which bore 103 HODDOM.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [holywood. incised crosses, the cross at the top of a narrow shaft being of the cross-formy or cross-patty t^'pe (fig. 7). Cross-slabs of this character in the museum at iMaidstone are attributed to the 13th century. On two, covering graves, at the extreme west end of the south chancel wall outside were a pair of shears in addition to the incised cross. In addition there are in the churchyard other two slabs worthy of notice. One, oval in outline, is surrounded by a raised border, and has carved upon it, in relief, an equal-limbed, trefoiled, hollow¬ angled cross, set on a shaft rising from a stepped base, the trefoil occupying the whole Fig. 79.—Crosses, Hoddom Churchyard (No. 274). angle from the outer extremities of the limbs. The slab measures 5 feet 9I inches in length, I foot 3 inches in greatest breadth, and tapers somewhat to the lower end. The cross with shaft and base has an extreme length of 5 feet 3 inches and a breadth across the arms of II inches. Another slab, also carved with a hollow¬ angled equal-limbed cross on a shaft, lies near. It has been broken across, but the lower end of the slab, showing the continua¬ tion of the shaft, which does not rise from a stepped base, lies beside it. Cf. fig. 165, Appendix. Other fragments are illustrated in fig. 79. A is a broken portion of a cross-head, the curvature of the arms lying on the circumfer¬ ence of a circle with a radius of 14 inches. In B the transverse arms measure 22 inches be¬ tween the extremities. Ivii. N.E. 15 September 1915. 275. Cross-socket, near Hoddom Church¬ yard.—At the side of the churchyard wall, in the field on the east, lies a massive socketed base of a cross 3 feet 8 inches in length by 2 feet 9 inches in breadth, chamfered at the angles. The socket is 7 inches square, and measures 5 inches in depth. Ivii. N.E. 2 October 1912. Sites. 276. Kirkconnel Tower. Ivii. N.E. The O.S. maps also indicate sites as under :— 277. St Fechan’s Church, Ecclefechan. Ivii. N.E. 278. Hallguards. Ivii. N.E. 279. Luce Church, north of Luce Mains. Ixvii. S.E. HOLYWOOD. Castellated and Domestic Structures. 280. Fourmerkland Tower. — Fourmerkland Tower stands 5^ miles north-west of Dumfries Fig. 80.—Fourmerkland Tower (No. 280). 104 HOLYWOOD.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [holywood. and I a mile west of Irongray Station, on marshy ground to the north of the highway. The building is complete (fig. 8o) and is kept in excellent repair by the proprietor ; the weather-beaten walls, terminating in a steeply- pitched roof with angle turrets, present a most picturesque appearance, as they emerge from the trees surrounding the site. Four SECOND noOQ. rLOOI^ ai^ouNO riooR 20 30 FEET. I .i ' J Fig. 8i.— Fourmerkland Tower (No. 280). storeys in height and oblong on plan (fig. 81) the tower measures exteriorly ig feet by 23 feet 6 inches; the upper floors are reached by a wheel-stair situated in the southern angle. The entrance is to the south-east by a lintelled doorway, which is ornamented with a simple edge-roll moulding and surmounted by a panel bearing on a shield the arms of Robert Maxwell, viz.: A saltire between a holly leaf in chief, a stag head in base, a star in dexter flank; the sinister flank is indecipherable ; above the shield, in raised characters, R 1590 M ; flanking the shield RM, and lower IG. Through the well of the stair the basement is reached, a vaulted chamber measuring internally some 12 by 17 feet and lit by a small loophole in each wall. The arrangement of the upper floors is simple ; each floor comprises one apartment, and the third floor has small chambers contained in the turrets projecting from the east and west angles. The building was in occupation until 1896. xlviii. N.E. 13 June 1912. 281. Portrack or Portract Castle.—Within the grounds of Portrack House, on the west bank of the River Nith, i mile north-north- west of Holywood, is a fragment of masonry, 12 feet high, 8 feet 6 inches long, and 4 feet thick, which formed the south-west angle of a keep, defended to the north and east by a burn and to the south and west by an artificial ditch. In 1588 Edward Maxwell of “ Tynwell ” was caution for James Maxwell of “ Porterak,” and James Maxwell of “ Portrak ” is a signatory to the General Band of 1602 {Reg. P.C., iv. p. 289 ; vi. p. 828). xli. S.W. 28 May 1912. 282. Cowhill Tower.—On an eminence 300 yards south-south-east of the Mansion of Cowhill, which lies f of a mile north-north¬ west of Holywood Station, are the ruins of an L-shaped dwelling. Only the east wall of the longer wing exists ; over the doorway within a panel is a stag’s head and neck affront^ ; on the neck is a saltire ; below, on another stone, is a shield ensigned with a stag’s head couped, bearing arms : A saltire with a holly leaf in base; at sides are the initials RM and BM for Robert Maxwell and Barbara Maxwell his wife. On the door lintel is the date 1597. A skew-put bears the date 1611— when probably there was an addition or alteration. A view of Cowhill Tower, dated 1789, is given in Grose’s Antiquities, vol. i. p. 146. The writer mentions that a few weeks after the drawing was made the proprietor started to demolish the dwell¬ ing with a view' to erecting a mansion on the site. The mansion, how'ever, w'as erected elsewiiere. xlix. N.W. 28 May 1912. 105 HOLYWOOD.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [holywood. Defensive Construction. 283. Entrenchment, Cowhill.—On the north face of a ridge lying along the right bank of the Nith, separated by a deep hollow from the plateau on which are the ruins of Cowhill Tower, is a segmental trench some VI0‘ MICH * • 5'y HIOM # 4k ► 44 HIGH ♦ 40" MiGtl- # side of the road between New Bridge and the farm of Kilness, are the remains of a very large stone circle (fig. 82). It now consists of eleven stones, of which only five remain upright. They are for the most part great m.asses of rock, whinstone, and coarse granite, with one or two boulders. The largest stone is that nearest to the gate into the field from the road on the west; it is wholly exposed, and measures 10 feet 6 inches in length, 7 feet 9 inches in greatest breadth, and from 3 feet 6 inches to 4 feet in thickness at base. On the ^ lower end, which has origin- 6 ' 0 ’HIGM been underground, are several natural cup marks. '' The highest of the upright stones—that diametrically op¬ posite the last, and the sixth from it following the course of the sun—stands some 6 feet o 3 inches in height above ground, and at 3 feet up measures in circumference 16 feet 3 inches. The stones have seemingly been placed with their flat faces in the line of the circumference. See Proc. Soc. Ant. Scot., xxviii. p. 84. xlix. S.W. 28 May igi2. stam A-b & DENOTES A STONE IN POSITION 0 10 50 ^ DENOTES A FALLEN STONE III t I H I f JOG HfEEl Fig. 82. — Stone Circle, “ Twelve Apostles,” Holywood (No. 284). 15 feet in breadth at the level of the counter¬ scarp, with a mound thrown up to the outside. It has been excavated on falling ground at the end of the ridge, so that the scarp has an eleva¬ tion of from 8 to 10 feet. It is not continued round to the south end of the ridge, xlix. N.W. (unnoted). 7 May 1913. Stone Circle. 284. Stone Circle, “Twelve Apostles,” Holy- wood.—Extending into two fields, on the north Miscellaneous. 285. Bells, Holywood Church. —Hung in the tower of the present parish church, and still in use, are two mediaeval bells inscribed :— 1. At top of waist, V [shield] k Diameter i6| inches, height 15 inches. Note B. 2. 1 WEIGH A33AS SACR ME FIERI FECIT O ADQVIGE V. Diameter 18 inches, height 13^ inches. Note A*’ . No. I is somewhat beehive shaped, and a much clumsier and rougher casting than No. 2. The surface is poor, the shoulder large in to6 HOLYWOOD.J INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [hutton and corrie. proportion to the waist and soundbow, the rims are coarse and flat, and above the sound- bow there are three ridges with intermediate depressions like indeterminate mouldings. The crown is very high. The letters are large, coarse, and flat, and are immediately below a band of four large flat rims which run round the upper part of the waist, below the shoulder. The V is 2 inches high, the k if inches, and the shield about 2 inches. The latter is charged with a chevron between three cross- crosslets fitchy, being the arms of William Kennedy, Abbot of Crossraguel and Com- mendator of Holywood about 1527. No. 2 is better cast and well proportioned, has rounded shoulders, two rims above and below inscription, three rims above and below soundbow. The lettering is small and clear, about I inch high, except the V, which is about I inch high. The initial cross is the same size as the lettering, and is a plain cross patee. The letters are of the semi-lombardic type, which was frequently used just when Roman lettering began to be introduced; but the “e” of “ weich” is small black letter, and the “ t ” of “ fecit ” is more like a small Latin cross. The b’s in “ abbas ” are exactly like the Arabic numeral 3. The final V is like a large black-letter minuscule “ v.” Apparently the final group of letters stands for A d. [m] quinge- [ntesimIo, which with v gives the date 1505. SACR is for Sacri Nemoris, “ Holy Wood.” Welch or Welsh is a name closely con¬ nected with Holywood in various capacities, but it is not possible to identify this particular abbot, unless he is the John who, as abbot, preceded the above Kennedy, c. 1522 {Bhcc. MSS., p. 70; Trans. Dumf. and Gall. Antiq. Soc., 1889). Site. The O.S. map indicates a site as under :—■ 286. Holywood Abbey, Kilncroft. xlix. N.W. HUTTON AND CORRIE. Castellated and Domestic Structures. 287. Gillesbie Tower.—The ruined walls of this building are situated on a high rocky bank overhanging the western bank of the Dryfe water, about J a mile to the north-east of Gillesbie House. The tower has been oblong on plan (fig. 83), measuring externally some 41 feet by 35 feet, with walls averaging 6 feet in thickness and some 10 feet in height at the southern end. The greater part of the east and west walls are demolished, as indicated on plan, and no features of interest remain. The tower has been well defended by the river, with its rocky bank on the eastern side; and, on the western side, at a distance of some Fig. 83.—Gillesbie Tower (No. 287). 50 feet from the exterior of the west wall, it has been secured by an artificial rampart, some 10 feet in height and from 6 to 15 feet in width at the apex, with an outer ditch averaging 25 feet in width and 9 feet in depth. At a distance of 34 feet and 18 feet respect¬ ively, measured from the north and south extremities of the tower to the crest of the rampart, the outer defences turn at right angles and have followed an easterly course till they joined the river bank, thereby completing the outworks on the landward side. Owing to the construction of a modern cart road along the west bank of the river, the north and south returns of the rampart 107 HUTTOX AND] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [CORRIE. have been cut through and partly removed, and the ditch has been filled up. The nature of the site and the earthen outworks indicate an earl}’ date, but the remains of the building are too indefinite to suggest any particular period. This was a residence of the Grahams of GiUesbie, who are credited also with a prior residence at “ Maskersa ” or Mosskesso (No. 313, Stat. Acct., vol. xiii., p. 578). " John Grame of Gillesbye ” is on record in i486, and James Graham of Gillesbie in 1531, 1555, 1560, and 1579 [Johnstone MSS., pp. 13, 15, 21, 22, 27). xxxiv. N.E. 27 August 1912. 288. Lunelly Tower.—Some 3I miles east of Lockerbie and 800 yards due south of Corrie- mains, on the north bank of the Water of Milk, is a fragment of the south wall of a tower. It is 3 feet 8 inches thick, 20 feet long, 25 feet high, and bears trace of vaulting interiorly. Of old it was called “ the Lun ” [Stat. Acct., vol. xiii., p. 578). The estate of Corrie passed from the family of that name in the 15th century (Cf. Introd., p. xxviii.). In 1516 James Johnstone received by Crown charter the land of Lund, which had been resigned by Robert, Lord Maxwell [Reg. Mag. Sig. s.d., No. 99). xliii. S.E. 24 August 1912. Defensive Constructions. 289. Fort, Castle Hill.—This fort is situated on the east side of the summit of the Castle Hill, a slight eminence in an undulating moor¬ land on the watershed between the Boreland Burn and the Dryfe Water, and about J mile to the south-east of Upper Fenton, overlooking the road from Lockerbie to Eskdalemuir. It appears to have been an oval construction, measuring interiorly from north to south some 220 feet, placed with its longest axis north and south, just where the hill begins to dip to the eastward. The ground on the west is comparatively level at either end, to north and south it falls away in a gentle decline, while to the east, for the first 30 feet, it has a steep declivity, thereafter falling to the roadway at its base in an easy gradient. Only on the west half of the oval do the defences exist, and on this arc their strength is somewhat remarkable when contrasted with their total absence now on the opposite side. They consist of a bold rampart, some 26 feet in breadth at base, which is formed of earth and shivers of rock and rises to a height of from 4 to 5 feet on the inner side with a trench in front 10 feet deep below the crest of the scarp. The trench, cut through rock, has a width from crest to crest of some 32 feet and a depth below the counterscarp of 3 to 4 feet. It has not been formed at the base of the hillock but about halfway up, so that the scarp of the mound to the outside of it has an elevation of some 8 feet above the actual base of the eminence on which the fort is situated. There are two entrances : one 10 feet wide from the south-west, carried over the trench on unexcavated ground and through the rampart; and another of similar width, ap¬ proached from the north-west by a track along the crest of the counterscarp of the trench, thence over the trench on unexcavated ground, and through the rampart at the highest level of the interior. Towards the east and lowest side of the interior are one or two artificial hollows and small excavations in the rock at the top of the steep slope, the latter possibly due to quarrying for stones at no remote date. xxxiv. N.E. 26 July 1912. 290. Fort, Range Castle Hill. — This fort occupies the summit of a grassy hill at an eleva¬ tion of 958 feet over sea-level, falling very steeply for some 400 feet to the valley of the Dryfe on the west, sloping away by an easy gradient into the moorland to the south and east, and dipping with a steeper declivity to the glen of the Murthat Burn on the north. A regular oblong eminence forms the summit of the hill, having its longest axis north-east and south-west and measuring on its plateau some 260 feet by 115 feet. All round, except towards the south-west, it is steeply scarped for a vertical height of some 20 feet, giving on a somewhat irregular terrace with an average breadth of 10 to 12 feet on the west and 20 feet on the east, faced where it crosses the slope at the north-east end by a rampart some 4 to 6 feet high, which is traceable also along the east side, where it has a ramp to the base of the eminence of some 6 to 8 feet. There is no apparent parapet mound. Towards the south- HUTTON AND] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [CORRIE. west end the defences draw in to the end of the plateau, where there has evidently been an entrance, in front of which is a large circular hollow with a diameter of 44 feet, while a well-defined semi-lunar terrace interposes between this hollow and the steeper slope to the south-west. Some 20 feet below this terrace is a well. xxxiv. N.E. 26 July 1912. 291. Fort, Carthur Hill.—This fort is situated to the south-west of Waterhead, on the extreme east edge of the summit of Carthur Hill, an eminence which rises very steeply for some 250 feet above the right bank of the Dryfe Water. On the west a hollow cuts off the site of the fort from the extended plateau of the hilltop, while along the east side the slope is almost precipitous. The enceinte, which is oblong with rounded ends, lies with its longest axis north-east and south-west, and measures some 210 feet by 120 feet. A narrow ridge of rock running along the crest of the hill from the north-east appears to have been utilised as the line of the defence on the north-west side. Along this has been con¬ structed a stony rampart, formed to a consider¬ able extent by the cutting away of the rock on the interior, which has apparently been sur¬ mounted by a wall, the whole rising to a height of about II feet. On the exterior it is formed with two distinct gradients, divided by a shoulder at about 5 feet below the crest, the lower portion being steeper, while the upper is probably encumbered with debris from the wall. The rampart or wall of the fort has been carried across the space intervening between the ridge and the edge of the summit, and, at a diminished height, also along the east edge. There has been an entrance into the lowest part of the interior from the south-east, opening into a hollow which appears to have been excavated, and above which the ground slopes upward to the crest of the rampart some 15 feet in height. In rear of the hollow, and at a slightly higher level, is a circular depression cut out of rock on the upper side and with its longest axis towards the entrance, measuring some 34 feet by 28 feet; and, on the right of the hollow, a slight ridge on the surface indicates an oval foundation, probably of a hut-circle. There seems to have been a second entrance by a ramp a little to the south-east of the other, approached along the face of the steep bank. The fort on Range Castle Hill looks down on this from the opposite side of the Dryfe. xxxiv. N.E. 26 July 1912. 292. Fort, Hencastle Rig, Old Craighouse.— This fort is situated on the end of a low saddle- backed ridge called the Hencastle Rig, which runs down from the east, expanding somewhat suddenly, with a circular plateau at its western extremity, where it projects into the haugh- land by the left bank of the Corrie Water, some 100 yards distant from the stream. A burn flows down a ravine on the north, and the sides of the plateau present a somewhat steep escarpment, steepest at the west end, where it has an elevation of some 30 feet above the meadowland. The fort on plan appears to have been oval with its longest axis north- north-east to south-south-west; but the north- north-east extremity is now bounded by the somewhat straight edge of the bank overlook¬ ing the burn. The interior now measures 164 feet by 131 feet. Surrounding it, except above the bank, is a broad rampart of earth and stone some 4 to 5 feet high in the interior, increasing on the exterior from 4 feet on the east to 8 or 9 feet on the south and west, with a concentric trench in front some 18 feet wide from the base of the rampart and now showing a very slight concavity. There has been a broad entrance to the area up the steep west face, and on the east side there appears to have been a narrow opening not far back from the bank of the burn. xliii. S.E. 6 August 1912. 293. Fort (remains of), Parkcleuch, Corrie Common.—About | mile to the west of Corrie Common and J mile east by south of Park- cleuchfoot, is a segment of a curvilinear fort outside a wood on the north, measuring along the curve some 240 feet in length. It has apparently rested on a steep bank on the north side of a linn, but beyond the segment in the rough pasture no trace remains. The existing portion consists of a massive rampart rising some 6 feet above the interior level and dropping steeply for 12 feet at most to a trench, in part a natural hollow, 38 feet wide from crest to crest with a depth of 7 feet below the crest of the counterscarp. As the 109 HUTTOX AND] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [CORRIE. rampart diverges at the north-west from the line of the original hollow, the trench ceases. In rear of the defences at the north-west end is an oval hut-circle, measuring interiorly 38 feet by 32 feet, with its longest axis north¬ west and south-east and surrounded by a low bank of earth some 7 feet wide with a narrow break through it at either end, as if for entrances. xliii. N.E. 6 August 1912. 294. Fort, Carterton.—This fort lies adjacent to the farm of Carterton at the head of the glen of the Corrie Water. The site is the termination of a long, narrow, saddle-backed ridge running in a south-westerly direction and forming the watershed between the Back Burn and the Corrie Water, which flow by the base of the fort on either side, uniting some 70 to 80 yards distant from the point of the ridge. The fort occupies the crest, the ground declining towards it by an easy gradient from the north-east, dropping steeply to the respective burns on either flank, and running out to the termination of the ridge at a gradient of about i in 9. The fort lies at an elevation of about 600 feet above sea-level, some 36 feet above the burns at its higher end and at its lower about 30 feet. It is oblong on plan, with its longest axis north-east and south-west, measures 198 feet by 108 feet, and is surrounded by a massive rampart of earth and stone, which rises above the interior level some 8 feet at the north-east end and diminishes to 3 to 4 feet along each flank. On the north-west side the fort is not far distant from the pre¬ cipitous edge of the ravine of the Corrie, but on the opposite or south-east flank a stretch of level ground intervenes between the base of the ridge and the burn. On the former side outer defences have been deemed unnecessary ; but on the latter, some 15 feet below the crest of the parapet at the higher end and 10 feet at the lower, is a terrace varying from 13 to 18 feet in breadth, commencing directly below the line of the rampart at the north¬ east end and carried round to the south-west, where, faced with a rampart, it forms a small forecourt 18 feet broad in front of the entrance. At 80 feet out from the rampart, at the north-east end, another mound 25 feet broad at base but now low in elevation, crosses the higher part of the ridge, having in front of it a trench-like hollow 36 feet in width and 4 feet deep below the crest. The lower half of the interior appears to have been to some extent hollowed by excavation. The entrance has been through the centre of the lower end, passing through the outer rampart over the forecourt and between two massive mounds for a distance of 52 feet. These mounds are probably for the most part natural, formed by the excavation of the ground on either side ; and on the top of each is a hollow, the site either of a hut or of a massive parapet wall which has been torn out. The width of the entrance is 6 feet. In the north angle there is another smaller entrance or postern, which, judging from the inward curves of the base of the rampart on either side, has originally been narrower than it is at present. The crest of the rampart is stony and has possibly been surmounted by a parapet wall carried across the summit of the mounds on either side of the main entrance. The interior has a fall of some 8 feet from back to front. xxxiv. S.E. 7 August 1912. 295. Fort, Dormont Rig, Carterton.—This fort is situated on the crest of the Dormont Rig, at an elevation of 807 feet above sea-level, and about | mile east-north-east of Carterton. It occupies the north-east extremity of the Rig, which declines in front of it slightly to rise to a higher eminence a mile away, and holds in view to the southward the vale of the Corrie Water and to north-west and north¬ east a fringe of distant mountain tops. The fort has been curvilinear, either circular or oval, but its original shape has been lost by the destruction in cultivation of the south arc. The remaining portion is surrounded by a rampart, rising some 3 feet above the interior level and dropping 6 or 7 feet to a trench some 30 feet wide, cut through rock, and about 3 feet in depth below the crest of the counterscarp. Across the moor¬ land from the north-east a mound some ii feet in breadth and about 2 feet in height, with a trench some 13 feet in width on the west side, is to be seen approaching directly towards the highest point of the fort. As it draws near to it a second parallel mound is discernible on the west side of the trench; and at 52 feet distant from the outer rampart of the fort, the two mounds diverge to join no HUTTON AND] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [CORRIE. with the counterscarp mound as it passes along each flank, forming an angular enclosure in front. There is no perceptible junction of these mounds with the actual defences of the fort, but they merge into them as if they were a part of the original construction. xxxiv. S.E. 7 August 1912, 296. Mote of Hutton. — The Mote of Hutton is situated in an elevated and unusually exposed position for a mote- hill, about I mile south by east of Nether Hutton, and at an elevation of 718 feet over sea-level. It rises on the west from the edge of a deep natural ravine, the bottom of which lies some 30 feet below the base of the hillock, to a height of about 23 feet. It is in form a truncated cone, is constructed on a knoll of rock, and is surrounded at base by a trench, which, when cut through higher ground on the west, measures some 28 feet in width at the level of the top of the counterscarp and some 7 feet in depth ; on the lower side, where the trench is girt by a mound on the outside, it has a width of about 10 feet and a depth of 3 feet. The summit is very uneven, and the surface has been considerably disturbed ; but it appears to have been circular, and to have had a diameter of some 24 feet. The cincture of the trench is complete ; but from it, on the north-east, a branch extends outwards for some 28 feet in a north-easterly direction, thereafter making a return for some 52 feet towards the south-west. The ear¬ shaped area thus enclosed has a mound above this trench, and, except at its northern end, another along the side of the trench that encircles the mote, leaving between them a slight depression, which seems to have been an approach to the base of the hillock, whence the means of access rose to the summit. The area is too small to have formed a base-court. In 1664 we have a grant of part of the lands of the barony of Hutton, including the “ land of Liverknow and Mote ” {Reg. Mag. Sig., 1660-1668, No. 685). xxxiv. S.E. 9 August 1912. 297. Fort, Broom Hill.—This fort is situated at an elevation of 700 feet above sea-level about I mile to the north-west of Show farm and on the eastern extremity of a ridge, which falls sharply on the north to the burn 100 feet beneath it and on the south declines by an easier gradient in a south-easterly direction. On plan it is pear-shaped, measuring from front to back or west to east 172 feet by 176 feet across. Facing the west is a bold rampart, rising some 4 to 5 feet above the interior level and dropping on the exterior 7 feet 6 inches to a trench in front 27 feet wide from crest to crest. The trench does not appear to have been carried around the flanks, where also the rampart is much slighter, but presents a scarp to the outside, 8 feet in height on the east, 10 feet on the south, and less above the steep ground on the north. The interior is neither levelled nor hollowed but rises from either side to a ridge crossing it from west to east. In front, at the west end, there is a semi-lunar area with a depth of 50 feet at centre, which is enclosed by a broad mound meeting the sides of the fort on north-west and south-east. How much of this outwork is original it is difficult to determine, but in part it is certainly secondary. The main entrance into the fort proper is from the west, at a point where the ground commences to decline to the north, and it is significant that there does not appear to be a corresponding gap in the outer mound. At the point of junction of the two mounds on the north the main rampart shows distinct signs of having been cut back. xxxiv. S.W. 12 August 1912. 298. Fort, Whitcastles. — This enclosure is adjacent to the farm of Whitcastles, at an elevation of 700 feet over sea-level. In form it is circular, measuring some 175 feet in diameter, and has been surrounded by a ram¬ part of earth and stone, rising some 2 to 3 feet on the inner side and about 5 feet on the ex¬ terior. In the centre there has been a quarry, and the rampart has also been dug into to some extent. It is doubtful if it has been encircled by a trench, though at one or two points on the circumference a trench-like hollow is observable. The construction lies within a wood, and is thickly overgrown with vegetation. xliv. N.W. 13 August 1912. Enclosures. 299. Enclosure, Dryfe Lodge.—This appears to have been a square or oblong enclosure, III HUTTON AND] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [CORRIE. but two sides have disappeared beneath the buildings of Dryfe Lodge and its outhouses. It is situated where the bank facing the Dryfe Water on the east side of the valley makes a sudden bend outwards at right angles towards the stream, which flows by about loo yards distant. The bank has an elevation of from 20 to 30 feet and in itself forms a consider¬ able defence on the north and west. The area, the original dimensions of which are un¬ obtainable, is surrounded on the two remaining sides by a broad rampart of earth and stone, varpng from 3 to 5 feet in height on the interior and sloping uniformly with the bank to its base. The west side is straight, but the north curves somewhat. There are two breaks through the rampart into the interior; one at the north-west angle, which is modern, and the other near the end of the west side ; the latter, if original, has been altered, and a road leads up to it from the meadow below. xxxiv. N.E. 26 July 1912. 300. Enclosure, Cocklawrig.—On the south point of the Cocklawrig, at an elevation of some 800 feet over sea-level, and at a distance of about J mile to the north of Waterhead, is a circular enclosure measuring some 120 feet by no feet in diameter. The rig forms the water¬ shed between the Dryfe Water and the Water- head Burn, and faces straight down the valley of the former. The enclosure has been formed by a stone wall some 6 feet in thickness at base, now reduced to a low stony bank over¬ grown with turf. It is not surrounded by a trench, and, except for its site, affords no suggestion of a defensive construction. It appears to have been hollowed to a depth of some 2 feet in the interior. xxxiv. N.E. (“ Fort ”). 26 July 1912. 301. Enclosure, Peat Hill.—On the east slope of Peat Hill, at an elevation of some 800 feet over sea-level and looking down a hollow southward into Dryfesdale, are portions of an oval enclosure. It lies on sloping ground, dipping considerably to a hollow on the east, in which there is a spring of water. The construction consists of two segments of rampart; the sides of the oval, some 200 feet and 160 feet in length respectively, being covered by a trench 23 feet wide and 4 feet deep. On the north-west and south-east—the highest and lowest points—the cincture does not appear to have been completed, but from end to end of the south-west segment a low bank has been erected, forming an enclosure on that side, at the upper end of which is a circular excavated hollow some 31 feet in diameter. xxxiv. N.E. {" Fort ”). 26 July 1912. 302. Enclosure, Waterhead. — Behind the shepherd’s house at Waterhead, on the left bank of the Dryfe and some 40 feet above the base of the hill, there is a four-sided enclosure with rounded angles, measuring about 100 feet each way, dug out of the falling ground, with a scarp from 5 to 8 feet high on three sides, and the apparent foundations of a wall along the fourth, from the front of which the ground falls sharply for some 20 feet to a low terrace above the haugh-land reaching to the river. The floor of the interior has been cultivated, and is level. This enclosure is similar to that at Burnfoot and many others seemingly associated with farms and shepherds’ houses. They present features much resemble certain enclosures which have a surrounding trench, and do thus show some defensive character. xxxiv. N.E. (unnoted). 26 July 1912. 303. Enclosure, Craighouse.—About J mile south by west of the farm of Craighouse and on the face of a bank which slopes down to the haugh-land by the Corrie Water, is a pear- shaped enclosure, lying with its longest axis north-east and south-west and measuring 260 feet by 194 feet. To north and south it is bounded by a small natural ravine. The con¬ struction has been surrounded, except where the ravine on the south serves a similar purpose and at the base of the slope, by a trench some 30 feet in width from crest to crest, flat- bottomed on the north-east or higher end and V-shaped on the lower. In the former position the scarp and counterscarp have, respectively, heights of 5 feet and 4 feet, and in the latter of considerably more. Above the trench on the interior is a parapet mound, in rear of which the ground appears to have been dug out, so that from the crest of the mound there is a gradual decline merging into the slope of the interior. The enclosure is on falling ground and opens with no defences at its lower end on to the low 112 HUTTON AND] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [CORRIE haugh-land, while it has rising ground im¬ mediately above it. For the last 82 feet of the longer diameter to the bottom of the slope the gradient is steeper. xliii. S.E. (“ Fort ”). 6 August 1912. 304. Enclosure, Pyatshaws Rig. — On the crest of the Pyatshaws Rig, a prominent ridge running northward from the valley of the Water of Milk, to the south-east of Corrie Common, and forming the watershed between the Trout Beck and the Stidriggs Burn, is a circular enclosure lying at an elevation of 744 feet over sea-level and measuring 162 feet in diameter. It is surrounded by a ram¬ part of earth and stone some 2 to 3 feet in height on the inner side, which drops some 6 feet on the exterior to a narrow trench 19 feet wide from crest to crest. The entrance is on the east, some 9 feet wide, and passes over the trench on unexcavated ground. On the south there is a narrow passage from the trench, expanding slightly to a circular hollow some 6 to 8 feet in diameter as it passes inward through the rampart, flanked on either side by a broad mound or talus projecting out from the rampart, and bounded on the inside beyond the line of the rampart by a mound 24 feet long and 14 feet broad. A narrow passage, slightly higher than the level of the interior, passes by the end of this mound inwards. The counterscarp in front of this excavated hollow is not broken down, nor is the trench intermitted. Lying within the rampart, on the south side, is an oval mound, 2 feet in height and measuring superficially 18 feet by 13 feet, which is overgrown with grass. xliv. S.W. (“ Fort ”). 7 August 1912. 305. Enclosure, Shaw.—In the angle of the burn, to the north-east of No. 303 and about |- mile north by east of Shaw, is an oblong enclos¬ ure lying with its longest axis north and south, four-sided but not quite rectangular, the sides measuring respectively 76 feet, 122 feet, 60 feet, and 116 feet. There are two entrances, one in the north end and the other in the west. In the south-west corner is a small circular depression, with a diameter of 10 feet, possibly the site of a hut. The surrounding bank of earth is some 12 feet broad at base and from 3 feet 6 inches to 4 feet in height. A slight trench-like hollow in front on the north and east sides is probably due to the excavation of material. This is no doubt an ancient sheepfold, and the modern pens lie adjacent to it. xxxiv. S.W. (unnoted). 12 August 1912. 306. Enclosure, Shankend.—This is a small circular enclosure, measuring 100 feet in dia¬ meter, situated on a plateau sheltered on the west and north by rising ground but with a steep slope on the east to the margin of the Caldwell Burn, flowing nearly 200 feet below. The interior has been hollowed by excavation to a depth of from 4 to 5 feet, and is sur¬ rounded by a slight stony rampart or wall. The entrance has been from the south into the lowest part of the interior area. xxxiv. N.W. (“ Fort ”). 12 August 1912. Stone Circle. 307. Stone Circle, Whitcastles.—Situated on open moorland about 150 yards to the north of the hill road, at an elevation of some 850 feet over sea-level, and nearly i mile to the north¬ west of Whitcastles farm, is a stone circle, con¬ sisting of nine massive blocks, for the most part whinstone boulders. All the stones lie pros¬ trate, with the possible exception of one on the western arc, the point of which protrudes about I foot 10 inches above ground. The largest blocks are on the north and south : the former a roughly oblong mass measuring 7 feet 4 inches by 4 feet 3 inches by 2 feet 10 inches in thickness, and the latter a pointed boulder, 7 feet i inch long by 3 feet 9 inches in greatest breadth, and 2 feet in thickness. The diameter of the circle has been approxi¬ mately 160 feet : the individual stones are irregularly placed, those on the eastern half of the circle being more widely spaced than those on the western. xliv. N.W. 13 August 1912. [Miscellaneous. 308. Heraldic Stone, Berryscaur.—Forming the lintel of a cottage door facing the high road, at the east end of the hamlet of Berryscaur, is a slab measuring 3 feet 4 inches in length by I foot 2 1 inches in depth. It is carved in 8 113 HL-TTOX ANDCORRIE.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [johnstone. relief as follows ;—In the centre, occupying almost the entire depth by i foot in breadth, is a shield bearing arms : A lion rampant within a bordure fiory counter-flory, the Royal Arms of Scotland : on the dexter side a holly leaf, and bej^ond it a saltire : on the sinister side, at the extreme end, in Gothic form, the letters A.B., separated by a scroll or reversed S. The top of the shield, hidden by the iroii gutter at the edge of the cottage roof, is partly damaged on the sinister side. This stone is said to have been found in 1783, underground, in the remains of an ancient building at Westside, on the Black Esk, in the parish of Eskdalemuir, and in that year it was transferred to Berryscaur and used as a lintel. The saltire and holly leaf are respectively the arms and badge of the Lords Maxwell ; the letters A.B. are probably the initials of a member of the Beatty family, one of whom, in 1532, was King’s sergeant and officer in Eskdale. On the map of 1590 the tower of Ally Battle is marked at a place correspond¬ ing with Westside, and the O.S. map marks a spot as “ Sergeant Know,” within 2 miles of it. xxxiv. S.W. 12 August 1912. 309. Gravestone, Corrie Churchyard. — At the north-east angle of Corrie Churchyard, outside a railed enclosure, which forms the burial-place of the Grahams of Dunnabie, is an upright slab commemorating peter graham IX BARNSDEL WHO DEPERTED THIS LIFE OCT. 2IST 1753 AGED 12 YEARS. On the front is a figure of a man dressed in a long skirted coat with deep cuffs, holding in his right hand a crown and in his left a sceptre. On his left a skeleton stands on a skull grasping a spear in his left hand. Above the man’s head is an hour-glass, and at the apex of the stone an angel with outspread wings. xxxiv. S.E. 26 July 1912. 310. Cross-slab, Corrie Old Churchyard.— On a mound in the centre of the churchyard, evidently covering the ruins of the old church, lies a squared block of sandstone, measuring 6 feet 5I inches in length, 9 inches in thickness, i foot ii inches in breadth at the head, diminishing to i foot 6 inches at the foot, whereon is carved a foliated cross in the form of a cross-potent with a lozenge¬ shaped boss in the centre, having a long shaft set on a calvary mound. A broad- bladed symmetrically pointed sword is incised on one side of the shaft, the handle of which is entirely worn away. The cross is carved in relief, but is much weathered. The edge of the stone has a border of projecting dog¬ toothed bosses, 6 inches apart, rising from a 4-inch chamfer, the interspaces on the chamfer decorated with a leaf - and - bead ornament. xliii. S.E. 6 August 1912. Sites. The O.S. maps indicate sites as under :— 311. Corrie Church, Corriehills. xliii. S.E. 312. Chapel, near Carterton, xxxiv. S.E. 313. “ Mosskesso,” about 500 yards north¬ east of Closs. xxxiv. N.E. (C/. No. 287.) 314. “ Covenanters’ Graves,” Caldwell Burn, xxxiv. N.W. JOHNSTONE. Castellated and Domestic Structure. 315. Lochwood Tower.—This tower occupies a naturally strong site some 6 miles south of Moffat, which is defended to the north by broken and wooded ground and by the Loch¬ wood Moss — once an almost impenetrable morass—in other directions. The tower is placed at the southern end of the site, with a range of outbuildings extending 160 feet northwards, where it abuts on the southern defence of a circular mote-hill — Lochwood Mount (No. 316). The buildings are in a ruinous state, the main building being com¬ paratively complete to the level of the first floor, while the south-eastern angle stands some 20 feet higher; the outbuildings are mere shells. On plan (fig. 84) the tower is L-shaped, and was entered at the first-floor level. There appears to have been no external opening to the basement, an unusual feature. The larger wing, with its main axis running east and west, measures 43 feet 5 inches by 34 feet exteriorly, with walls 6 to 9 feet thick. Internally, the basement, entered from a 114 OQ ••*''’ Vr J-# -3^/0 _• ^ ^ - ^__^^._ j j 'I ^1 4 ^ lit BLOCK. PLAN GROUND ■ PLAN PLAN OF PRISON Fig. 84.—Lcchwocd Tower and Mote (Nos. 315 and 316). JOHNSTOXE.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [johnstone. wheel-stair contained within the short wing, is divided into two vaulted compartments, each furnished with a ceiling hatch and lit by long narrow windows with jambs widely splayed to the interior. Beneath the easter cellar is a prison 21 feet 4 inches long and 5 feet 7 inches wide ; access to this is gained by a straight stair entered from the short passage between the wheel-stair in the smaller wing and the eastern cellar. The door in the passage appears to have folded vertically in leaves, for a recess and groove I inch deep, in which such a door might slide, are worked on the lintel. There is no window, but a ventilation outlet is provided in the east wall. A bulkhead is formed in the cellar above to obtain the necessary headroom in the dungeon. From the general arrange¬ ment of the plan the tower appears to have been built in the 15th century. The outbuildings to the north, which were apparently erected at a later period, formed the eastern boundary of a courtyard, while traces of a smaller courtyard to the west are evident. Lochwood was the principal seat of the Johnstones. As described by its English captors in 1547, “ It was a fair large tower, able to lodge all our company safely, with a barnekin, hall, kitchen, and stables, all within the barnekin, and was but kept with two or three fellows and as many wenches.” ^ In April 1585 Robert Maxwell with his friends and some Armstrongs attacked and burnt Lochwood, ” the Lardes {i.e. Johnstone’s) owne howse,” and its provision of victuals.^ Cf. further Introd., p. Ixv. ^ Cited in History of Westniorland and Cum¬ berland, Nicolson and Burn, vol. i., p. liv. ; '^Border Papers, i. No. 303. xxiv. S.E. 10 May 1912. Defensive Constructions. 316. Mote, Lochwood. — This mote, com¬ monly called “ The Mount,” is situated in a wood of aged oak trees, just to the north of the ruins of Lochwood Castle, and looks out to the eastward over the plain of Upper Annandale. It is formed from a natural hillock crowning a slope rising from the east and mounting from a hollow on the west. From the latter direction it has a vertical elevation of some 22 feet; but from its base on the east, along which runs the roadway, it rises to a height of 44 feet or thereby. Two terraces encircle it: the upper one on the west side, at some 10 feet below the summit level, dipping on the longer western slope to 20 feet, and the lower one varying from 8 to 10 feet further down. On the east side and round by the north, both terraces show a parapet, and on the south the lower takes a trench-like aspect with a bold rampart cutting it off from the ground beyond, on which the later castle has stood. On the north the hillock does not slope directly to its base from the parapet of the lower terrace, but presents a narrow bench crowned at its edge with a rampart, from which there is a scarp some 5 to 6 feet in height to the lowest level. To the eastward this bench gradually merges with the narrow terrace above it, and to westward it slopes away to a lower level, leaving the rampart extend¬ ing onwards in that direction, and containing within it an area too low lying to have formed a base-court. Towards the south-east the upper terrace forms a salient angle; and directly below it there is a gap, which has probably been an entrance through the outer mound, some 7 feet in width, towards which, between two parallel mounds, what may have been a roadway maj^ be seen approaching directly to it on the opposite side of the present road. Westward from this supposed approach, and facing the south, there is a space, some 12 to 14 feet in breadth, reach¬ ing downward from the edge of the summit, interrupting the upper terrace and scarped at a flatter angle than the rest of the mound, up the west side of which there is a distinct suggestion of a track, which makes a sharp turn to the eastward at the highest level before entering on to the summit at its south-east point. On the east side of this space, stretching from the summit to the trench-like hollow of the lower terrace, there is visible a stony artificial ridge. The summit is oval, measuring superficially some 24 feet by 16 feet, and has been hollowed to a depth of some 18 inches, with a wall formed in part of natural rock left around the edge. xxiv. S.E. (unnoted). 14 September 1912 116 JOHNSTONE.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [johnstone. 317, Mound behind Lochwood Tower.—To the south of the ruins of Lochwood Tower lies a green level meadow, probably the garden, and near its centre there rises an artificial¬ looking earthen mound surmounted by four ancient gean trees, some 9 feet in height with a diameter at base of 36 to 40 feet, fallen away somewhat towards the east, and measuring across its level circular summit some 10 feet. Around its base is a shallow trench with a width of 12 feet. This mound in character and situa¬ tion bears a resemblance to that which rises from the centre of the garden at Logan in Gallo¬ way, similarly within sight of the old castle. xxiv. S.E. (unnoted). 14 September 1912. 318. Fort, Mote Cottage.—On the east bank of the Kinnell Water, about J mile east by south of the farm of Ross Mains, rises a grassy hillock marked as a mote on the O.S. map. It is a natural gravel mound, lying with its longest axis north and south, with an eleva¬ tion rising from 18 feet at the north end to 26 feet at the south, steeply sloped on the north and west and falling by an easier gradient to its base on the south and east. The ground around is low-lying meadow land ; and, while the Kinnell Water at the present day flows by some 150 yards to the w’estward, an old channel marked by pools of stagnant water lies at its base. The summit has been surrounded by a bank of earth and stone enclosing an area measuring some 100 feet by 40 feet. It slopes from west to east as well as from north to south, and at no point has been levelled, as would be the case in a mote-hill. At the lowest point on the east side, towards the north end, there is an entrance 8 feet wide approached up the slope from the base; and on the right of it, against the bank, there appears to be an oblong foundation, probably of turf, at the east end of which, at a level some 5 feet lower, is a circular hollow dug out of the face of the bank, measuring ii feet in diameter. On the highest point of the hillock, in the line of the enclosing bank, is a small oblong depression measuring superficially 7 feet 6 inches by 5 feet 6 inches and sunk some 2 feet below the surface; while on the east edge, also on the line of the mound, is another hollow, which may mark the site of a hut. xlii. N.E. (“ Mote ”). 8 August 1912. 319. Fort (rerriains), Kirkhill Cottage.—This cottage, in an angle between two roads about I mile to the north of Johnstone Church, apparently occupies the site of a fort, of which a small portion of a rampart remains on the north. xxxiii. N.E. 14 August 1912. 320. Fort, Tanner’s Linn, Mollin.—This is a small semi-oval fort about | mile south¬ east of the farm of Mollin, the oval bisected obliquely, and in its periphery, exclusive of the chord, presenting four distinct facets of vary¬ ing dimensions. It rests on the edge of the precipitous left bank of the Linn, which flows through a wooded ravine some 50 feet below. The main axis of the oval, if complete, would have been north and south, and the basal line of the fort lies from north-east to south-west, measuring 93 feet in the latter direction from crest to crest with a bisectional diameter of 60 feet. The defences consist of an inner rampart of earth and stone, a deep regularly formed concentric trench, and an outer rampart. The inner mound has an elevation of some 5 feet above the portion of the interior directly behind it, and has a scarp at the highest point of 7 to 8 feet in height above the floor of the trench, but along the north-east arc of only some 3 feet 6 inches ; the trench from crest to crest measures 30 feet, except on the north-east face, where it measures 25 feet, and has a depth below the counterscarp of 7 feet where deepest, near the centre of the curve on the north, and diminishes in depth towards the edge of the ravine at either end. At either extremity the outer rampart has a height of from 5 to 6 feet on the e.xterior, where the ground level declines to the edge of the ravine. There is much stone at places in the interior, especially at the north-east end, but no distinct foundation is traceable. The entrance has probably been from the north-east, past the end of the rampart, and flanked by the precipitous side of the linn— an arrangement frequently observed in this class of fort. From the west there is a slight filling of the trench, to form a gangway to the interior ; and thence southward to the edge of the ravine the inner rampart has an eleva¬ tion some 2 to 3 feet lower than to north- 117 JOHXSTOXE.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [johnstone. ward. It is doubtful, however, if this is original, as there is no break in the continuity of the outer mound as it passes it. xxxiii. N.W. 21 August 1912. 321. Fort, Mollin.—This fort is situated on the crest of a ridge, at an elevation of 600 feet over sea-level and about J mile north by east of the farm of Mollin. It appears to have been an oval enclosure, with its longest axis north and south, measuring interiorly 156 feet by 140 feet, divided by a cross rampart or wall in such a way as to cut off a semi-lunar segment, amounting to about oae-third of the interior area at the north end. The whole enclosure has been surrounded by a stony rampart, now of low elevation and completely eradicated for a distance of some 80 or 90 feet on the south-east. xxxiii. N.W. 21 August 1912. Exxlosures. 322. Enclosure, Crunzierton Wood, Raehills. —Situated on a shelf on the steeply sloping ground within the Crunzierton Wood in the policies of Raehills, a short distance from the gamekeeper’s cottage, is a circular en¬ closure, measuring 102 feet in diameter, surrounded by a wall or stony mound, now of low elevation and sc me 13 feet in thick¬ ness. The interior has been hollowed by ex¬ cavation on the upper side to a depth of from 3 to 4 feet and is completely overlooked by the rising ground to the west of it. There is a wet spot, which is probably a spring, at the west end, and there is an indication of a cross-wall cutting off a segment towards the north, as in the last-mentioned enclosure. It lies at an elevation of 600 feet over sea- level and some 30 to 40 feet above the road which passes along the base of the slope. In its position it bears a strong resemblance to the hollowed enclosures in Eskdale. xxxiii. N.W. (“ Fort ”). 21 August 1912. 323. Enclosure, Duff Kinnel Bank, Raehills. —Situated on a level plateau on the western slope of the valley of the Kinnel Water, about mile north-north-west of Raehills House, is a fort at an elevation of 600 feet over sea-level and some 30 feet above the roadway. It is a circular construction, measur¬ ing interiorly some 114 feet in diameter, and has been surrounded by a massive wall or rampart of stones, now structureless, measur¬ ing at most 19 feet across and 5 feet in height. Towards the edge of the slope to the roadway the mound of stone is much more massive than on the opposite side of the enceinte, which is overlooked by the slope rising steeply above it; nor is there any indication that the defence has originally been so great on this side. The entrance from the south-east is clearly defined and measures 7 feet in width. 324. Enclosure, Duff Kinnel Bank, Raehills. —Higher up, just on the brow of the hill and some 200 yards distant from the last, is an oval construction, lying with its longest axis north and south and measuring 130 feet by III feet, also surrounded by a ruined wall or stony rampart, of much slighter dimensions, however, than that of the last enclosure, measuring some 9 feet over all. About one- third of the interior area at the south end has been cut off by a cross-wall, and there are indications of hollowing by excavation. The wall of a later construction, probably enclosing a wood now blown down, is partially super¬ imposed. xxxiii. N.W. 21 August 1912. 325. Enclosure, Edgemoor.—This enclosure is situated by the edge of the steep east bank of the Kinnel Water, about | mile to the south¬ west of the farm of Edgemoor. It has been oval in form, lying with its longest axis east by north and west by south, measuring 150 feet by 130 feet or thereby. Its north-west arc has been destroyed in the formation of a road, and the rest of its defence has been greatly pillaged for stones. It has been surrounded by a massive stone wall, of which only a small section, some 40 feet in length, remains on the north-west, adjacent to the hedge bounding the road. The large blocks which have formed the lower course on the outer face still remain in situ, indicating a breadth for the wall of some 10 feet. The interior is lower by 2 or 3 feet than the sur¬ rounding ground and is wet and overgrown with rushes. 118 JOHNSTONE.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [keir. The entrance has been from the east, now some 12 feet wide, and one large stone sunk firmly in the ground in an upright position still stands in situ on the north side. xxxiii. S.E. (“ Fort ”). 21 August 1912. Site. The O.S. map indicates a site as under :— 326. Chapel near Johnstonecleuch. xxxiii. S.E. KEIR. Castellated and Domestic Structure. 327. Barjarg Tower.—This house, which is distant some 4 miles by road to the south-east of Penpont, is situated in a woodland dis¬ trict on the western side of the river Nith and commands an extensive view of the valley to the north and south. The existing build¬ ing consists of a central dwelling-house of modern construction, to the east of which is the old tower and to the west a modern wing containing stable offices built to imitate the outward appearance of the older wing. The eastern tower is L-shaped on plan ; the main portion is vaulted on the ground floor, and measures some 20 feet 6 inches by 13 feet 6 inches within walls averaging from 4 feet to 5 feet 6 inches in thickness, while the short wing has a projection of some 13 feet. Origin¬ ally the staircase leading from the basement to the upper floors has been formed at the re-entering angle. The entrance doorway is situated in the north wall of the short wing and is secured by an iron “ yett ” ; while the staircase is contained within a circular tower of late date measuring some 8 feet inside the walls, and projecting to the south. Access to the wheel-stair is gained from the interior of the short wing, as shown on plan. Doorways formed at various levels in the western wall of the tower communicate with the central part of the modern ‘house, which is three storeys in height. Circular turrets rest on corbelled projections at the angles of the old building and form separate chambers with access from the attic floor level. On the east face of the west or modern wing is a panel bearing in relief the date 1680 and the initials I G ; G K, for John Grierson and Grizzel Kirkpatrick his wife. The panel is an insertion, and doubtless refers to the eastern tower. Barjarg is said to have been given by the Earl of Morton to Thomas Grierson in 1587 {New Slat. Acet., vol. iv.). A panel stone lies near, bearing on the dexter side the date 1603 above an ornament, the initials, much defaced, R M (? Robert Maxwell), and a shield charged with a human face ; on the sinister side two shields, one above the other, each bearing Maxwell arms, A saltire with a holly leaf in base. xxxi. S.E. 12 June 1912. Cairns. 328. Small Cairns, Barjarg Moor.—Stretch¬ ing up the glen of the Auchenage Burn on Barjarg Moor, from about 100 yards north¬ west of the dyke that encloses the cultivated land on Glenlaugh farm, are several groups of small cairns. The first group nearest the dyke contains about four low cairns apparently of earth and stone, which measure in diameter about 17 feet. Some 200 yards along the hillside, and at a slightly higher level, is another small group comprising about half a dozen. Associated with these there appear to be two hut-circles, the enclosing banks of both of which have been formed with small stones and a very few large ones. The most northerly has an interior diameter of ii feet 6 inches and has been entered from the south-east. The floor is slightly sunk beneath the surrounding level. In the other circle, to the west of the group, the interior is less well defined, but measures some 14 feet in interior diameter, while the over-all measurements are 20 feet by 20 feet 6 inches. The position of the entrance is doubtful, but it has probably been from the east-north-east. On the east side of the glen, about 150 yards west of the end of a wood which lies parallel with the burn, is a group of four small cairns —one of which has been partially cleared out, —while about 100 yards farther down the glen there are other si.x, measuring from 10 to 12 feet in diameter and about 2 feet in 119 KEIR.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [kirkconnel. elevation. All these small cairns lie between the Soo and 900 feet contour lines of the O.S. map. xxxi. S.W. and xl. N.W. 6 June 1912. 329. Long Cairn, Capenoch Moor.—On open moorland, about 150 yards south-east of the most south-easterly point of Capenoch Big Wood, lies a large cairn of oblong form, having its longest axis north-east and south-west and expanding somewhat to the latter direction. In length it measures 120 feet, in breadth at the north-east end 31 feet, at 66 feet onwards 53 feet, and at the south-west extremity 57 feet. For a distance of about 35 feet from the north-east end it has been almost entirely removed; thereafter it rises in elevation from 6 to ii feet. The front has been very straight, and there is no indication of a frontal semicircle. At one or two places on the west side and at the south-west end, small portions of a built facing wall are still discernible. At some points slight excavation has been attempted with no results. xxxi. N.W. (unnoted). 9 June 1912. Miscellaneous. 330. Foundations, Capenoch.—In a meadow enclosed b}^ woodlands, about | mile south¬ west of Capenoch House, are foundations which appear to be those of a small rectangular tower with buildings surrounding a courtyard attached to it. Capenoch was the property of one of the branches of the Grierson family. (Macfarlane’s Geographical Collections, hi. p. 207.) xxxi. N.W. (unnoted). 6 June 1912. Site. The O.S. map indicates a site as under ;— 331. “ Kilbride,” Kirkbride Mains, xl. N.E. KIRKCONNEL. Ecclesiastical Structure. 332. St Connel’s Church. — This church, which was appropriated to the canons of Holywood, was situated in the churchyard, on a plateau on the southern base of Kirk¬ land Hill, some if miles north of Kirkconnel. The foundations, which are all that now re¬ main, are some 74 feet long and 27 feet wide and indicate that the structure was double chambered, comprising nave and chancel. These dimensions and features suggest that the building was erected in the late 12th or early 13th century. In the dry stone dyke bounding the church¬ yard are many fragments of dressed stone work. On the south side may be seen a base moulding of apparently early section as well as late 15th-century voussoirs and arch stones. In the west side of this wall a slab of stone was found, on which was worked a circular basin 8 inches in diameter—probably part of a stoup ; also a fragment of a cross¬ shaft 6 inches thick, i foot i inch broad, and I foot 3^ inches long, ornamented on back, front, and sides with interlaced work. A shallow basin i| inches deep, formed in a stone 2 feet 6 inches long, i foot 5 inches broad, and 8 inches thick, now forms the top step of the stile at the east side of the churchyard. Many other fragments of cross-shafts, deco¬ rated with interlaced ornament, may be found in this and the other dykes in the vicinity. A portion of the stone roof ridge was found under a tombstone within the churchyard. The present parish church was built in 1729. Within it is preserved the socket of a large cross. A small piece of land in the parish, paying, under a charter of 1444, is. Scots to the ofhciant at the altar of the Blessed Virgin was made liable for this amount to the parish minister. {Stat. Acct., x. p. 453.) V. N.E. May 1912. Miscellaneous. 333. Cross-Socket, Orchard.—In a meadow between the road and the Crawick Water, and some 30 or 40 yards to the south of the cottage at Orchard, is a large square block of stone, 3 feet 6 inches long, i foot 5 inches broad, and 3 feet high, with two rectangular oblong sockets, inches-apart, sunk on the top (fig. 7 of Introduction). The stone has been very roughly squared, and the edges of the hollows are much worn down. On the west side, beneath each mortise, there has been rudely carved a cross, apparently of the cross-paty or formy type, with curved 120 Ancient and Historical Monnmoits — Dumfries. Fie;. 85.—Isle Tower (No. 337). To face p. 121 KiRKCONNEL.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [kirkmahoe. angles of intersection, rising in each case from a shaft, which in turn seems to pass into a solid base. In both crosses the upper arm has been almost entirely worn away. vi. N.E. (“ Font Stone ”). i8 June 1912. Sites. The O.S. maps indicate sites as under :— 334. Castle Robert, Corsebank. ii. S.E. 335. St Connel’s Grave, Half Merk Hill. i. S.E. 336. Cairn, Bail Hill. vi. N.W. Note .—The “ Deil’s Dyke ” is recorded under Sancpihar Parish (No. 566). KIRKMAHOE. Castellated and Domestic Structures. 337. Isle Tower,—This small but unusually complete structure of the i6th century stands on the right bank of the river Nith, 5 miles north-north-west of Dumfries (fig. 85). Oblong on plan (fig. 86) and measuring exteriorly some ig feet 3 inches by 22 feet 3 inches, with walls 3 to 4 feet thick, the building closely resembles Fourmerkland Tower (No. 280), about two miles away, in outward appearance as well as in the simple arrangement of plan. The tower is three storeys and an attic and ter- minates without a parapet in a steeply pitched roof with X ^ turrets projecting from the I V north-east and south-west Each floor is a single apart¬ ment : the basement has a Fig. 86.—Isle Tower Stone-vaulted ceiling, and (No. 337). the two upper floors are carried on oaken beams. A wheel-stair with an inward projection is contrived in the south-east angle, and ascends from the basement to the top floor. The principal apartment is situated on the first floor (see plan) and, although adapted to meet modern requirements, retains most of its ancient features. A garderobe is situated in the north-eastern angle, and a wall cupboard is provided in each wall ; the east wall cupboard is evidently intended as a place of concealment for valuables, as, on passing a hand through an opening in FIRST FLOOR PLAN , 59 ip FEET the soffit, a similar cavity is discovered above. The entrance to the tower is in the south wall; above is a panel (fig. 87) i foot 3^ inches high by I foot 5 inches wide, containing a shield bearing arms: A lion rampant within a bordure charged with a mullet—for Ferguson of Isle. Over the shield is carved the date 1587, and flanking it the initials I F, for John Ferguson. and B R (? ome), his wife. Beneath the initials are the attires of a stag. The entrance was defended, in addition to an outer wooden door and bars, by an iron gate, still in situ, also ha\ing a bar with a loophole in the staircase wall over the door. These defences are minutely described in Proc. Soc. Ant. Scot., x\ii. pp. 107-g. The ' modern mansion abuts on the south wall of the tow’er, and to the west extends a range of outbuildings, on the south wall of which are the initials M A A and the date 1700. The tower now forms a habitable porti n of the mansion, and is in excellent repair. xli. S.W. 10 Julv 1912. 338. Dalswinton Old House.—The ruins of this hc'“ ' ' .mimmimi.e; .llinM'M'M\l|llllll|l>l ' '\'I'I'II„ ‘ii/ ^'fo 7 .. /^»rrraisraniai5Sffl^^^ /^^SssKSffl^ VSZSSSSSSIS^ A " .. vvnillMl||Mi|niM\lll|VVI'''l'l|MMn!i!')lI|n'W;///,,. W, ''yAA '■t—===r=rni ' '“:3i iVTTlnP^ / V' e ^ IQ 0 50 100 150 , 2TO Pttl -fH — I — I — I— I— M — I—I ' I I—I—I—I—I—I—I — I — I — I —f” E.LE.VATI0N (SHEWING BUTTRESS') OF STOREHOUSE Cj Fig. III. —Roman Camp, Birrens (No. 462) 162 MIDDLEBIE.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [middlebie. were formed of large stone flags set on end, and from certain remains it was concluded that its upper margin had been surrounded with a parapet bearing a moulded cope. In this central room the standards of the garrison must have been kept : it was the Sacellum ; while the sunk chamber served the purpose of a general bank or strong room, being always under the watch of a special guard. With the exception of that immediately to the east, the other rooms showed shallow square sink¬ ings in the pavement, on which probably had stood pedestals supporting we know not what. Immediately to the west of the Principia, also at the eastern extremity of the line, and again to the south-east across the street between the east and west gateways, were narrow buildings with walls about 3 feet 8 inches thick and thickly buttressed, which were undoubtedly storehouses or granaries (Cl, C2, Cg). Sleeper walls running length¬ ways showed that in each case the floor had been raised so as to keep it free of damp, while the thickness of the walls and the presence of buttresses indicate that the roofs had been heavy. At the west end of Cg was a small loading platform (c), about 5 feet square and about i foot above the level of the street, much worn on the surface at one place. It will be observed that, owing to lack of evidence, no attempt has been made to appropriate to special uses others of the buildings, but these must include quarters for the Commandant and the officers, as well as workshops, stables, etc. The position of the hypocaust, B, comparable to that in the “ Commandant’s house ” at Balmuildy Fort on the Vallum, may suggest that here, too, the building on the west of it was of this class of residence. The streets generally are composed of a hard beaten bed of gravel margined by two layers of stones laid on the flat with solid gutters outside. In the shorter cross-streets, however, there was but one gutter along the central line. The street running north and south is blocked midway by the Prin¬ cipia, and the northern part of this street was paved over the gravel with cobbles of whinstone. Here, too, the gutter occupied the middle of the street, but, instead of being formed of channelling cut in solid stone, was built up with stone kerbing at the sides on a bottom of flags. There were numerous drains throughout the camp, but the general drainage system could not be made clear. As already mentioned, a very little of the walling remains above the foundations, and that only in portions. The north wall of Cl shows to a height of the seventh course, and is the largest fragment in position. The dressing of the stones on the front wall of Cg is of the familiar reticulated pattern within a plain margin. A portion of this elevation, including the base of a buttress, is shown on the plan. Floor tiles of stone, a piece of roofing tile, and pieces of window glass were also found. Inscribed Stones. —The inscribed stones that are stated to have been found at one time or another on this site number twenty-three,* on two of which the lettering has been entirely defaced. Two additional stones are doubtful, but probably from the same site. Most of the stones are in the National Museum, the others in private possession. A few are little better than fragments. Eleven are either altars or altar-slabs dedicated to divinities, some of them otherwise unknown ; four are probably pedestals, three are legionary stones, one a dedicatory tablet, one a statuette of Bri- gantia, one a sepulchral stone, while three do not furnish sufficient material for decision. Of the two doubtful stones, one, in the porch wall of the parish church, is a dedication to Jupiter by the I. Cohort (Nervana) of Ger¬ mans, and is very probably from Birrens Station, where an altar to Fortune by the same Cohort was found. The other stone has been lost. All these are dealt with in detail in the description of the excavations, and only the more important will here be selected for illustration. First we have the dedicatory tablet (fig. 112), the inscription of which may be ex¬ panded and read as follows :— imp(eratori) caes(ari) t(ito) ael(io) hadr(iano) an- TONINO AVG(VSTO) (PIO), PONT(IFICI) MAX- (IMO), TR(IBVNICIA) POT(ESTATE) XXI., co(n)- S(VLl) IV., coh(ors) II. tvngr(orvm) mil- (iaria), eq(vitata), c(ivivm) l(atinorvm). * The number is now increased by one exposed in the foundations of the church at Hoddom in 1915, {cf. No. 271). 163 MIDDLEBIE.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [middlebie. SVB jv[lio vero] * leg(ato) avg(vsti) pr{o)-pr^etore) ; translating, “ In honour of the Emperor Caesar Titus /Elius Ha- drianus Antoninus Augustus Pius, Chief Pon- of Latin citizenship, under Julius Verus, Legate of the Emperor, as Governor (of Britain).” The importance of this tablet is that it enables us to fix a date, namely, 158 A.D. Certain of the altars give us informa¬ tion as to the garrison at Birrens, and on one (fig. 113), which had been thrown into the well of the Principia, we have a dedica¬ tion to the Discipline of Augustus or the Emperor by the 11 . Cohort of the Tungri, 1000 strong, with its proportion of cavalry, i.e. 760 foot and 240 horse. On one side of the altar are the sacrificial axe and flay¬ ing knife ; on the other is a patera. Erom tiff, invested with the tribunician power twenty-one times, four times consul, [erected MARTIETVICP: TMIiniNCO-l 1lTVNGR-CVI PMEESTSILWS -PRAEF, ^'■1 -n,, I! nil Iiini ”! i/iH Fig. 113.—Altar. by] the Second Cohort of Tungrians, a thou¬ sand strong, with its due proportion of cavalry, and in possession of the privilege * Name supplied from a similar slab found in 1903 at Newcastle (Proc. Soc. Ant., vol. xxxviii. (1903-4), pp. 454 ff.). Fig. 114.—Altar. the nature of the dedication,* and from the character of the lettering, it may be in¬ ferred that this altar dates from the 2nd century a.d. On top it has the usual volutes on each side with rosette ornaments, while the upper part of the cornice is heavily ornamented with details, some of which * The Emperor Hadrian established the " Dis¬ cipline ” cult, but it was continued by Pius, so that a limit to Hadrian’s term is unwarranted. 164 MiDDLEBiE.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [middlebie. suggest the later Saxon work. The focus on the surface is wider and more shallow than usual. Another altar (fig. 114) contains both the name of the Cohort and that of its commander, and is to be read as follows :— MARTI ET VICTORIAE AVG(VSTI). c(iVES) RAETI milit(antes) in coh(orte) secvnda tvn- gr(orvm), cvi praeest silvivs avspex, praef(ectvs), [f(ecervnt)]. v(otvm) s(olve- rvnt) l(ibentes) m(erito) : “ Sacred to Mars and the Victory of Augustus. Rhaetian citi¬ zens, serving in the Second Co¬ hort of Tun- grians, com¬ manded by Silvius Auspex the Prefect, erected this. They performed their vow will¬ ingly, deserved¬ ly.” The altar of the Cohort Nervana is shown in fig. 115. Each of these altars is 4I feet high, with the exception of the first, which is only 3 feet 2 inches. An ex- ample of an altar to Fortune is figured and described in No. 273. Of the dedications to non-Roman deities, two are by architects who, possibly, had been occupied in the planning of the station. One is to Brigantia (fig. 116), the eponymous deity of the tribe of Brigantes, in or near whose land Birrens lay, and is made by Amandus. It will be observed that the figure is in Roman attire, bears the globe, a Roman symbol of Victory, and like Victory is vinged.* The other dedication is on a small altar (fig. 117), 2 feet 3J inches high, with wide mouldings at base and cornice and volutes and focus or hearth at the top, which was raised by the architect Garnidiahus to a strange goddess Harimella. One stone was raised to Afutianus, son of Bassus, by his wife Flavia Bae- tica, and is excep¬ tional in so far as it is the solitary tombstone from the station (see also No. 273). Birrens Station has been identified with Blat obiil- gium in Route II. on the Antonine Itinerary, which in its origin dates probably from the close of the 2nd century. Blatohulgium was 12 miles from Castra Exploratorum (Nether- by). The most important items of evidence are those provided in the relics from the station itself. The date 158, as deduced from the year of the Em¬ peror’s reign, probably points to the period of the rebuilding of the station, which brings it into line with the general reconstruction of the defences in Scot¬ land about that time, of which evidence also appears at Newstead and along the Vallum, as well as in northern England. This recon¬ struction was probably due to a temporary Fig. 117.—Altar, abandonment or loss of these positions consequent upon a successful revolt of the powerful tribe of Brigantes in northern England and on the Scottish border. By the Birrens slab and others from English sites Julius Verus is restored to us as Governor * Other dedications to Brigantia have been found on the English side, one as far south as the neighbour¬ hood of Leeds. 165 Fig. 116.—Brigantia. [ Me HARIMEL ,■( LAE-SACGA) M IDIAHV5 ' ARC■^V^LLAl MIDDLEBIE.] HISTORICAL MONUMENTS (SCOTLAND^ COMMISSION. [middlebie. of Britain at the time. There is no evidence of occupation earlier than the 2nd century, no¬ thing so far that would take the original erection of the station beyond the time of Hadrian (117-13S A.D.), or even necessarily so far. The normal garrison seems to have been a miUiar}-’ cohort of auxiliaries nominally at least from the Rhineland ; a force made up of 760 foot and 240 horse and therefore suitable for an outl3dng post. The Second Cohort of Tungi'i is known also at Castlesteads, Cum¬ berland : and the First Cohort of Germans can be traced both at Burgh-on-Sands and Netherby. Two fragments of slabs indicate the presence at Birrens at some time of legionaries of the Sixth Legion, called “ Vic¬ torious,” whose headquarters were at York ; and a companion inscription of the same legion has now been found at Hoddom in the foundations of a church (No. 271), for which the ruins at Birrens Camp seem to have sup¬ plied much if not most of the building material. Proc. Soc. Ant. Scot.,vo\. xxx., 1895-96, pp. 81-199; xxxviii., 1903-4, pp. 454-59 ; Bir¬ rens and its Antiquities, Macdonald and Barbour, Dumfries, 1897 ; A Roman Frontier Post, James Curie, 1911. hdii. N.W. 16 September 1915. 463. Fort, Stockbridgehill.—On the open moorland 700 feet above sea-level, about J mile to the west side of the Darlawhill Grain and a little over that distance to the east-north-east of Stockbridgehill, is a small rectangular, oblong enclosure, lying with its longest axis east by north and west by south, measuring some 126 feet by 106 feet, surrounded by a slight mound formed from the upcast of a narrow and shallow trench measuring some 18 feet in breadth and not more than 3 feet in depth below the scarp and 2 feet below the counterscarp. The line of the scarp is not very regular, and along the east side it is very ill defined. The angles are rather rectangular ; the interior is not level, there being a fall of some 5 feet from the highest point near the centre to the edge of the trench on the east. The entrance has been from the west, somewhat to the south of the centre, measuring from 8 to 9 feet in width. li. N.E. 14 August 1912. 464. Fort, Birrens Hill, Carruthers.—This fort (fig. 118) is situated on the crest of a ridge at an elevation of 800 feet over sea-level, overlook¬ ing to eastward the valley of the Kirtle Water, to southward the 10 miles of cultivated land that stretches to the Solway, and to westward rolling uplands with the table-like mass of Birrenswark Hill dominating the horizon. Fig. 118.—Fort, Birrens Fiill, Carruthers (No. 464). To the north the ridge inclines by an easy gradient to the summit of Grange Fell, i mile distant. The fort is oblong and approximately rectangular with rounded angles, lying with its main axis north-north-west and south- south-east, measuring 230 feet in length by 180 feet in breadth through the centre, and containing an area just under an acre. The south-east side appeared to be curved some¬ what, and there seemed likewise to be a con¬ siderable convexity on the north-east flank. The angles are rounded with a sweeping curve. A bold rampart surrounds the interior area. 166 MIDDLEBIE.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [middlebie. rising some 3 to 4 feet in elevation on the inner face, covered by a trench, no longer measurable on the north-east, some 29 feet in width from crest to crest along the north¬ west and south-west sides, and 25 feet on the south-east, having a fairly uniform depth of from 7 to 8 feet below the crest of the scarp and of about 4 feet below that of the counterscarp, the scarps being smooth and sharply cut. The entrance has been through the centre of the south-east end. The profiles are considerably impaired along the north-east and south-east sides by secondary constructions and on the latter side also by excavations. To the north angle there stretch across the moorland two broad parallel mounds, some 10 feet apart and with a height of a couple of feet or thereby, having a slighter parallel mound on the north-west. This last is carried across the trench of the fort, up the scarp and along the crest of the rampart, curving inwards towards the position of the entrance as it approaches the south-east end. The central mound of the three terminates on the counterscarp of the trench, while the third is continued along the north-east flank at the base of the rampart, almost entirely filling the trench—which has probably been shallower on this side—and onward beyond the end of the fort, terminating with a right-angled return to the westward. Immediately within the position of the entrance there is traceable the end of a rectangular foundation, the western end of which has been destroyed by quarrying ; and beyond the trench, on the south-west, is another rectangular foundation and also a triangular fold-like enclosure. In front of the position of the entrance is a hollow, and there are several other smaller depressions adjacent, from which stone appears to have been quarried. From these constructions and excavations two parallel mounds again trend away from the fort in a southerly direction. The hollows in front of the position suggest an analogy with the rectangular fort on Gotterbie Moor (No. 451), where deep quarry - like holes passed across the front of the entrance. The ptisition of these holes at Gotterbie, sunk directly on the level and not into the face of the rising ground adjacent, is peculiar. ith regard to those at Carruthers, though some appeared undoubtedly to be excavations for stone, a hollow running down from the front was rather like a natural ravine crossed by a bank near the fort, so as to form an enclosure within it. hi. N.W. 16 August 1912. 465. Fort, Minsca. — In a meadow sur¬ rounded on three sides by woodlands and separated by a deep glen from the farm of Minsca to the south-east, are the remains of a fort. In form it is circular, measuring some 205 feet in diameter. Only on the southern semicircle are the defences preserved to any extent, consisting of an inner mound some 3 feet in height above the interior level and a trench 30 feet in width. The trench has a scarp some 7 to 8 feet in height and, wiiere highest, a counterscarp of 5 feet. On the north-west quadrant the periphery is almost effaced, and on the north only the scarp from the summit level exists. There has been an entrance from the north-east, and to the south of it the rampart makes a bend outwards for some 15 feet. On the south a roadway leads into the trench, flanked by a mound on either side : that on the left continues along the counterscarp, while that on the right runs out across the trench, where there is an altera¬ tion in the general level of the floor. The meaning of this roadway is not apparent, but it may have led by the trench to an entrance through the part of the defences now demolished. lii. N.W. 16 August 1912. 466. Fort, Purdomstown. —- On a bluff immediately adjacent to the south end of the Annan waterworks, near Purdomstown farm and contained in a loop of the Middlebie Biirn, are the rernains of a small quasi-rectangular oblong fort l3fing with its longest axis north and south. The south end rests on a steep bank, some 20 feet in height, overlooking the stream. On the east the ground from the fort falls sharply away ; on the west side lies a hollow, gradually deepening as it approaches the face of the bank on the south ; towards the north the original contour of the ground and the lines of the fort have been destroyed by the waterworks. Along the east and west sides are the remains of an earthen rampart, showing a breadth at base of some 16 feet and low in elevation, and there is an indication 167 MIDDLEBIE.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [middlebie. of its continuance along the south end. The length of the enceinte, as remaining, is 173 feet; its breadth at the south end no feet, and at the north about 126 feet. The situa¬ tion is about miles north-north-west of the Roman fort of Birrens ; and, though there is no positive endence of a Roman origin, the fort’s position and the regularity of its lines are such as to make it worthy of notice from this point of \ iew. lii. S.W. (unnoted). 24 April 1913. 467. Dunnabie Fort.—About J mile due north of the farm of Dunnabie, on a some¬ what level bench on the hillside 150 yards back from the highway, near the junction of the roads from Langholm and Waterbeck, and at an approximate elevation of 60 feet above it, is a fort. It lies on the south side of a small glen down which there flows a burn. The form is approximately circular but is flattened on the north by the glen of the burn ; the fort has been surrounded by a broad earthen rampart, now much ploughed down, and, except on the south, by a trench. Where the trench opens on to the bank of the burn, on the upper and lower faces of the enceinte, it is broader and deeper than elsewhere on the periphery, showing on the west side a breadth from crest to crest of 42 feet and a depth of 16 feet below the top of the scarp. The interior diameter has been some 132 feet. lii. N.W. (unnoted). 24 April 1913. Exxlosure. 468. Enclosure, Muckle Snab Hill.—On the east face of Muckle Snab Hill, at an elevation of some 900 feet over sea-level as it slopes down to the glen of Williescleuch, is a frag¬ ment of a circular or oval enclosure surrounded by a low stony bank. Within the interior area, on the lower side, is an artificial-looking hollow, measuring some 50 feet in diameter, and excavated to a depth of from 2 feet to 5 feet as the ground rises from one side to the other. xliv. S.E. (" Fort ”). 31 July 1912. Miscellaneous. 469. Gravestone, Pennershaughs Graveyard. —On the north side of Pennershaughs grave¬ yard, situated on the Carlisle road about i mile to the east of Ecclefechan, there lies, close to the edge of a steep bank, a grave-slab, 6 feet 3 inches in length by i foot 6 inches in breadth, slightly diminishing to the foot, and 9 inches in thickness, on the top of which is incised a double-armed cross, i foot i inch in length, crossed horizontally by arms 5 inches and 4 inches in length ; in the centre of the stone appear a ploughshare and pruning-hook. The stone bears no date or inscription. Ivii. N.E. 14 August 1912 470. Carruthers Old Churchyard.—In the wall of the old graveyard of Carruthers, situ¬ ated some 200 yards south by east of Crowdie Knowe farm, and in the lowest course on the right side of the gate, is a slab, 5 feet 4 inches in length by i foot 7 inches in breadth, on which is incised an equal-limbed cross of Maltese form, i foot 7 inches in length, set upon a stem, the whole 4 feet 5 inches long. Forming the gate-post on the left side of the entrance is a richly-carved slab, partially hidden by an iron stop for the gate. On the south-west corner lies inverted a font-like basin in the form of a scalloped capital suggestive of Norman design, which measures i foot 7 inches across the base, I foot across the top, and i foot 2 inches in height. lii. N.W. (Graveyard, unnoted). 4 October 1912. 471. Standing-Stone, Winterhopehead.—On a knoll on the left bank of the Kirtle Water, some 300 yards to the south-west of the farm, there has been recently erected a standing- stone some 5 feet 9 inches in height above ground, which was brought from a spot where it originally stood in a moss about ij miles to the north-east of its present position. The two large stones, which may be seen on the moor a short distance to the north-east, have also been recently set up on end. There is no evidence to warrant the supposition that they have formed part of a stone circle. xliv. S.E. (unnoted). 24 April 1913. 472. Templehill, Waterbeck.—To the east of the house of Templehill, at Waterbeck, and rising from the edge of the burn, is a large flat-topped mound, which has certainly MIDDLEBIE.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [MOFFAT. been shaped to some extent by art. It does not seem to be a mote-hill, as its elevation on the side towards the house is only a few feet, though on the other three sides it has something of that form. lii. S.W. 24 April 1913. Sites. The O.S. maps indicate sites as under :— 473. Carruthers Village, Carruthers. lii. N.W. 474. Pennershaughs Church, Burnfoot. Ivii, N.E. MOFFAT. Castellated and Domestic Strlxtures. 475. Breckonside Tower,—Breckonside (fig. 119), an oblong tower of the late i6th century, is situated in a secluded val¬ ley 2^ miles south-south¬ east of Moffat. It now forms part of a farm-house, and apparently the ground floor is the only portion which remains in its original state. This floor is vaulted and is now divided into two apartments, each lit by a window, probably later in¬ sertions. There is a wheel- staircase in the north-west 4)i u ii) u | -angle, hrom north to south Fig. ii9.-Breckonside ^^e tower measures 36 feet Tower (No. 475). and from east to west 22 feet. The walls are 3 feet 6 inches to 4 feet 6 inches thick. The building is in occupation and is in a fairly good state of preservation. James Johnstone of “ Brakenside ” is among the Johnstone “friends” in 1590. {John¬ stone MSS., No. 65). xvi. S.E. 8 May 1912. 476. Cornal Tower.—The ruins of Cornal ToM'er stand on the southern bank of the Cornal Burn, about if miles south-east of Moffat. The site has evidently been chosen with regard to its defensive capabilities. It is a plateau, in shape roughly oval, measuring from north to south 50 feet and about 100 feet from east to west. On the north it is steeply scarped down to the burn, while the ground on the three other sides is considerably lower than the plateau. The tower stood at the eastern extremity of the site and has evidently- been a tower of the type common to the dis¬ trict in the i6th century. It measures 21 feet from north to south and 37 feet from east to west. Only a portion of the north wall is now left. It is 12 feet 6 inches long, 12 feet high, and 5 feet thick. Twenty-nine feet west of the tower is an out-building, 46 feet 6 inches from north to south and 22 feet 6 inches from east to west. The walls are 3 feet 10 inches thick. The name Cornal is a shortened form of the old Polcornal or Pocornal, usually associated with or included in the Estate of Logan. “ Logane tenement and Polcornare {sic) ” appear in a precept of sasine following on a charter by James IV. in 1512 granting to Symon Carruthers of Mouswald among other lands “ ten mark lands in Logane tenement.” {Bucclench MSS., No. 120. Proc. Soc. Ant. Scot., 1888-9, P- 50)- Simon Carruthers left only two daughters, and his lands passed from them to Douglas of Drumlanrig {Bucc. MSS., No. 126; pp. 62-5. See further No. 517). xvi. S.E. 8 May 1912. 477. Blacklaw Tower. —The ruins of Black- law Tower are situated on the west bank of the Blacklaw Burn, 2 miles west-north-west of Moffat and 200 yards east of the high road to Glasgow. The tower is oblong on plan, and measures from north to south 32 feet and from east to west 25 feet. The walls are 5 feet thick. Only the vaulted lower storey now remains. It is lit by narrow^ windows in the north and south w^alls. The entrance, off wLich the new'el-stair enters, is in the east w'all. The tow’er is enclosed by out-buildings and yards, some of wiiich are probably secondary. A courtyard extends to the north-east, with an entrance in its north wall. North of this are the abutments of a little bridge over the burn. The enclosure to the north-w'est may have been a garden, and the space between this and the south-west enclosure another entrance. Entrance to the tow^er can only be obtained by passing the windows in the north and south walls. The building appears to 169 MOFFAT.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [MOFFAT. have been erected in the i6th century, and in a bad state of repair. The vault at the southern end has already fallen in, and other portions are likely to do so in the immediate future. James Johnstone “ in Blacklaw ” (1592) is probabh’ a reference to this tower. { John¬ stone MSS., p. 34). xvi. N.W. 8 May 1911. 478. Raecleugh Tower.—The ruins of Rae- cleugh Tower lie some 20 ^^ards north-east of the present farm of that name, on the Evan Water, 5 miles north-west of Moffat. From the fragments of walling visible the building seems to have been oblong on plan, with a vault covering the ground floor. From north to south it measures some 36 feet and 21 feet 9 inches from east to west. The walls vary from 3 to 5 feet in thickness, and are roughly built vdth the whinstone of the neighbourhood. Robert Johnstone of “ Raecleuch ” appears in the first quarter of the 17th century. {Johnstone MSS., pp. 41, 42, 47). ix. N.W. 7 May 1912. 479. Mellingshaw Tower. — Mellingshaw Tower stood on the banks of the Mellingshaw Burn, some 300 yards from the junction of the burn with the Evan Water and 3-4 miles north-west of Moffat. Little now remains of the building save the south-east angle, which contains a small wheel-stair lit by two small windows, one above the other, in the east wall. This wall is 8 feet 6 inches long and 2 feet 4 inches thick. The height of the angle is some 15 feet. In the wall is one jamb of the door. There is now no evidence to show the form or extent of the tower. It belonged to a branch of the Johnstones, as “ Malynshaw ” or “ Mallingshaw,” in the i6th century. {Cal. Scot. Papers L, No. 396 ; Johnstone MSS., p. 22). ix. S.W. 7 May 1912. 480. Frenchland Tower.—Frenchland Tower (fig. 120) stands on the banks of the Frenchland Burn, less than a mile east of Moffat. The build¬ ing was originally oblong on plan, unvaulted, with a parapet carried on corbels, and was probably built in the i6th century. At some ^:Jr} later period, possibly in the 17th century, the tower was modernised, the parapet was re¬ moved, and a high-pitched roof substituted. A wing was thrown out on the west—making the tower a mansion of the L-plan type—to contain a staircase wider and more suited to the requirements of the dwelling than the original wheel-stair. As usual, this wide scale-and-plat stair only reached to the first floor, the wheel-stair serving the upper floors. The upper portion of the wing is divided into two floors — possibly used as bedrooms. The tower measures some _ 26 feet 6 inches by 21 feet ^ 6 inches, and the walls are 120.—Frenchland 3 feet 6 inches thick on an Tower (No. 480). average. The wing projects 15 feet 6 inches from the west wall, and is 13 feet broad, with walls 2 feet 6 inches thick. The wall-head is some 20 feet from the ground. Large portions of the east and west walls have fallen, and the general condition of the other walls is unsatisfactory. The tower is named from the early owners of the property, Franche, or French. “ Adam Franche of Franchland ” flourished in 1535. {Johnstone MSS. p. 18). But the line is of much older date ; William Franciscus (French) in the late 12th century and Roger his son early in the 13th century held land from the Bruces in the neighbourhood of Moffat, no doubt including that which still bears the name. (Bain’s Calendar L, No. 705). xvi. N.E. 7 May 1912. Defensive Constructions. 481. Fort (remains of), Tail Burn.—On a plateau forming the eastern bank or extended channel of the Tail Burn, adjacent to the spot where it debouches on the Moffat Water, are the fragmentary remains of a curvilinear fort. The position on the top of a 20-foot bank, which now lies some 30 feet back from the actual bed of the burn, commands a view up and down the glen of the Moffat Water ; while to the north-west the burn, dashing over 300 feet of cliff, forms the Grey Mare’s Tail Water¬ fall. Of the fort all that remains is a segment of a rampart, concave towards the edge of the 170 MOFFAT.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [MOFFAT. bank, and a trench outside ; but of the area which these defences enclosed there remains at no part a greater breadth than 12 feet, to such an extent has the plateau been eroded by the stream. The rampart extends for some 125 feet measured along the curve, is stony and is covered by a trench some 20 feet in width from crest to crest, 6 feet deep below the crest of the scarp and from 2 to 3 feet below that of the counterscarp. Many large stones lie along the edge of the counterscarp, but they do not show signs of construction. A bank, extending from the south end of the rampart towards the road for a distance of some 50 or 60 feet, is traceable over the trench and is probably secondary. iv. S.E. 16 September 1912. 482. “Meg Tod’s Mote,” Moffat.—About 300 yards soixth of the Auldton mote, and 100 yards or thereby to the east of Ball- play Road, are the remains of a small oval enclosure on the summit of a hillock which rises some 15 feet to 20 feet above the lowest level of the field, with steepish slopes to north and south. The enclosure has lain with its longest axis east and west and measured in diameter 66 feet by 48 feet. The summit has been levelled, but under the action of the plough all other traces of art have disappeared. xvi. N.E. 24 September 1912. 483. Mote, Auldton, Moffat.—The mote of Moffat (fig. 121) lies to the east of the Birnock Water, and about mile to the north-east of the parish church, engirt by a semicircular set¬ ting of hills. It is of the mote and bailey type —the bailey or base-court resting on the edge of a former river bank, some 10 feet in height on west, south-west, and south-east, while the mote has been constructed on rising ground to the north-east of it. The mote rises to a height of 28 feet, is steeply scarped all round and measures in diameter on the summit some 37 feet by 32 feet. Unfortunately at some period, probably in complete ignorance of the true character of the construction, excavation has been conducted on the summit, vith the result that it has been hollowed to a depth of 6 or 7 feet, while the material through a break on the south-west forms a talus over the trench at its base. The mound is entirely surrounded by a trench, some 30 feet in width, at the level of the crest of the counterscarp, which has a height of some 6 feet on the south, and, where passing through higher ground, on the east a height of 8 feet. Where the trench cuts off the mote from the base-court it is now shallower. At the north end of the base-court a deep trench, 34 feet broad, ii feet deep Fig. 121.—Mote, Auldton, Moffat (No. 483). below the crest of the scarp, and from 5 to 6 feet below that of the counterscarp, springing from the trench which encircles the mote, cuts it off from the plateau, continu¬ ing along the bank to the northward. Over¬ looking this trench a breastwork runs along the base-court with an elevation of 5 feet on the inner side. The base-court, which is approximately semi-lunar in form, measures 196 feet from north-west to south-east in longest diameter and 118 feet at right angles to that line across the centre. xvi. N.E. 20 September 1912. 484. “ Mote,” Granton.—On the west side of the road to Ericstane, some 300 yards beyond the entrance to Granton, is a gravel hillock which rises somewhat abruptly to a height of from 12 to 14 feet. From the highest point at the south end it tails away for some 80 feet or 90 feet towards the north-north-east, with a round-backed and somewhat narrow ridge. The situation is on absolutely level and low-lying ground with the steep slope of the eastern w’all of the valley mounting up¬ wards some 50 yards distant from it. There 171 MOFFAT.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [MOFFAT. is no surrounding trench, and the character and situation in no way suggest a mote ; the hillock is probably natural. ix. S.E. (“ Mote ”). 17 September 1912. 485. Fort, Auchencat Burn.—This is a large semicircular or horse-shoe-shaped earthwork Fig. 122.—Fort, Auchencat Bum (No. 485). which lies on the north side of the Auchencat Burn, about J mile above the spot where it is crossed by the road to Ericstane. Its situa¬ tion is the top of a lofty bank, which has a steep and, in places, weU-nigh precipitous slope down to the burn some 70 to 80 feet below. The defences consist of a massive rampart of earth and stone, bold all round except towards the edge of the ravine, rising to a height on the interior face of some 5 to 6 feet and covered in front by a trench, the floor of which lies at deepest some 8 feet below the crest of the rampart and 4 to 5 feet below that of the counterscarp. The trench has an average width from crest to crest of 33 feet and appears to have been dug entirely out of soil, the counterscarp having a gentle slope and giving no evidence of rock cutting. The base of the fort, with the defences along the edge of the bank, measures 242 feet, and the bisectional diameter is 136 feet. The entrance, with an approximate width of 8 feet, has been a short distance in from the edge of the ravine on the west side, past the banked-up end of the trench on the north and through the rampart, which, to the south of it, has evidently taken a sharp turn eastward along the edge of the ravine, at this point less steep than elsewhere. ix. S.E. 17 September 1912. 486. Fort, Ericstane.—About J mile to the west of Ericstane, at an elevation of some 750 feet above sea-level, is a fort. It is formed on the edge of the steep northern bank of the ravine, down which flows the Braefcot Burn to the Annan some 400 yards distant, and has been constructed from unusually adapted natural features. The enceinte is a semi-oval with its chord east and west, lying along the edge of the ravine and falling considerably in level to the for¬ mer direction. It measures in¬ teriorly 240 feet by 129 feet and is enclosed by a massive rampart, seemingly a structureless mass of small stones, rising some 5 feet to 6 feet above the level of the interior area. In front of this rampart is a trench, 36 feet wide, 8 feet deep below the crest of the scarp, and from 5 to 6 feet below that of the counterscarp. To this extent the defences appear to be in the main artificial; but beyond this a remark¬ able conformation of the ground has been utilised to provide on the upper half of the fort 172 MOFFAT.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [MOFFAT. two outer ramparts, separated by a trench¬ like hollow, with a second trench to the outside, which are transformed on the lower half of the periphery, through the convergence of the two hollows and the consequent elimination of the outer rampart, into a single mound and trench. This trench or hollow passes round the lower end of the fort and opens on to the face of the ravine. These outer mounds and hollows are clearly discernible as natural features on the higher ground above the fort; and, as they come into its immediate vicinity, they have been dug out and shaped so admir¬ ably as to substitute and simulate artificial lines. The entrance, some 8 feet wide, has entered the outer hollow at its lower end and, probably passing up it for some 50 yards, mounted to a little plateau at the south end of the enceinte, whence it lead to an opening in the east end of the inner enclosure. ix. S.W. 17 September 1912. 487. Fort, White Hill, Ericstane.—This fort occupies the crest of a ridge l5dng north and south, rising slightly from the east and falling about 100 feet by a long gradual descent to the bottom of the Annan valley on the west : to the northward there is likewise a consider¬ able gradient to the bed of the Stotfield Gill. It lies I mile north-north-east of Ericstane and at an elevation of 700 feet over sea-level. The interior of the fort is oval, lying with its longest axis north and south, and measures 176 feet by 143 feet. It is surrounded by two stony mounds, not truly concentric, the distance over the flat intervening space being 26 feet and 15 feet on the north and wxst respectively and 40 feet on the south and east ; to the outside is a trench, with an average width of 20 feet from crest to crest and a greatest depth of 4 feet below the crest of an outer mound, which forms the counterscarp. The inner mound has an elevation of not more than 3 feet ; while the outer mound, with a similar elevation on the inner face, presents a scarp to the trench varying from 4 to 6 feet in height. There are indications of stonework on the crest of the ramparts. The interior area has, in places, been hollowed by excavation, especially towards the south end. The entrance, some 7 feet in width, has been on the east side into the lowest part of the area, passing through the line of the trench, which is banked up on either side, and directly through the rampart, being seemingly flanked by a wall on either side as it crosses the interspace. So as to increase the length of the passage, the rampart, forming the counterscarp of the trench, makes a bend inward on either side. On the right or north side of the entrance various irregular mounds and heaps of stone may indicate the existence of structures of some kind between the ram¬ parts ; and around the base of the inner rampart, at the south end, there is much stone lying, overgrown with turf, which may have a similar indication. ix. N.E. 17 September 1912. Ex’CLOSURES. 488. Enclosure, Roundstonefoot.—About I mile to the north of Roundstonefoot, and partially destroyed by the road up Moffatdale, are the remains of an oval enclosure. It now consists of little more than a semi-oval, the chord being marked by the wall at the road¬ side, measuring 108 feet by 60 feet and sur¬ rounded by a massive stony rampart, with an elevation of some 5 feet at highest on the exterior and not more than 3 feet on the inner side. The interior lies at two distinct levels— the north half ha\ang apparently been hollowed by excavation to a depth of several feet. The entrance, some 6 feet wide, has opened into the low'er area from the north-east. X. S.W. (“ Castle,” remains of). 16 Sep¬ tember 1912. 489. Enclosure, Selcoth.—This enclosure is situated on a point running out into the valley of the Moffat M'ater, on its eastern side, and some 250 yards south-w^est of Selcoth. It is pear-shaped in form, sur¬ rounded by a single broad stony rampart, wEich, except wEere it crosses the neck of the promontory on the east, follow's the contour at the edge, from wEich there is a fall of from 20 to 40 feet to the floor of the valley. The longest axis lies from south-east to north¬ west, and the interior measures some 191 feet by 124 feet. The rampart, some 12 feet broad at base, has a height of 4 feet at most above the interior and nowEere more than 3 to 4 feet on the exterior. The interior area shows evidence of hollow'- 173 MOFFAT.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [MOFFAT. ing by excavation, especially at the upper or south-east end, where the floor lies some 6 feet below the natural ground-level outside. At the extreme point within the rampart, near the highest part of the defences, there is an oblong enclosure measuring 44 feet by 22 feet, the inner wall of which appears to be a bank of unexcavated soil and rock now surmounted by a 'modern stone dyke, and which is pierced at no \'isible point by an entrance to the larger enclosure. The main entrance has been from the east-north-east into the lowest area of the interior. Crossing the interior diagonally to¬ wards the entrance, and cutting the area into two divisions, is a broad bank, which seems to have been formed of natural ground left by excavation on either side. xvii. N.W. (“ Fort ”). 16 September 1912. 490. Enclosure, Hunterheck.—This enclosure occupies the summit of a plateau in an angle formed by a bend of the Frenchland Burn, as it changes its course from a westerly to a southerly direction to the north of Hunter¬ heck cottages, and appears to have been irregu¬ larly circular in form, measuring 183 feet by 164 feet in diameter. With its north arc rest¬ ing on the glen of the burn it has been sur¬ rounded, except at the north-east, as after- mentioned, by a broad stony mound rising at most some 3 to 4 feet on either face. On the north-east this mound, instead of being carried forward to the edge of the glen to complete the circle, is turned away sharply to the eastward for a distance of some 66 feet, terminating at 23 feet back from a steep bank l}flng parallel and falling in the direction of the burn. The interior has been to some extent hollowed by excavation, the floor level on the south being some 3 feet below that of the ground immediately outside. It has been crossed by a broad bank from east-south-east to west-north-west, cutting off about | against the south arc, at the west end of which a circular hollow appears to have been formed, measuring some 60 feet by 52 feet in diameter. Another cross-wall runs in a north-easterly direction from a point somewhat to the east of the centre of the main divisional bank and forms a triangular enclosure, against the east arc of which, however, the north angle is unclosed. The entrance, 6 feet wide, has been from the south-west, flanked on the left by an inward return of the rampart for a distance of some 12 feet; from it to the French¬ land Burn on the west a roadway is traceable, where its course has been cut through oppos¬ ing rock and down the bank of the burn. Beyond the enclosure, on the point of the plateau to the north-west, is an area which appears to have been hollowed ; and some 250 yards to the eastward, at the base of the rising ground, towards the upper end of the held and close by the bank of the burn, are a number of indeterminate foundations. xvi. N.E. (“ Fort ”). 16 September 1912. 491. Enclosure, Auldton.—On a bench on the hillside, about J mile due east of Auldton and 100 feet above it, are the remains of a cir¬ cular enclosure, which has been surrounded by a stony rampart or wall and measures, with its longest axis north and south, 96 feet by 78 feet in diameter. The position is completely commanded by the hill rising abruptly behind it. The inner face of the bank on the east has been formed by excava¬ ting the interior to a depth of 5 feet. xvi. N.E. (“ Fort ”). 20 September 1912. 492. Enclosure, Corehead.—On the brow of a ridge which forms the end of the watershed between two burns coming down from Cock- law Knowe and Spout Craig respectively, at an elevation of 900 feet over sea-level and about I mile to the south-east of Corehead, is a circular enclosure, measuring interiorly 152 feet by 138 feet, formed with a single rampart, composed of small stones, now reduced to a low level towards the exterior, but, owing to the hollowing of the interior, having an eleva¬ tion on the inside towards the higher level of from 3 feet to 4 feet. Against the south-east arc on the interior lie some low heaps of stones, irregularly circular, suggestive of hut foundations ; and on the north-north-west there is an elevated circular platform with a diameter of some 40 feet, to the south of which appears to have been the entrance coming from the west. ix. N.E. (“ Fort ”). 17 September 1912. 493. Enclosure, Meikleholmside. — On a plateau on the western slope of the Annan 174 MOFFAT.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [MOFFAT. valley, some | mile north-north-west of Meikle- holmside and at an elevation of 700 feet over sea-level, are the remains of an oval enclosure, lying with its main axis north-north-west and south-south-east and measuring in diameter some 149 feet by 113 feet. It has been sur¬ rounded by a single stony mound, now of low elevation and much spread. The interior rises to a ridge crossing the centre from north to south, and the lower part of the area on the west is completely overlooked by the higher ground outside and immediately adjacent. On the east there is a steep ascent from the plateau to the bottom of the valley. ix. S.W. (“ Fort ”). 24 September 1912. Miscellaneous. 494. Foundations, Frenchland. — In the second field, to the north of Frenchland Tower, about J mile to the east of Moffat, are a number of foundations shown in ridges over the turf and covering several acres in extent. Some appear to be those of houses, while others are of larger enclosures, and there are also areas hollowed by excavation. Some of the wall founda¬ tions are very broad. An old roadway is clearly discernible leading into the group from the direction of Frenchland Tower. It is said that these lands belonged to the Knights Templar, and it is assumed that here are the remains of some settlement belonging to that body. xvi. N.E. (“ Site of Walls, Chapelry of Knights Templars ”). 16 September 1912. 495. Sundial, Ericstane.—Standing within the garden at Ericstane is a sundial which is said to have come originally from Wamphray Place. From a base 2 feet square, which has originally been moulded at the upper edge with half-round vertical mouldings at the angles, there rises a shaft, a square of 8 inches in section and 2 feet in height, supporting a cushion cap, on which rests the table, incised with the dial, a double hexagon in form, i foot 9I inches in diameter, each facet measuring 6 inches across, and, except those respectively facing north and south, bearing a cup-shaped hollow in the centre, 3| inches in diameter, crossed obliquely by a thin iron rod. The hollows facing north and south have been sunk with a straight-edge to a depth of J-inch, and have had some metal object projecting from the centre. The shaft has been carved on its four sides. On the south, incised at the top are the names— Robert Johnstoune— Isabella Rollo. Beneath, in relief, two coats-of-arms, placed vertically, blazoned thus : —The upper—a saltire, on a chief three cushions (tasselled) ; Crest—a fox. Motto— Nunquam non paratus (Johnstone). The lower—a chevron between three roes’ heads erased. Supporters—two stags. On a baron’s coronet a crest—a stag’s head. Motto—“ For¬ tune mate for fortune” (RoUo). (The first two words of the motto are doubtful). Both shields are enriched with mantling. At base the date 1701. The west face bears the table of the sun’s rising and setting, with the names of the months and the hours : the north two columns, evidently of hours and minutes. On the east the table of the golden numbers. The names inscribed are those of the last laird of Wamphray and his Mufe, who was a daughter of the third Lord Rollo. The dial is supposed to have come from Wamphray Place, which ceased to be a residence about the middle of the i8th century. Previous to its being removed to Ericstane, some thirty- two years ago, it stood for many years in a garden in Holm Street, Moffat, that had belonged to a family of clockmakers. ix. S.E. 17 September 1912. 496. Bell, Moffat.—Serving as the bell for the town clock of Moffat is a bell taken from the Old Tolbooth, measuring in height to the canons i foot 7 inches and in diameter at the mouth i foot 4 inches. Braised on to the sides, a little above the middle of the bell, are a number of small rectangular plates, containing initials and a coat-of-arms, the former measuring about i inch in height and the latter 2| inches in height by if inches in breadth. The initials, in relief, are W E I. L and I H. ; the date is 1660. The arms are those of Lord Hartfell, by whom the bell is said to have been presented, and are blazoned thus:—A saltire, on a chief three cushions, for Johnstone—all beneath an Earl’s coronet. xvi. N.E. 24 September 1912. W5 MOFFAT.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [MORTON. 497. Moffat Churchyard.—Built into the re¬ maining fragment of the south wall of the church which stands in the old churchyard, beneath a raking cornice, are two shields bearing arms :— Dexter, beneath the initials A I, a saltire, three cushions in chief, in base a heart (Johnstone). Sinister, beneath the initials N.D. three piles, in chief three mullets, in base a heart (Douglas). Beneath, the date 1665. xvi. N.E. 24 September 1912. Sites. 498. Earthwork, near Bridgend. — On the north side of a small plantation, about 150 yards to the north of the school at Bridgend, on a slight eminence, are the very indefinite traces of an artificial earthwork. ix. S.E. (“ Eort ”). 17 September 1912. 499. Forts, Well Burn, Moffat.—On the rising ground immediately behind the well-house at Moffat, and bordering' on the west bank of the burn, is the site of a fort. The remains consist of a segment of a stony rampart towards the edge of the burn, with a portion of what appears to have been a semi-lunar outer court at the south end. The interior does not appear to have been hollowed by excava¬ tion. On the higher slope, about 100 yards to the north-west and about 50 feet higher in eleva¬ tion, are the remains of another fort, which appears to have been enclosed by a stone wall. The outline is now difficult to follow, but there is evidence of two or three subsidiary hollowed enclosures at the north end, in one of which a large quantity of road metal has been collected. xvi. N.E. 24 September 1912. 500. Fort, Craigieburn.—To the west of the entrance to Craigieburn, and occupying a hillock on the east side of the burn, there has been a large oval fort, whose outline can now only be made out with difficulty. To the north of the cottage and kennels which stand upon the site, there is traceable a portion of a trench, and a broad rampart is discernible curving around the meadow in front. The entrance has evidently been up the slope from the bottom of the valley into the south end of the fort by a track passing obliquely upwards. xvi. N.E. 16 September 1912. 501. Enclosures, Crofthead and Cornal Burn. —The enclosures at both these localities may be noted as sites. That at Crofthead is shown on the 25-inch O.S. map as having been circular, with a diameter of about 150 feet. xvii. N.W. and S.W. (" Eorts ”). 16 Sep¬ tember 1912. The O.S. maps also indicate sites as under :— 502. Tweeds Cross, Flecket Hill. ix. N.W. 503. Fort, Roger Gills, ix. N.W. 504. Fort, near Moreland, ix. S.E. 505. Fort, Hunterheck Hill. xv. N.E. 506. Fort, near Moffat Well. xvi. N.E. 507. Fort, near Archbank. xvi. N.E. 508. Fort, near Auldton, Moffat, xvi. N.E. 509. Fort, Wait Hill. xvii. S.W. MORTON. Castellated and Domestic Structure. 510. Morton Castle.—In the hill country of Upper Nithsdale, some miles south-east of Carronbridge Station and 2| miles north¬ east of Thornhill, the ruin of Morton Castle stands on a promontory at an elevation of 6 o 2| feet above sea-level overlooking a small loch which has been artificially embanked on the east side. The flanks of the site are thus defended by water, and it only remained for the engineer to throw a strong screen of building from flank to flank to secure the position. Whether a ditch returned before the building is indeterminate, but the con¬ figuration of the ground at either corner shows that at these points at least there was this feature. The castle is a 15th-century structure and possesses a gatehouse on the west, entered between D-shaped towers, from which a curtain returned eastward to a circled tower on the east angle, then northward, probably to the apex of the site, where there may have 176 MORTON.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [MORTON. been a second circled tower, and thence probably returned to the gatehouse. The area thus enclosed is triangular, has a fall from south to north, and measures i6o feet from north to south by 133 feet from east to west. The southern half at least was occupied by structures in which the curtains were incor¬ porated. What remains of these are the eastern tower and wall of the gatehouse, a small portion of the eastern angle tower and, between these, the main wing, a structure of two storeys containing the principal living apartments of the castle, which is fairly complete to the wall-head of the south curtain Fig. 123.—Morton Castle (No. 510). The stone employed is a coarse warm- coloured granite, the masonry is of ashlar, built in II to 15 inch courses, the stones averaging 20 inches in length. The curtain walls are battered exteriorly towards the base. Owing to the inclination of the enceinte and the level of the basement floor, the sill of the entrance is about 10 feet above the ground level and one course below the upper edge of the batter on the wall. As shown by the details of the remaining jamb, the entrance was defended by a drawbridge, which gave access over a built fosse extending 7I feet within the entrance and 16 feet out from the front wall, with a width of about 8 feet, and which, when raised, lay within a recess against the front of the building, forming an addi¬ tional barrier. Immediately within this was a gate opening outwards. The actual en¬ trance was through a pointed arched doorway, beyond which was the portcullis, while an inner gate admitted to an arched pend, which penetrated the gatehouse as at Caer- laverock and was apparently the only vaulted portion of the building. Above the entrance lay a chamber from which the portcullis and drawbridge were raised. The gatehouse tower, as now standing, is hve storeys high. The lowest storey is much below the basement floor level, with which it communicated by a hatch. It is D-shaped internally and consists of one chamber ; the only opening in the walls is an upward flue for ventilation in the south wall, which suggests the use of this chamber as a prison. The ceiling was formed of wooden joists and flooring, the former borne on pro¬ jecting corbels. The floors above are similar in shape. The basement floor has a window looking east and another to the south. These have stepped soles and shouldered scoinson heads. A squint in the west wall commands the entrance pend. The first floor was entered from the adjoin¬ ing main wing. It has a window to the south with a seat in its west jamb and a garderobe in its east jamb with a soil flue discharging from the south wall. The scoinson arch is ribbed and segmentally pointed. A window to the east has a shouldered scoinson head and stepped sole. A fireplace in the west wall has had a projecting hood and jambs. The floor joists of the room above were borne on a scarcement on the south and east and by a large projecting corbel over the fireplace on the west. Access to the second and third floors, which call for no special mention, was by a wheel-staircase starting at first floor level and contained within the walls at the junction of the south curtain with the gate¬ house. The arrangement of the circled tower in the east flank has been somewhat similar and need not be described. The main wing between these towers is 30 feet wide from north to south and has a mean length of 93 feet, all within walls 4^ to 6^ feet in thickness ; at its western end there is a triangularly-shaped area between it and the entrance passage, of which in its present state nothing can be said. Against the outer face of the north wall there have been two structures ; the western has been of one storey, as evidenced by the corbelling 12 177 MORTON.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [MORTON. and weather-table of the lean-to roof, and had no internal communication with the main wing. The eastern was of two storeys and, on the upper floor only, communicated with the main wing through a window with widely- spla^'ed ingoings and shouldered scoinson heads, the northern being double shouldered. This opening appears to be too small for a doorway and probably was provided only for the supervision of the outbuilding. The basement of the main wing has a slight scarce- ment along the north and south walls. The windows, 15 inches wide and 12 inches high, are placed eight to the south and two to the east, all set at a considerable height above the floor level. The soles are stepped, the heads shouldered. Three have had stone basins with slop drain discharging through the south waU. Between the east windows is a fireplace with projecting hood and jambs. The south-east angle contains the entrance to an angled passage admitting to the east tower. In the floor of the passage is a hatch, probably a soil outlet, set over a drain cutting through the gorge of the tower and having outlet to the exterior at each end. On the south the outlet has been too small for the admission of a human being ; the external batter is cut away as though to permit the passage being sluiced. At the northern end it has sufficient height to permit entrance, but a gate or grate was placed in a chase immediately north of the hatch to prevent the opening being used for this purpose. This was necessary, as there are instances of castles ha^■ing been captured by a surprise entrance through such a passage, e.g. Chateau Gaillard. The first floor was entered from a doorway in the north wall at its west end, which was reached from a wooden forestair and gallery. The doorway has a pointed arched head with arch-and-hood mouldings dying on a “ discontinuous impost ” ; the jambs have three contiguous rounds—derived from the engaged shaft—and are received on con¬ tinuous splayed bases. Adjoining the door¬ way to the west are a window and the jamb of a second. These have shouldered heads and splayed arrises. East of the doorway, at about the mid-point of the north wall, there has been another window, and beyond it to the east is the window already described as communicating between the main portion and outbuilding. The south wall contains a fireplace with projecting hood and jambs and four lintelled windows with mullions and transoms ; that on a slightly lower level is a later insertion. These windows have been heavily barred, the scoinson arches are pointed and ribbed and there have been seats in the ingoings. The Fig. 126.—Morton Castle: doorway at first floor level (No. 510). east wall contains two windows, the northern is similar to those in the south wall, the southern is smaller. At the south-east angle a doorway leads to the east tower. The roof has been an open timber roof with braces beneath the beam. The ruin is in a fair state of preservation. The erection of this castle probably followed on the acquisition of the barony of Morton by James Douglas of Dalkeith in 1440. (Cf. Introd., p. xxvi.). xxii. N.E. 25 July 1919. ‘ 178 A}tcieut and Hisiorical Momtments — Dumfries. Fig. 124.— Morton Castle : Gatehouse Tower (No. 510). Fig. 125.—Morton Castle : Interior (Xo. 510). To face p. 178 . MORTON.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [morton. 179 Front. Back. Right side. Fig. 127.—Cross-shaft from Closeburn, now at the Grierson Museum, Thornhill (No. 514). MORTON.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [mouswald. Defensive Constructions. 511. Fort, Morton Mains Hill.—Morton Mains Hill is a prominent grassy eminence, which rises to an elevation of 1076 feet above sea- level to the north of Morton Mains farm. The east side of the hill has a very abrupt declivity to the Kettleton Burn, while the gradient from the other points of the compass, though less severe, is also fairly steep. Across the south end of the summit and along the west flank, at an elevation some 20 feet below the crest, runs a rampart, formed evidently from the upcast of a slight ditch facing it to the exterior. This rampart and ditch are not carried round the north end of the summit, nor are they elsewhere continuous, though the one always accompanies the other. On the south front there are no less than three wide gaps, the widest 35 feet across, where neither trench nor rampart has existed. Both run continuously along the w^est for a distance of 380 feet. The peculiar imperfection of this defensive work gives it the appearance of a construction whose progress has been arrested before completion. xxii. N.E. (“ Earthwork ”). 10 June 1912'. 512. Fort, East Morton.—On the left bank of the Carron Water, by the edge of a cliff, with an altitude of some 50 feet above the stream and I mile to the south-west of East Morton, is a small entrenchment. The position is strongly defended on two sides : on the north by the deep glen of a burn and on the west by the rocky wall rising sheer from the Carron. The enclosure is oblong, set with its longest axis north-east and south-w^est, and measures 114 feet by 82 feet. It is enclosed on the east and south sides by a well-defined trench some 24 feet wide and 6 feet deep, supplemented by a rampart on the south where the ground level declines ; while along the east front, in the interior, there is also a low and broad bank above the scarp of the trench. Some 54 feet back from the edge of the bank on the w’est, a roadway, 14 feet wide, is carried over the trench on the south at the natural level into the interior. A second break in the rampart, and a gangw'ay across the trench towards the north-east corner, are less clearly original. From in front of this gap a track leads down to the burn at the bottom of the glen on the north. Towards the east the ground rises sharply immediately adjacent to the fort and completely overlooks it. xxii. N.E. (unnoted). 8 May 1913. Cairn. 513. Cairn, Burn.—In a grass park some¬ what less than J mile to the eastward of the farm of Burn and above a high bank over¬ looking the Cample Water, is a grass-covered cairn slightly spoiled on one side, wEich measures in diameter 37 feet by 31 feet, and in elevation about 5 feet. It is unex¬ cavated. xxii. S.E. (“ Tumulus ”). 7 June 1912. Cross-Shaft. 514. Cross-shaft from Closeburn.—In the grounds of the Grierson Museum at Thornhill is a cross-shaft or slab, which was taken from a site at the village of Closeburn. It is illus¬ trated in fig. 127. The slab is of white sand¬ stone, and is 3 feet 6 inches high, 13 inches wide by 7 inches deep at top, and 14I by 8| inches at bottom. The right side is orna¬ mented with the scroll foliage involving birds pecking at the fruit, which is so marked a feature of the Ruthwell Cross and other local examples. xxxi. N.E. 10 June 1912. MOUSWALD. Castellated and Domestic Structure. 515. Mouswald Place.—Adjoining the man¬ sion of Mouswald Place, f of a mile north of Mouswald Village, is the ruin of a tower once the home of the Carruthers family. The building is oblong on plan, measuring 23 feet II inches by 17 feet 9 inches within walls 6 feet thick. There were three storeys below the wall-head; the ground floor is provided with gunloops and does not appear to have had a vaulted roof. The north wall is fragmentary ; the south, east, and west walls are 30 feet high and have been recently but¬ tressed where necessary by the proprietor. The building now presents no features of interest. Ivi. N.W. 4 October 1912. 180 MOUSWALD.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [mouswald. Defensive Construction. 516. Fort, Panteth Hill, Mouswald.—On the brow of the steep brae which rises above the U.F. Church Manse of Mouswald, and at the south end of the High Plantation, is a small circular fort at an elevation of some 400 feet over sea-level. In diameter it has measured some 200 feet, and it has been surrounded by a single rampart and trench, the latter some 27 feet in width. The defences have been much destroyed, and for a short stretch towards the east entirely demolished: the rampart is little more than a glacis, and the trench is apparent only at one or two places in the periphery. Ivi. S.E. (unnoted). 25 April 1913. Miscellaneous. 517. Effigy and Armorial Stone, Mouswald Church.—(i) Effigy. —An effigy lies in an angle between the church and a burial en¬ closure but used to be within the church {Stat. Acct., vii. p. 298). It is very much worn by exposure to the weather, but represents a man in plate armour of the end of the 15th or the early part of the i6th century. The name of Sir Simon Carruthers of Mouswald is associated with the figure, but there were four Simons of the name between 1411 and 1548. The figure of a lady, which used to accompany the present effigy, has totally disappeared. The effigy should be under cover (see Proc. Soc. Ant. Scot., xxix. p. 389). (2) Armorial Stone.— The burial enclosure mentioned above is that of the Griersons of Lag, but earlier of the Douglases of Mouswald. Over the entrance is a stone with the Douglas arms—heart, and three stars in chief—and the initials I D and A R, for James Douglas and Agnes Rome, flanking the chief, with the date 1655 below. This James Douglas was the grandson of James Douglas, second son of Sir James Douglas of Drumlanrig, who had acquired the Mouswald barony on the failure of male heirs to Simon Carruthers. James Douglas was the last of IMouswald. His daughter was mother of Sir Robert Grierson of Lag {Proc. Soc. Ant. Scot., 1888-9, PP- 62 ff). Ivi. S.E. 17 October 1914. 518. Tombstones, etc., Ruthwell U.F. Church, Mouswald.—Built into the vestry wall of this church are two grave-slabs (fig. 128) carved in high relief, which originally had place in the parish burying-ground at Ruthwell.^ Each, as inserted, is about 5 feet 10 inches long by I foot 7 inches at top to i foot 4I inches at bottom. That on the left (A) bears a cross standing upon a calvary of three steps and terminating at top and at the arms in an open lozenge. The quarters of the cross are filled with circles, which are tied with a striated band at the points of contact with the shaft and each other. Below the cross, on the sin¬ ister side, is a sword with conventional quillons and a loz¬ enge-shaped pom¬ mel, while on the other side is an ob¬ ject which cannot be identified. The slab on the right (B) bears a cross on a calvary of two steps ; the head and the arms expand at the end in lozenges, and the intersecting portion is en¬ closed within a square with indented comers, while four circles, as in the companion case, occupy the angles. A sword of similar type again appears on the sinister side, and on the dexter are what appear to be the sock and coulter of the old plough. The implements in both examples would indicate the occupation of the person thus memorialised, and the grouping of objects as well as the character of the crosses suggest that the stones are pre-Reformation in origin but comparatively late in that period. Below the crosses projects the half of a crudely moulded stone circled on plan and 2 feet 5 inches in diameter. The upper member is enriched with a debased water-leaf ornament, and the lower mouldings are rolls. ^ New Stat. Acct., iv. pp. 22J-S ; Trans. Dumf. and Gall., 1909-10, p. 215. Ivi. S.E. 29 April 1920. Fig. 128. — Grave-slabs at Ruthwell LLF. Church, Mouswald (No. 518). 181 MOUSWALD.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [penpont. 519. Tumulus.—Situated on Rockhall moor, about I mile to the north of Blountfield, is a low circular stony mound, 36 feet in diameter, which is overgrown with long heather but shows two or three small hollows in the interior and one or two large stones on the circumference. It is probably the site of a cairn, but there is a possibility of its being a hut-circle ; the heather, however, makes it impossible to determine. Some 30 yards due north is a small oval depression, with a slight bank around it, measuring some 16 feet in diameter over all. This has the appearance of being a small hut- circle. Ivi. N.E. 4 October 1912. Sites. 520. Fort, Burnhill.—Burnhill is a slight natural eminence rising to a height of some 15 feet above the adjacent ground, which is partially occupied by a school and cottages. Only at the north angle is there any trace of handiwork, where about halfway up the bank there appears a small portion of a terrace some 20 feet in width. i hd. N.W. 4 October 1912. The O.S. maps also indicate sites as under 521. Tower, Bucklerhole. 1 . S.W. 522. Tumulus, Elfknowe. 1 . S.W. 523. Tumulus, Rockhall Moor. 1 . S.E. 524. “ Tryal Cairn,” near Byeloch. Ivi. N.E. PENPONT. Defensive Constructions. 525. Mote, Druidhill Burn.—This mote is situated on the eastern boundary of Penpont parish, about i mile to the north of the farm of Merkland. It is formed from a rather large grassy hillock by a trench and terrace near the base at one side, cutting off the higher and major portion, which rests on the bank of the Druidhill Burn, at the opposite side. The area thus enclosed is oval in shape. From the bed of the burn on the east it rises very steeply to a height of from 70 to 80 feet, while from the south-west it rises by an easy gradient from above the scarp of the trench. The summit itself presents no level plat, but 182 from the crest overlooking the burn it falls with a slight inclination towards the north¬ west. At the south-west edge of the summit, and just at the point where the gradual slope upwards from the trench meets the steeper declivity to the stream, there appears to be a small artificially levelled area with a diameter of some 23 feet. The area cut off by the defences measures some 206 feet from north¬ east to south-west and at right angles 118 feet, with a rise of about 12 feet and 17 feet re¬ spectively from the top of the scarp to the highest points at the edge of the hillock. The trench extends along the south-west, cut with a steep scarp to a depth of about 16 feet and a width, where widest, of about 55 feet from the level of the top of scarp and counterscarp, and has been continued round to the west, where it appears to have stopped at a hollow crossing its direction to the base of the scarp. Beyond this a terrace takes its place along the north side to the edge of the burn. From this direction the hillock rises with a longer slope from its base, and the terrace is therefore at a relatively higher level than the trench. The scarp above it has a height of about 10 feet. A gap in the scarp, adjacent to the hollow on the north-west, is suggestive of an entrance, and in rear of it a slight depression is visible leading towards the summit. The situation of this mote is very secluded, at the head of a remote glen and among high rolling hills, at an elevation of 800 feet over sea-level. In its construction, from a simple hillock unaffected by art, except for its surrounding trench and terrace, it bears a remarkably close resemblance to the mote at Moatland in Glencairn parish (No. 239). xxi. N.E. II June 1912. 526. “ Fort,” Arkland, Glendinning Cleuch. —Immediately to the south of the farm-house of Arkland is a formation which is very puzzling. The site is a plateau overlooking the deep glen of the Arkland Burn, with a return on the face of the bank to westward and a sharp decline to eastward. The plateau is encircled by two great mounds, leaving an uneven hollow in the centre dipping towards the south-east. At the west side a level pitch has been laid across the end of the plateau, cut through the mound on the north and reach- PENPONT.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [PENPONT. ing to the face of the glen on the south—this is modern. The mound on the north is trace¬ able as it follows the edge of the glen for a long distance westward and is without doubt natural, while the mound on the east and south sides of the plateau is apparently natural also. On the south side, however, at the base of the mound on the exterior, is a distinctly arti¬ ficial terrace along the side of the glen, which is continued round the east to the north side, where it lies some 30 feet below the crest of the mound. The level of the interior, which is quite small, is some 15 feet below the crest of the north mound. It is not at all likely that this was a fort, though the O.S. map marks it as such. xxi. S.E. (“ Fort ”). ii June 1912. 527. Entrenchment, Eccles House.—On a level plateau, about | mile to the south-east of Eccles House, is an entrenchment, four¬ sided with rounded-off angles. A line bisect¬ ing it lies east-north-east and west-south-west and measures 127 feet in length, while the length of the west-south-west end is 132 feet and of the east-north-east end 112 feet. The ditch which encloses this area measures, where best preserved, some 31 feet in width, and has a depth of about 4 feet 6 inches below the scarp and from 5 feet to 9 feet below the crest of the counterscarp, the latter measurement being at the west angle, where a mound, some 3 feet high from the outside and preserved by the growth of a large ash tree, surmounts the counterscarp. Along the south-east side the counterscarp has been considerably ploughed down. The area of the enclosure has been levelled and cultivated, and lies at a slightly lower level than the ground to the north-west and west-south¬ west of it. There is still evidence of the existence of a parapet mound around the edge. The north corner of the ditch has been filled with cartloads of stones in recent times, brought from the outside, but there is no indication of any previously existing roadway across the ditch. xxxi. N.W. (“ Castle, site of ”). 20 June 1912. 528. Grennan Hill Fort.—On a shouldei" of Grennan Hill, overlooking the valley of the Scar Water which flows some 200 feet below, is a small fort at an altitude of about 500 feet above sea-level. The actual site is a rocky hillock lying with its main axis north-north¬ west and south-south-east, measuring some 152 feet in length by 62 feet in breadth and with an elevation of from 12 to 14 feet above the general surrounding level, except on the west, where the ground falls away at a steep gradient to the base of the hill. The summit is oval in form and has risen by an easy slope from the northward, while towards the south it terminates in a wall of rock. The defence consists of a trench some 35 feet in breadth and at most some 16 feet in depth, commencing at the brow of the steep declivity on the west and carried thence round the east side to terminate at the commence¬ ment of the wall of rock at the south end. The soil of the trench has been thrown up to form a rampart above the counterscarp, while a parapet mound crowns the scarp at the north end. The trench has evidently in places been cut through rock, and has been slightly expanded at its southern termination where at the date of visit it contained a deep pool of water. The end of the trench is covered by a return of the rampart to the hillock. A well-defined approach leads to the summit past the end of the trench, entering into a slight hollow at the lowest point of the interior of the fort. Flanking the southern side of the approach is a low bank, which abuts on the rock face and seems to pass in a curve, with a greatest depth of 30 feet, round to the west, forming a small enclosure at the base of the rock. Except adjacent to the entrance to the fort, however, this out¬ work is somewhat ill-defined. The summit of the hill does not appear to have been artificially levelled, but there are some three slightly hollowed circular areas which may be hut sites ; and facing the valley of the Scar Water, at the north end of the hillock, is a more considerable hollow, which likewise appears to have been artificially formed. XXX. N.E. (unnoted). 5 !May 1913. C.VIRNS. 529. “White Cairn,’’ Honeyhole.—On a prominence just to the west of the wood that lies up the hillside to the west of Honeyhole, PEN-poxT.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [PENPONT. Aluiiiit a -id Historical — Dumfries. Fig. 130.—Comlongon Castle (No. 537). Fig. 132.—Comlongon Castle, S.W. corner of Hall (No. 537). I'o face p, 18 PENPONT.] INVENTORY OF'MONUMENTS IN DUMFRIESSHIRE. [ruthwell. on the road from Druidhall Bridge to Penpont, are the remains of a large circular cairn which has had a diameter of from 8o to 90 feet. The stones have been almost entirely cleared away with the exception of the debris, and it is very doubtful if the interment remains undisturbed. xxi. S.E. II June 1912. 530. “Tumulus,” Auchenbainzie Hill.— This has been a small cairn with a diameter of about 24 feet, which has been reduced to ground level and probably had the interment disturbed. xxi. S.E. II June 1912. Miscellaneous. 531. Cross-Shaft, Nith Bridge.—In a field and surrounded by an iron railing, some 200 yards west of the Nith Bridge and 20 yards south of the road from Penpont to Thornhill, stands a cross of led sandstone (fig. 129) 9 feet in height above ground, 17 inches broad at base, 13 inches at top of shaft, 8 inches in thick¬ ness. The length of the shaft is 6 feet 6 inches. The shaft on all sides is richly carved with zoomorphic and foliageous interlaced work. The cross is fully described in Early Christian Monuments, pp. 449-450, from which the drawings in fig. 129 have been reproduced. It is much covered with lichen, and in consequence the ornament is difficult to follow. xxxi. N.W. 10 June 1912. 532. Entrenchment, Grennan Hill.—Along the west flank of the isolated summit, which rises about J mile to the east of Grennan Hill, is a slight trench with a well-defined scarp on the upper side and a slight mound on the counterscarp, occurring about halfway up the slope of the hill. The trench is lost as it passes round the south end and cannot be traced over the summit to the north. The position of an entrance through this construc¬ tion is evident directly south-south-cast of Grennan Hill. There are a number of low banks running about the hill, which seem to be the remains of old feal dykes, and though this entrenchment is more prominent, it can hardly be regarded as a permanent defensive work. x.xxi. N.W. (unnoted). 5 May 1913. Sites. 533. Fort, Virginhall Plantation.—About ^ mile to the west of the Nith Bridge, on the left Mde of the road from Penpont to Thornhill, lies the Virginhall Plantation, within which is the site of a fort. A plateau with a steep bank some 20 feet in height, facing the main road and curving round to the west, stretches away to the northward, having a steep scarp likewise on the east side. Cutting off the point formed by the meeting of these banks, a broad but somewhat shallow trench has been dug across the plateau, the bottom of which lies some 3 feet below the scarp and i| feet below the long slightly defined counterscarp. The indication of a rampart above the scarp of the trench is very faint, and there are no suggestions of a mound along the sid(?s. The remains of the fort are altogether obscure. xxxi. N.W. 10 June 1912. 534. Cairn.—A short distance to the north¬ east of No. 530 is the site of another similar cairn which has been entirely removed. xxi. S.E. II June 1912. The O.S. maps also indicate sites as under :— 535 - Auchengassel Castle, x.xi. N.E. 536. Tower at Arkland, xxi. S.E. RUTHWELL. CASTELL.4TED AND DOMESTIC STRUCTURE. 537. Comlongon Castle.—This is an un¬ usually complete and well-preserved example of a 15th-century tower, which stands, adjoining a modern house, on the flats bordering the Solway Firth, 8 miles south-east of Dumfries. It is oblong on plan, measuring exteriorly 42 feet 7 inches from north to south by 48 feet 10 inches from east to west, with walls ii to 132 feet in thickness at base, and contains a vaulted basement of two storeys, above which are three storeys beneath the wall head and the parapet walk. The total height from ground to the top of the cap-house is 641 feet. The masonry is of rubble with freestone dressings ; a weathered basement course returns round the structure at base ; the jambs and lintels of openings have cham¬ fered arrises. The tower is entered through a semicircular 185 RUTHWELL.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [ruthwell. headed doorway, some 3 feet above ground level towards the east end of the north wall, which still retains its iron gate (described in Proc. Soc. Ant. Scot., vol. xvii. p. no), in front of which there was a wooden door opening outward (fig. 133). Behind the ii'on gate a recess or seat is formed in the west ingoing ; in the right ingoing is the stairfoot of the main wheel-stair, contained within the walls of the north-east angle, which serves all the floors and terminates in a cap-house at the parapet walk. Beyond the stairfoot and to PLAN OF PARAPET WALK PLAN OF BASEMENT PLAN OF FIRST FLOOR Fig. 131.—Comlongon Castle (No. 537). the south of it is the entrance to the lower basement floor, which is a gloomy chamber dimly lit by a narrow aperture in the east and west walls, widely splayed internally and with a stepped sole. There is a curious alcove in the south-east angle for some indeterminable purpose ; it is circled on plan. A small secondary stair at the west end of the south wall ascends only to the first floor level. There is a circular well. The north and south walls have a scarcement to receive the timbers of the upper floor beneath the semicircular barrel-vault. This chamber is lit by apertures similar to those below ; it was entered from a doorway in the east wall, now built up, reached from the main staircase. The first floor or hall is a fine apartment 29 feet 4 inches by 21 feet 2 inches, with a stone-flagged floor and a ceiling of beams sup¬ porting the floor joists of the room above tind resting on corbels carved with heraldic or other devices, only a few of which can now be made out. On the north wall the westermost corbel shows a saltire and cushions in chief for Johnstone. The corresponding corbel on the opposite wall has a saltire only, the next three arrows paly with points uppermost and no feathering ; on another an angel holds a shield, now blank. An heraldic panel is inserted over the fireplace in the west wall and bears a shield filled with what is obviously intended for the royal arms, though the lion is regardant and the unicorn supporters are sejant. On each side is a moulded corbel with an angel bearing a shield carved on the face. The shield on the right shows the three stars within a double tressure flory-counter-flory for Murray : that to the left parted per pale has on the sinister side the Murray arms, and on the dexter a couped panel of chequers in five rows with a mascle in base. The east wall contains a large kitchen fire¬ place which was shut off from the body of the hall by a partition with a serving hatch to form a separate chamber for cooking. In the south-east angle is an L-shaped mural chamber with two small windows. The west wall has a fine central fireplace with shafted jambs, moulded bases and capitals, the last enriched with two rows of foliaceous orna¬ ment. Of the lintel stone only a portion over each jamb remains ; a wooden lintel of considerable age now spans the jambs, and the remainder of the lintel space is filled with rubble; above is the original cornice, on which is carved a vine scroll. The cornice has been re-built but with the original stones, as a human head, matching one over the south jamb, is not in situ. On either side of the fire¬ place there is a doorway with a small window over ; the north doorway leads to a small square mural chamber in the north-west angle, that on the south is the entrance to the wheel- stair reaching to the lower basement. In the south wall at the west end there is an ornate stone buffet, which has been shelved. The jambs, resting on a projecting stone sill, are shafted and have moulded bases and capitals surmounted by crocheted finials ; the head is a semicircular arch crocheted on the extrados and cinquefoliated with foliaceous cusps. 186 Ancient and Historical Monuments — Dmnjries. Fig. 133. — Iron “ Vett’’ or Gale, Comloiifjon Castle (No. 537). To face p. 186. RUTHWELL.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [ruthwell. There are two windows in the south wall, one with an aumbry in the ingoing ; and there is one in the north wall. All have pointed segmental scoinson arches and window seats in their deep embrasures ; the breasts project inwardly as a shelf or table, and the lintels and upper part of the jambs are grooved for glazing, while behind the reveals there has been a shutter. At the east end of the north wall an angled stair leads down to a prison within the thickness of the wall, comprising two inter¬ communicating vardted chambers each 6| feet wide and 12 feet in length, lit by a small window fitted with seats in the outer wall. The inner chamber has a garderobe with soil flue to the south, and the outer has a hatch in the floor (fig. 134), the only access to a lower prison (“pit ”) which is ii feet deep and 171 feet long by 4 feet wide. This prison is stated in Proc. Soc. Anh Scot., vol. xvii. p. no, to have a ventilation flue of unusual type. Above the prison there is a mural chamber with a window, garderobe, and soil flue entered from the stair on the ascent to the second floor. There have been two ■cSiUpril”®” chambers on the second (No. 537). floor, each with a fireplace. The eastern chamber has a small window in the east wall north of the fireplace, an L-shaped mural chamber in the south-east angle, a window with seats in the south wall, and beside it to the west a garde¬ robe, adjoining which, but within the next chamber, is another, as at Drumcoltran Castle.^ The west chamber has a window with seats west of the garderobe and a small window south of the fireplace. The fireplace has an angled recess in its north jamb. The north wall contains one window, and from its ingoing one mural chamber is entered to the east and another to the west. The western has two windows, a garderobe with lamp recess, and a squint overlooking the western apartment. The eastern chamber has one window, and there are two recesses in the south wall. The arrangement of the third floor is very similar, but the mural chamber in this floor is placed in the south-west angle, and there are no garderobes in the south wall. Within the roof there was a garret entered from the east parapet walk. The walk is corbelled out on moulded discontinuous corbels and returns round each wall within a crenellated parapet 56I feet above ground, of which the only por¬ tion not reconstructed seems to be that on the east. Towards the late i6th century the western walk was enclosed and roofed in to form a gallery. It has an entrance at either end from the parapet walk, and is lit by windows in the crow-stepped gables and the west wall. The gallery contains a fireplace, slop drain, and lamp recess. On its floor is seen the gutter¬ ing of the original walk. Similarly, the south¬ east angle within the parapet is now occupied by a little structure contemporary with the gallery. The inner face of the east wall is carried up to receive the roof and is crow- stepped with unusually broad steps in order to give access to two watch platforms within crenellated parapets, the one situated over the cap-house, the other, slightly higher, on the east chimney stalk at the roof ridge. The building is very typical of its period and, being entire, shows very clearly the arrangement and accommodation of the 15th- century tower. It is well roofed (modern) and the walls on the whole are in excellent pre¬ servation, but attention may be drawn to a rent in the masonry at the west end of the mural chamber immediately above the upper prison. The wooden flooring and joists have disappeared, but many of the bridging joists and bressumers are in situ. This tower was a residence of the Murrays of Cockpool, etc. (cf. Inirod., p. xxix.). ^ Inventory of Monuments in the Stewartry of Kirkcudbright, No. 276. Ixi. N.E. 24 July 1919. Miscellaneous. 538. Ruthwell Cross.—See Appendix. 539. Heraldic Panel, Ruthwell Church.—At the end of the south aisle in Ruthwell Church, beneath a window to the memory of the Earl of Mansfield, is a coat-of-arms on a bordered panel, blazoned thus :—Quarterly, ist and RUTHWELL.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [sx mungo. 4th, three mullets uxthin a bordure flory (Murray) ; 2nd and 3rd, three crosses patee (Barclay) ; supporters, on each side a hound: over all a viscount’s coronet. Motto— spero MEL iOR.\. Crest—a stag’s head with a cross between the horns. At base the date 1687. In the ke^^stone of the arch of a recess be¬ neath is a shield charged with three mullets within a double bordure flory. 540. Tombstones.—To the east of the south end of the church is a richly-carved slab, in¬ scribed :— MY DEAR MOTHER NEIR I LY TEN BEFORE ME THE IITH AM J. DEAR SPOUS WHILE YOU BEHOLD THIS SHRYNE THINK ON YOUR BONY BABS AND MYNE. Ixi. N.E. 24 July 1912. Sites. The O.S. maps indicate sites as under :— 541. Tumulus, Bent Mote. Ivi. S.E. 542. Cockpool Castle. Ixi. N.E. Hic JACET Gilbertus Conder. He WAS FACTOR TO THE ViSCOUNT OF STOR¬ MONT 41 YEARS. He degased (sic) the II OF March 1709 of age 75 years VIVIT POST FUNERA VIRTUS. Immediately to the south of the south end of the church lie a number of stones to the memory of members of the Young family. The most southerly of the group measures 9 feet 8 inches by 3 feet 3 inches. At the head is carved a cherub’s head, with outspread wings. Beneath, heir lyes margrat young SPOUS TO EDWARD CARTEL OF LIMKILLS WHO DECEASIT MAY 24 1665 OF HIRE AGE 48. OF VERTUE, WIT, GRACE TRUTH LOVE PIETIE THIS tVOMAN IN HIR TYME HAD STORE ON SMAL MEANES SHE UPHELD GRIT HONESTIE & IN REWARD HES ENDLES GLORE. In the centre of the stone is a shield charged with three piles, for Young. At the base, skull, cross-bones, and memento niori. The adjacent stone, of similar size, is in¬ scribed ;— HEIR LYES THE EARTHY PAIRT OF CHRISTIAN YOUNG SPOUS TO JOHN WILSON WHO DE- PAIRTD I OF MARCH 1653 OF HIR AGE 27. In the centre is carved a skull, with memento mori motto, a sand-glass, and spade. Two shields charged ;— Dexter, a lion rampant with a dagger, for Wilson : flanking the shield the letters 1 . W. Sinister, a chief pily, for Young : outside the letters C. Y. At the base of the stone is this inscription :— ST MUNGO. Ecclesiastical Structure. 543, Church of St Mungo.—The ruins of this church stand on the left bank of the river Annan, ij miles north-north-east of Dalton village. Only a portion of the chancel, con¬ sisting of the east gable, 8 feet 6 inches of the north wall and 36 feet of the south wall, now exists. A door between two windows in the south w'all wLich is dated 1754 on the lintel, and a window in the east gable, are probably contemporaneous. The east gable, enshrouded in ivy, is 24 feet wide, 3 feet thick and some 18 feet in height ; a string-course of two unequal splays, separated by a vertical plane, runs along the outside about 3 feet 6 inches from the ground. It is doubtful whether any of the work now remaining is earlier than the i8th century. This parish was anciently called Abermelc. The Inquisitio of Earl David in 1116 found that the lands of old belonged to the bishopric of Glasgow, and the church continued to be a mensal church of the bishopric till the Reformation. In the notice of the parish in the New Statistical Account, dated 1834, the minister says : “ A considerable part of the church seems to have been rebuilt from the remains of the original structure, as the present walls contain the remains of broken pillars and door and window rybats. The church was in the form of a passion cross, having very narrow lancet-shaped windows. In 1754, the south wall (not the whole church, as stated in the former report), was rebuilt, as appears from an inscription over the old chancel door; 188 .-itil it'llf and Historical Monuments — Dumfries, To face p. 189. Fig. 137.— Sanquhar Castle : Entrance to Inner Courtyard (No. 551). Fig. 138. —Sanquhar Castle ; The Tower (No. 551). ST MUNGO.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [s.\nquhar. and, in 1805, to give room for celebrating the holy sacrament of the Lord’s supper, the north transept was removed and a new one built, and the whole structure repaired and modernised. . . . The ancient baptismal font is still preserved ; and, on a late repair of the church, two grave-stones of great antiquity were discovered ; these are preserved care¬ fully. The church was covered with heather till I 754 -” Ivii. N.W. 9 June igi2. Defensive Construction. 544. Fort (remains), Nutholm Hill.—At the north end of Nutholm Hill, an eminence which lies about J mile to the north-east of Nutholm- hill farm, is a fragment of a rampart some 60 feet in length with a trench to the outside, the bottom of which lies some 3 feet below the crest of the scarp. These are the remains of a fort which at one time evidently crowned the summit. The enceinte has probably been curvilinear. The summit does not appear to have been levelled or excavated. It com¬ mands a fine prospect of both Dryfesdale and Annandale to the northward. li. S.W. 30 July 1912. Enclosure. 545. Enclosure, Nutholm. — On the lower slope of Nutholmhill, some 150 yards to the east of Nutholmhill farm, are the very faint remains of an oval enclosure, partially destroyed by the road and greatly demolished by the plough. A broad inner bank which has surrounded it is discernible towards the west, where also a slight hollow indicates a trench. li. .S.W. (“ Fort ”). 30 July 1912. Sites. The O.S. maps indicate sites as under :— 546. Castlemilk. li. S.W. 547 - Malls Castle, Norwold farm. li. N.W. SANQUHAR. Ecclesiastical Structure. prebend in Glasgow cathedral with consent of the patron, Crichton of Sanquhar, whose right of patronage continued. The remains of the old church were de¬ molished in 1827 and a new building erected ; but the foundations of the old church were laid bare in 1895 and covered with large flag¬ stones, so as to mark out their exact size and position in the churchyard to the east of the present church. These foundations show that the old church was some 96 feet long by 30^ feet broad. The choir had three large but¬ tresses projecting from the walls at ground level. A splayed base was found along the foundation of the choir wall, but not along that of the nave. This suggests that the choir was added in the 15th century on a more elaborate scale. vi. S.W. 20 October 1915. Sepulchral Monuments. 549. Effigy of a Priest.—Within the present parish church, near the south-west corner, is preserved a fine though somewhat worn and mutilated effigy of a priest in the eucharistic vestments (fig. 135). After being removed to Friar’s Carse as a curiosity, it was brought back in 1897 by the Marquess of Bute, 14th Lord Crichton of Sanquhar, and set up on a plain modern altar tomb. The figure is of fine sandstone, about 6 feet long, and is represented in apparelled amice, albe vith sleeve apparels, maniple, stole, and chasuble. The hands, now much mutilated, are joined upon the breast, and the hair is straight, showing a tonsure. The amice apparel is very badly mutilated, but it has been 4 inches wide. The albe is very full ; no apparel is shown at the skirt, but there arc small ones at the wrists of the sleeves. The chasuble is large and ample, and falls in very graceful folds ; more of the forearms is shown than is usual in English effigies. It has a Y-cross orphrey. The whole figure is so badly worn that it is only on the lower parts, nearest the slab on which it rests, that one can see that the orphreys and amice apparel have been richly ornamented. The effigy probably dates from the first half of the 15th century. 548. Old Church, Foundations.—In the 15th 550. Cross-slab.—A small slab, carved with century the rectory of Sanquhar was made a a floriated cross, is built into the south-east 189 SAXQUHAR.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [sanquhar. wall of the churchyard. It is i foot 2 inches at the south-eastern end of the burgh of long and 9 inches broad. Sanquhar. The site is enclosed by curtain vi. S.W. 20 October 1915. walls, which form an area trapezoidal in shape, Fig. 136.—Sanquhar Castle (No. 551). Castellatf.d and Domestic Structures. 551. Sanquhar Castle. —Sanquhar Castle is situated on a high bank sloping steeply to west and north and overlooking the river Nith measuring 163 feet from north-west to south¬ east by 130 feet from south-west to north-east. These curtains, which date from the late 15th century, replace earlier walls, which can be traced on west and south, where they abut on a SANQUHAR.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [sanquhar. square tower at the southern angle of the en¬ closure, evidently the earliest portion of the castle and probably dating from the early 15th century. The eastern flank of the site is protected by a dry ditch 35 feet wide across the scarps, which returns from the declivity on the north along the east curtain and was crossed by a bridge now represented only by its abutments. The buildings are extremely ruinous, and their chronology and arrangement are further obscured by a modern partial restoration. The entrance is immediately west of the bridge through a 17th-century gateway in the east curtain, which opens on the outer court¬ yard. This is an enclosure measuring 85 feet from north-west to south-east by 130 feet from south-west to north-east, and is bounded o*n the south-east by a late 15th-century range of building, which is pierced by a central vaulted pend (fig. 137) communicating between the outer and an inner courtyard on the south¬ east. South-west of the pend a circled tower projects into the outer courtyard, and con¬ tains the well chamber in its lowest storey. North-east of the pend and entering from it is a vaulted chamber with a closet in its south¬ west wall. Between this chamber and the east curtain are sundry returns and foundations, while the foundations of a projecting quadrant¬ shaped tower appear south-east of the cham¬ ber. From these structures a ruinous range of kitchen offices returns to the south-east along and embodying the west curtain, and terminates against the square tower at the south angle. This tower (fig. 138) is four storeys in height and measures some 23 feet square over walls which are 5f feet thick on the basement floor. A splayed basement course returns round the building and is also carried along the outer and inner faces of the original curtain. The tower was entered apparently at basement and first-floor levels from the south curtain through doors in the north-west wall. The third floor had doorways opening on the parapet walk of the south and west curtains. The basement contains one unusually small vaulted chamber measuring ii| by 101 feet, which has no in¬ ternal communication with the first floor. As the greater part of the tower is a modern reconstruction, little can be said of the features now seen, but the window on the second floor in the south-west wall appears to be mainly original. It is a two-light window, mullioned and transomed ; the heads are ogival, and the soffit is strengthened by seg¬ mental arch ribs. The three-light window of the south-east wall on the third floor may also be original work, but a close examination is impossible in the present state of the building. It has unusually broad mullions and transoms, splayed and banded. The garderobe entrance at the same level in the north-w'est w'all is original. The head is obtusely pointed and is formed in two stones, wEile a bold splay re¬ turns round the intrados and down the jambs. The tow^er is built of browmish rubble with dressed margins ; the later building betw'een the two courtyards is of light coloured ashlar in unusually large stones. Red tiles, where they appear, mark the line between the old building and recent restoration. In connection with the castle there w^as an extensive deer park to the north-west, wEere portions of the boundary w'all are yet standing. Cf. historical references in Introd., pp. xxvi, lx. vi. S.E. ig August 1919. 552. Elliock House.—Standing on the left bank of Elliock Burn, 3 miles south-east of Sanquhar, is a 17th-century manison, H-shaped on plan, wEich incorporates older buildings in the northern of the two parallel wings. A cir¬ cular tow'er, containing the main entrance to the house and a spacious wheel-staircase, projects westwards from the central wing, and is dated MDCLViii. The older portion of the building extends from the angle of the northern wing for a distance of 18 feet 6 inches, and is the full breadth of the wing, 21 feet 3 inches. It contains a vaulted apartment, 14 feet broad, within walls 3 feet 7 inches thick, with shot- holes to the north and south. There is a charter, c. 1327, of the lands of “ Elyoc ” (Eliott, Elietis) to Richard Edgar (Reg. Mag. Sig., p. 530), but in 1388 Sir William de Dalzell is styled lord of Elyok (Reg. Hon. de Morton, Bann. Club, ii. p. 163). In 1463 Robert Charteris of Amisfield and Eliock granted the lands of Eliock to Sir Robert Crichton of Sanquhar, and in 1593 Robert Crichton of Cluny sold Eliock with the castle, etc., to James Stewart of “ Ballaquhane.” But in 1574 Robert Dalziel of Dalziel gave to 191 SANQUHAR.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [sanquhar. his son Robert the lands of Eliock to the extent of 26 inerks, and in 1596 Sir Robert Crichton of Cluny, son and heir to Robert Crichton of Eliock, King’s Advocate, with consent of Sir James Stewart of Ballaquhane, sold to the same Robert Dalziel junior a further portion of the Eliock lands, vith the fortalice, etc. {Reg. Mag. Sig., s.d.). xiii. N.E. II June 1912. Defensive Constructions. 553. Fort, Wanlock 'Water.—Far up the glen of the Crawick Water, some 8 miles above San¬ quhar and on the south side of the Wanlock Water, just before it debouches on the glen, is a small circular entrenchment or mote. It is situated at a ford where the north slope of Clackleith Hill terminates in a steep grassy bank above the bed of the Wanlock Water, which flows by about 100 yards distant. Towards the north-east is the steep bank before mentioned; on the north-west is a slight hollow, which in wet weather is probably the bed of the rivulet ; while on the south-west the gentle downward slope of the hill takes a steeper gradient as it passes the south-east flank of the construction. The central plateau, which has been very slightly, if at all, raised by art, rises some 5 feet above the bottom of the trench which surrounds it, is flat and measures some 40 feet in diameter. The trench from crest to crest has a breadth of from 20 to 23 feet, except towards the steep bank on the north-east, where its width and depth considerably diminish. Toward the east a traverse projects across the trench, which ma}^ have been the line of entrance to the plateau. The soil of the trench has been thrown up to form a mound on the counter¬ scarp, which along the east arc is of uniform level with the central plat. Somewhat to the west of the centre of the interior is a circular hollow, some 8 feet in diameter and 2 feet in depth—an excavation which has not been made in recent times but is probably not original. ii. S.E. {“ Fort ”). 18 June 1912. 554. Fort, Old Barr, near Euchan Cottage.— The remains of a fort, represented by a broad defensive wall for the most part overthrown and grass grown, are to be seen on the north side and north-east end of a long round- backed ridge, which rises steeply from the right bank of the Euchan Water about mile south-south-west of Old Barr and 2 miles south-west of Sanquhar. The wall is not now continuous in its course ; and the plan of the fort, which appears to have been curvilinear, is not ascertainable. The area has been under cultivation, and the ruins of farm buildings lie at its western extremity. xiii. N.W. 19 June 1912. 555 - Fort, South Mains, Elliock.—On the haunch of the hill overlooking South Mains farm, and about J mile south by west of it, is a small entrenched earthwork. In outline it forms an oval, lying with its main axis almost north and south and measuring over all some 190 feet in length by 126 feet in breadth. It is divided into two distinct divisions at about one-third of its length from the north end. The southern division is a truncated oval or horse-shoe shape, and is surrounded by a trench some 34 feet in breadth and 6 feet deep, except where carried across to form the division, where it has a breadth of 20 feet. The area thus enclosed measures 89 feet in length by 58 feet in breadth. Around the periphery, except to¬ wards the north opposite to the other en¬ closure, there is a parapet rampart. The interior, except for a space of some 30 feet at the south end, has been hollowed by excavation. The north enclosure is faced on its south aspect, against the dividing trench, by a rampart and has also a rampart enclosing it to the outside. There is no indica¬ tion of a trench around it, and the interior area is sunk below the ground level on the west or upper side. This enclosure forms a semicircle, measuring interiorly 72 feet by 44 feet. There is a very wide entrance into the main enclosure at the south-east angle, at the end of the east side, on unexcavated ground over the trench. On the thick turf in front of it there appear some lines, suggesting foundations of a covering wall or bailey. There is now no visible entrance into the northern enclosure. xiii. N.E. 9 May 1913. 556. Mote, Ryehill.—Some 170 yards to the south-west of Ryehill farm, and on the edge of 192 SANQUHAR.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [sanquhar. a steep bank which falls to the haughs on the north of the Nith, is a knoll bearing the name of the Mote. It has the appearance of having been regularly scarped all round, and presents a fairly level summit; but on the side away from the river it has an eleva¬ tion of some 8 feet, and, except where the ground dips on the north-west side, it does not attain to a greater height than lo feet. There is no indication of a trench having existed around it. The plateau has been under cultivation, and, with its longest axis north- north-west and south-south-east, measures some 62 feet by 47 feet. From the east¬ ward a hollow approach leads upward on to the summit. The situation suggests a genuine mote, but the mound is singularly low in height, and there is no trench to com¬ pensate. xiii. N.E. 9 May 1913. 557. Fort, Kemp’s Castle, Euchan Water, Sanquhar.— This fort (fig. 139) is situated about I mile south-south-west of the town of San- (luhar, on a long narrow plateau occupying an acute angle between the confluence of the Barr Burn and the Euchan Water, and lies with its main axis east and west. From end to end it measures some 350 feet, in breadth Fig. 139.—Fort, "Kemp’s Castle” (No. 557). at the west end 55 feet, and at the east end 70 feet. It is highest at the west end, where it rises some 40 feet above the Barr Burn, flowing by its base on the north, and some¬ what less above the meadow intervening between it and the Euchan Water on the south; and dips to the east some 23 feet in the course of its length. Approached from the west there is first encountered a trench across the ridge in front ol the plateau, some 32 feet wide, which stops some 6 feet short of the bank on the south flank, so as to leave room for a road¬ way along the edge. In rear of the trench is a small harp-shaped court, measuring some 66 feet across the interior face, 66 feet along the south, and diminishing to 37 feet on the north. Separating this court from the front of the plateau is a trench 6 feet deep below the counterscarp and 10 feet below the scarp, 42 feet in breadth from crest to crest and boldly cut. The track from the outer trench has been continued along the edge of the south bank, and has entered the interior up the scarp at that side. The plateau is in three divisions :—(i) An area some 50 feet in diameter at the west, at the front and sides of which are the re¬ mains of a stony parapet mound, and between which and the edge of the scarp of the trench is a berm some 3 feet wide ; (2) a constricted portion, some 50 feet in length, connecting the higher area with (3) the larger and wider part of the plateau towards the east. At the eastern extremity the steep flanks con¬ verge somewhat, and a track has led down the ridge between them from the interior on the line of a modern path. Half-way down this track to the burn level flat flags and outcropping boulders seem to indicate the foundation of a cross-wall ; and above the lowest level across the peninsula, here some 60 feet wide, is a distinct scarp some 4 feet in height. At its extreme eastern termination in the angle between the streams are a number of mounds with a distinctly artificial appear¬ ance, and in the centre an oval plat, which measures some 33 feet by 25 feet. The true character of these latter mounds and hollows is not, however, apparent. vi. S.W. (“ Vitrified Fort ”). 19 June 1912. Cairns. 558. Small Cairns, etc., Lochside.—On the brae face above the right bank of the Loch Burn, and about | mile north-east of Lochside, are a number of small artificial-looking hollows, one measuring 9 feet by 7 feet in diameter and 11 feet in depth, and one or two small circular hut foundations too indefinite for measurement. One mound was observed, oval in form and measuring 17 feet by 12 feet, with an interior hollow, entering from north¬ west, at right angles to its longest axis 13 193 SAXQUHAR.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [sanquhar. and measuring some 5 feet by 6 feet. At the back of the mound is a depression, from which the soil that forms it has been thrown out. (A construction somewhat analogous to this was seen near Doir’ a’ Chatha, Altassmore, Sutherland. See Sutherland Inventory, No. 58.) Several other oblong mounds, about i foot in height, show the same hollow behind. These objects occur between 700 feet and 800 feet over sea-level. xi. S.E. 19 June 1912. 559. Small Cairns, Elliock.—On the moor about I mile due south of Elliock House, on both sides of the Deil’s D3'ke and close to it, is a group of some fifteen to twenty small cairns. No hut-circles are visible among them. The altitude at which they occur is between 800 feet and 900 feet over sea-level. xiii. N.E. 8 i\Iay 1913. 560. Cairn (remains of). — On Glengenny Moor, at an elevation of some 880 feet, are the remains of a large cairn. From east to west it has measured some 86 feet and from north to south 94 feet : its elevation is now about 3 feet. There is no evidence that the interment has been disturbed. xiii. N.E. 8 Ma^^ 1913- 561. Cairn (Fort, supposed). Bogs Burn.—At the south base of Conrig Hill, on a shelf of the hillside and just to the north of the meeting of two rivulets that go to form the Bogs Burn, is a large sheepfold, enclosing in its centre the remains of an earlier construc¬ tion. The interior is covered with coarse grass and rushes, and it is difficult to deter¬ mine the true nature of the remains. A broad encircling ring of stones is very apparent, and, though the centre is hollowed to some slight extent, the presence of large stones all over it demonstrate that this is the site of a cairn from which the fold has been built. The diameters are 88 feet and 80 feet. vi. N.E. 19 June 1912. 562. Knockenhair.—The O.S. map marks a “ stone circle ” on the summit of Knocken¬ hair, a hill rising to a height of 1325 feet, on the east side of the Crawick Water. The “ stone ciixle ” is a ring of small stones and earth, the remnant of a demolished cairn. vi. N.E. (“ Stone Circle ”). 18 June 1912. Miscellaneous. 563. Details, Sanquhar. —Town House.— The town house, dating from 1735, is a Renais¬ sance edifice of two storeys, designed by William Adam (?). The north chamber on the upper floor contains a 16th-century fireplace of freestone 4 feet 3 inches broad by 4 feet 7 inches high. The head is ogival, and the jambs, 7 inches broad, rise from a splay at base and have mouldings of a late type which also return round the head. Burgh Cross, 1680.—This is represented only by its upper termination, a simply moulded stone preserved on a flanking wall of the Free Church. Old House.—A tenement on the north side of the High Street opposite the town house bears the date 1626 on one skew-put and has a heavily moulded Renaissance entrance. vi. S.W. 19 August 1919. 564. Cross, Mennock Pass.—Some 3J miles up the valley of the Mennock Water, where the pass which leads through the hills to Wanlock- head is at its wildest, is a cross formed of soil and gravel on a plateau to the north of the roadway. From the bottom of the glen, narrow at this point, the hills rise with steep ice-planed flanks for several hundred feet. In the angle formed by the confluence of the stream which flows down from the White Dod and the Mennock Water is a plateau which presents a steep scarp towards both streams. Here to the east of a sheep- fold lies the cross. It is raised from i foot to 11 feet above the surrounding level and is a Latin cross, measuring 52 feet in extreme length, 34 feet from the foot to the crossing, and 47 feet across the arms, with a general breadth of about ii feet. Where the surface on the cross is broken, it shows that it is formed of the soil and fine gravel which covers the rest of the plateau. The head of the cross points 230 degrees mag. vii. S.W. 19 June 1912. 565. Inscribed Stone, Blackaddie. — Built into the left wall of the passage leading from 194 SANQUHAR.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [tinwald. the square of the steading at Blackaddie farm is a stone inscribed in Gothic lettering. The stone was removed from the kitchen of the farmhouse about sixty years ago. The house was in part the old manse of Sanquhar, built in 1755, but the stone appears to have come from a much older building, probably its pre¬ decessor. The lettering is worn, but still shows the name of William Crichton, rector of Sanquhar, son of the late William Crichton of Ardoch (Durisdeer). The rector is witness to a deed dated 1513, and was dead before 1536, when the two sons were legitimated {Reg. Mag. Sig., s.d., Nos. 42, 1595). Ardoch was sold to the Douglases in 1507 (Brown’s Sanquhar, p. 386). vi. S.W. 9 May 1913. 566. Deil’s Dyke. — Running across the Elliock estate, to the south of Sanquhar, is a low bank or dyke bearing the name of Celtic or Deil’s Dyke on the maps. Its general direc¬ tion is east to west from the Durisdeer parish boundary above Elliock wood, and J mile east of the Euchan Burn, where it appears to have turned sharply to the north for about if miles ; thereafter, again assuming an east to west coui'se, it is lost by the edge of culti¬ vated land near the Polwarlock Burn, about \ mile east of the Ayrshire border. Its course is very irregular ; it does not appear to select defensive ground, and it follows the southern slope of the watershed at a distance of a mile or thereby from the crest. On the whole it is an inconspicuous turf bank, but between the Twenty Shilling and Elliock burns a foundation of somewhat large stones is exposed, indicating a breadth of 7 feet. vi. S.W. 8 May 1913. Sites. 567. “ Sean Caer ” (he.Old Fort), Broomfield. —About mile to the north-west of the church, at the west end of the town of Sanquhar, is a ridge trending from the north-west and end¬ ing in a swelling hillock, with a height on all sides, except along the ridge, of some 20 feet. This hillock is the site of a fort, and across the ridge there are evident the remains of a broad and deep trench, whose scarps have been greatly levelled down by the action of the plough. vi. S.W. 19 June 1912. 568. Crannog.—About | mile north of Loch- side is a small pool (O.S. Black Loch), around which innumerable sea birds were soaring. The site of a crannog is marked in it on the O.S. 6-inch map as “ Lake Dwelling.” For description of the remains found on drain¬ ing the loch, see Trans. Dumf. and Gall. Antiq. Soc., 1864-5, p. 5. Cf. also Introd., p. Ivii. vi. S.E. 19 June 1912. The O.S. maps also indicate sites as under :— 569. Cairn, The Manse, Sanquhar, vi. S.W. 570. Glenrae Castle, Glenrae Burn. iii. N.W. 571* Castle Gilmour, Auchengruith. vi. S.E. 572. Senchar Hospital, vi. S.E. 573. Cairn, Sanquhar Moor. vi. S.E 574. Cross (supposed Site of Ancient Church, Auchengruith Craig), vii. S.W. 575. Castle, near Goose Hill. xiii. N.E. 576. Castle, Ryehill. xiii. N.E. TINWALD. Ecclesiastical Structure. 577. Trailflat Church.—The parish of Trail- flat was suppressed and united with Tinwald in 1650. The church was given by Walter de Carnoc, lord of the manor, to the Abbey of Kelso in the 12th century. The grant was confirmed by Walter, Bishop of Glasgow, by William the Lion, and by Sir Thomas Carnoc in 1266. After the Reformation the rectory passed to the Earls of Roxburgh, from whom it was bought by the king and given to the Bishopidc of Galloway in 1637. The remains of the church are situated i.^ miles south-east of Shieldhill station, between Lockerbie and Dumfries ; the plan is a plain rectangle, and the internal dimensions arc 42 feet 3 inches by 16 feet, between walls 2 feet 4 inches thick. The highest part of the walls now standing is the south-west angle, where the stonework is still some 8 feet in height. The west gable contains a lintelled window, with a splayed jamb. xiii. S.W. 16 August 1912. 195 TixwALD.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [tinwald. Castellated axd Domestic Structure. 578. Amisfield Tower. — Five miles north of Dumfries this lofty tower, almost wholly detached from the later buildings, reveals itself as the culmination of design in the i6th- centuiA' domestic architecture of Scotland. The tower is undoubtedly from the hands of the same designer as Elshieshields (c/. No. 447), who, if he may be judged from these two pro¬ ducts of his skill, was an artist of imaginative and constructive ability, not in the foreign or “ grand manner ” which was already be- ir-- A FIRST FLOOR GROUND PLAN I THIRD FLOOR SECOND FLOOR O CV4 turret-stair is corbelled out at first-floor level and rises vertically for a height of 25 feet, where it develops into a square projecting tower of ashlar, roofed and gabled, which abuts on a cap-house carried still higher on the east gable. The cap-house has a considerable in¬ ward projection, which is apparently largely supported by the wooden trusses of the main roof. This is an ingenious and most unusual construction detail. The west gable Fig. 140.—Amisfield Tower (No. 578). ginning to come into use, but in the intimate indigenous style evolved from the adoption and continuance of the rectangular tower plan. While Elshieshields has the same air of dis¬ tinction and shows a similar careful proportion¬ ing and placing of ornament, it lacks the rich imaginative qualities of Amisfield, which is probably a slightly later and more fully de¬ veloped example of the designer’s genius. The masonry is of rubble with red ashlar dressings. On plan, at base, the structure is almost square, measuring exteriorly 28| feet from north to south by 3i|- feet from east to west, and rises to a height of 47 feet at the wall-head and to a total height of 77 feet above the ground. Beneath the wall-head there are four storeys ; within the roof is an attic with a garret above. On the south-east angle a containing the main flues is crow-stepped, with hollowed steps, as at Elshieshields. From the north-east, north-west, and south¬ west angles turrets of two storeys project and are covered with conical roofs. The door¬ way, which is in the south wall, is lintelled and has moulded arrises to head and jambs. The main windows of the south or front elevation have jambs and lintel moulded with a double roll separated by a nail-head enrich¬ ment, but the remaining windows throughout the building have rounded arrises. The dormer in the south wall over the entrance has a triangular pediment bearing a cable mould and shafted jambs with a machicolation in the sill. On either side of the window in the south wall at second-floor level there is a panel containing armorial bearings. The panel 196 Ajicient and Historical Monuments — Dumfries. Tcfaccp. 196 . Fk; 141. -Aniisfield Tower : North Face (No. 57S). Fig. 142.—Amisfield Tower : South Face (No. 578). 3 F # f IT i 4 *» 1 I i i 1 j Am tent and Historical Monutncnts—Dumfries. Fig. 143- To face p. 197- COLOURED PLASTER FRIEZE IN HALL AT AMISFIELD TOWER. TINWALD.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [TINWALD. mouldings are the double roll separated by a double row of nail-heads. The dexter panel has the date 1600 and the Lombardic initials I C flanking a shield bearing a fess for Char- teris. The sinister has the same date with the initials A. M. flanking the shield, which is charged : Quarterly, ist and 4th, a saltire with a label of three points in chief, for Maxwell, and 2nd and 3rd three hurcheons (hedge¬ hogs) passant, for Herries. John Charteris of Amisfield had married Agnes, daughter of Sir John IMaxwell, Lord Herries. A shield on the south gable of the cap-house appears to bear a fess for Charteris. The wall-head course has a billet moulding or rather two rows of miniature corbels with a nail-head enrichment between. At a higher level these mouldings are carried as a string round the stair turret, which is further ornamented with a cable moulding on the gable skews and chimney cope, and with a nail-head enrich¬ ment on the corbelling at third-floor level. When the building is entered a staircase is seen on the east and a mural chamber—a guardroom or porter’s lodge—on the west, which has a loop commanding the south, a slop-drain in the west wall, and a small recess in the west end of the north wall. Beyond this room to the north is the vaulted basement chamber, which is L-shaped, as a portion lies under the staircase. The south, east, and west walls have small gunloops, which are now built up. The straight staircase rises g steps to first-floor level and the foot of the wheel-stair at the south-east angle, by which access is obtained to the second and third floors. The first floor is an L-shaped chamber through the inward projection of the staircase, and measures 20 feet from north to south by 21 feet from east to west. It is floored with red sandstone flags and ceiled wdth wooden joists, which rest on corbels north and south and support the flooring above. It is lit by win¬ dows to south, east, and north ; the west wall contains a large fireplace with shafted jambs and moulded caps and bases of the common type of the period ; in the breast a later mantelpiece is inserted ; there is a dog-legged recess in the south jamb with the original moulded and carved oak facings. South of this the wall is recessed to contain a buffet. A cupboard with a projecting stone sill is formed in a bulkhead over the staircase. At either end of the north wall there is an L-shaped cupboard recess set some 7 feet above the floor. A garderobe with a lamp recess and small window occupies the northern end of the east wall, and at the south end there is a lamp recess. The walls have been plastered “on the hard,’’ and in one or two places the re¬ mains of a contemporary plaster cornice is seen. Early plaster cornices, it should be noted, are of a flowing plastic form suitable to the material, and are not based on stone and wood forms, as is the bulk of the Renais¬ sance and modern work. Between the first and second floors a small mezzanine chamber is contrived over the straight staircase, and was entered from the wheel-stair ; it is now filled with rubbish and inaccessible. The second floor is a chamber almost square, but broken by the inward pro¬ jection of the wheel-stair at the south-eastern angle. In each wall there is a window ; two have window seats. The fireplace in the west wall resembles that on the floor below and has a recess in its south jamb. There is an L-shaped recess at the western end of the north wall, a garderobe with window at the northern end, and a lamp recess at the southern end of the east wall. The walls have been plastered ; over the fireplace and on the north wall and on the stair projection, a pattern of colour decoration of white, yellow, red, gray, and black can be traced on the thin plaster (fig. 143). An oil medium appears to have been used, not a tempera, but this cannot be accu¬ rately determined without chemical analysis. The third floor contained two intercommuni¬ cating chambers separated by a timber partition ; each has a fireplace in the gable and is lit from a window in each wall. From this floor the angle turrets are entered at a higher level; the north turrets are reached from the stepped breast and through the in¬ going of the gable windows. They each contain a garderobe in the north wall. The sill of the south-west turret rises sheer above the floor and probably was reached by wooden steps. All have a window for light and narrow slits commanding the wall faces of the tower. The turrets have an upper storey entered from the attic chambers ; the only features therein are the loopholes placed to command all directions. The main wheel-staircase stops at the third- 197 TiNWALD.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [tinwald. floor level, and the ascent is continued by a smaller wheel-stair to the north-east, which gives access to two small floors, each of a single chamber, above the main wheel-staircase. The attic and garret call for no special mention. The cap-house, in which the small wheel-stair terminates, is an oblong chamber with an ingeniously contrived fireplace in its north gable and openings in each wall. There is preserved in the National Museum of Antiquities an oak door from the tower, 5 feet 7 inches by 3 feet by inches (fig. 143A), having carved upon it the date 1600 ; A. M. in monogram on the sinister side, and a fragmentary J. C. on the dexter; a shield quartering the arms of Charteris and Herries {cf. p. 197); a hat; and a full-size representa¬ tion of Samson, in contemporary costume, slaying the-lion. Grose says the figures were painted {Antiquities, vol. i. p. 158). The work of the same artist appears on a door dated 1601 from Terregles House {Book of Caer- laverock, i. p. 530). The structure is well roofed, but the walls urgently require pointing, and the floors are in bad condition. Close to the tower is the modern house, the western portion of which incorporates a structure described by Grose as “built in the reign of Charles L, a.d. 1631 as appears by a date inscribed over the chief entrance ” {Grose’s Antiquities, vol. i. p. 157). His view of it is dated 1789. The date is not now to be seen, and the old house has been absorbed in the large modern mansion. On Charteris of Amisfield see Introd., p. xxix. In 1636 Sir John Charteris and John, his son and heir, resigned the barony of Amisfield, which passed into the hands of Sir John Dalziel of Newton {Reg. Mag. Sfg., s.d.. No. 609. Cf. No. 552). xli. S.E. 23 July 1919. Defensive Constructions. 579. Fort, Pinnacle Hill.—This fort is situ¬ ated about 150 yards to the south of the sum¬ mit of Pinnacle Hill (687 feet above sea-level). It is somewhat oblong in form. The north¬ west end of the enclosure is represented by a broad mound of earth which crosses the ridge and passes down the hillside to the north¬ east. On the other three sides there are the remains of a rampart and trench, best pre¬ served on the south-east, where the former has an elevation of some 3 feet on the inner and 5 feet on the outer face, and the trench a width of 20 feet and a depth of some 3 feet below the counterscarp. For a considerable space in the middle of the south-west side the defences have disappeared. The dimensions of the interior are about 160 feet from north¬ west to south-east and 150 feet from north¬ east to south-west. The entrance has been from the south towards the west end of the south-east side. xlii. S.W. 20 August 1912. 580. Fort, Whitehill, Hightown Hill.—On the north end of the White Hill, an eminence which rises from a region of hillocks domin¬ ated by Hightown Hill to an altitude of 800 feet, and about J mile to the south-east of the latter hill, is the site of a defensive en¬ closure on a plateau recognisable by a low rampart curving across the brow of the hill towards the north, but the oval indicated on the O.S. map is not traceable in any other artificial work. xlii. S.W. 20 August 1912. 581. Fort, Barr’s Hill.—Crowning the highest point of the summit of Barr’s Hill, a pro¬ minent height which rises to an altitude of 714 feet over sea-level, and about miles to the north-north-east of Tinwald Church, is a well-preserved fort. The hill rises by a steep gradient from all sides except the south, where, below the actual summit, the slope up¬ wards is long and gradual, while the summit trends away to east and west in irregular hillocks and hollows. The hill commands a most extensive panorama over Nithsdale, the upper reaches of Annandale and, across the hills that bound the western side of the latter dale, to where the table-topped Birrenswark swells up on the horizon on the south-east. The fort is oval in form, lying with its longest axis north-east and south-west, meas¬ ures across the interior area 274 feet by 210 feet and has been surrounded by a parapet mound, now only discernible a few inches in height on the north. A steep scarp, with an average height of 10 feet all round, gives on a trench varying in breadth from 198 Ancient and Historical Monuments — Dumfries. Fig. 143A.— Door from Amislield Tower, now in the National Museum of Antiquities. To face p. 198 . TINWALD.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [TINWALD. 37 feet between crest of scarp and counter¬ scarp on the west to 29 feet at the north-east, flat-bottomed, with a breadth at floor level of some 12 feet and at deepest 3 feet 6 inches in depth below the crest of the counterscarp. Beyond an intervening mound is an outer concentric trench, which measures 25 feet from crest to crest, 12 feet across its flat floor, and at most 6 feet in depth below the crest of the scarp and 3 feet 6 inches below that of the counterscarp. This outer trench is, however, less regular in its continuance than the inner Fig. 144.—Fort, Barr’s Hill (No. 581). one, and, where the ground has fallen away sharply from the base of the scarp of the intermediate mound, it has probably been discontinued. A hedge crosses the interior from north-east to south-west, and on the south-east half of the periphery the defences have been much more destroyed by the action of the plough than on the other. The entrance, 10 feet wide, has been from the east, carried directly over the outer defences and into the interior by a roadway sunk as much as 5 feet below the level on either side as it passes over the scarp. There has been a considerable amount of rock¬ cutting in the formation of the fort, and the ramparts arc formed of earth and stone, the upcast of the trenches. xli. S.E. 22 August 1912. 582. Mote, Tinwald. — Situated on gently rising ground some 130 yards to the south of the parish church of Tinwald, with an exten¬ sive prospect over Nithsdale, stands the Mote of Tinwald—a truncated cone, some 9 feet to 10 feet in elevation, with an oval summit measuring 62 feet by 36 feet. The slope of the mound is by no means steep, and the whole construction has probably been reduced by cultivation. It is probably a natural gravel hillock, the levelled summit being the only really signiheant feature remaining. This was the capitate messuagiuni appellatum le Mote pixta ecclesiam de Tyndwalde upon which sasine of lands was given by the cere¬ mony of handing to the grantee, before witnesses, a handful of the earth and stone from the hillock (cf. Book of Caerlaverock, ii. p. 434, No. 41). xlix. N.E. 22 August 1912. Enclosure. 583, Amisfield Enclosure.—Rather less than 100 yards to the west of Amisfield Tower there is an irregular quadrilateral area, which has been enclosed by a stone wall, encompassed by a ditch ii feet wide at base and 7 feet deep, with an outer rampart on the crest of the counterscarp. The enclosure, which measures 154 feet from north to south by 143 feet from east to west, has apparently been entered from the south. It obviously belongs to a much earlier period than the tower. xli. S.E. 23 July 1919. [Miscellaneous. 584. Architectural Fragment, Tinwaldhouse Cottage. — Tinwaldhouse Cottage stands slightly more than a mile south-south-east of Tinwald, at the entrance to the loaning leading to Tinwald House ; in one of the gables is inserted a late 17th-century architectural fragment executed in red freestone. It con¬ sists of a portion of a delicately moulded archivolt surmounted by a heraldic achieve¬ ment contained within a cartouche. On either side of the cartouche is a graceful festoon executed with spirit and reminiscent of the carving on the front of Drumlanrig Castle ; the cartouche is roughly an upright oval in form and is elaborated with scroll work ; it is sur¬ mounted by a helm and mantling and ter¬ minates at foot in two cherubs’ heads placed on a label which is carried over the archivolt mouldings to resemble a keystone. Within the 199 tixwald.] historical MONUMENTS (SCOTLAND) COMMISSION, [torthorwald. cartouche is a shield charged ; Quarterly, ist and 4th, on a saltire a crescent; in base a holly leaf (Maxwell) ; 2nd and 3rd, three roundles or bezants (tor Mundell. Cf. Book of Caer- laverock, ii. p. 434, No. 41). Below the archi- volt, on an infilling, the initials J. S. MR (?) are incised, but these have no connection with the fragment. xlix N.E. 585, “Tumulus,” Broomrig.—At the bottom of a field, in low-lying ground some 350 yards to the south-west of Broomrig, is a gravelly knoll sparsely covered with vegetation, marked “tumulus” on the O.S. map. It has a diameter of 54 feet and an elevation of about 4 feet. Its sepulchral character is not certain. xlii. S.W. 20 August 1912. 586. Tinwald Churchyard.—Some 36 feet to the north of Tinwald Church lies a table-stone of red sandstone, 5 feet 10 inches in length by 3 feet in breadth, inscribed ;— HERE LYES THE CORPS OF JOHN Corbet who dyed the 17 of MARCH 1706 AND OF HIS AGE 63 YEARS WHO WAS TAKEN IN THE YEAR 1684 BY A PARTI OF CLAVERHOUS HIS TROUPE AND WAS BANISHED BY THE WICKED COUNSELL OF SCOTLAND TO EAST JARSEY 1685 AND RETURNED IN THE YEAR 1687 and at right angles to the foregoing ;— THIS STON LET SPEAK WHEN SPEECH IS FROM ME GONE HOW GOD ME LIEDD WHEN I WAS FAR FROM HOME BANISHT I WAS FOR COVENANTED CAUSE AND NONE COMPLYANCE WITH THEIR WICKED LAWES GOD WHOM I SERVD MADE ME THERE FIRME TO STAND BROUGHT BACK AGAIN UNTO MY NATIVE LAND MY SOBER WALKE IN EACH PLACE OF ABOAD MADE ME BELOVD OF ALL THAT LOVED GOD GAINST ALL ASSAULTS FROM FIRST UNTO THE LAST ASSISTED ME I GOTT NOT A WRONG CAST HIS PRECIOUS TRUTH FRAGRANT THO WAS TO ME FROM FIRST TO LAST AS I LIVE SOE I DY ONCE MORE IL LIVE, AND NEVER DY AGAIN AND SING HIS PRAISE IN A TRIUMPHANT STRAIN. On the north wall of the burial enclosure of the family of Charteris of Amisfield is a panel, dated i6i8, bearing a shield charged with a fess. On the dexter side the letters (?) I C ; on sinister D M F. xlix. N.E. 22 August 1912. 587. Trailflat Churchyard. — Towards the west end of the burial ground are a number of table-stones commemorating various genera¬ tions of Patersons in Skipmyre—the family of the founder of the Bank of England. One of these, date 1694, shows a well-cut inscrip¬ tion in relief, surrounded by a vine border with bunches of grapes. xlii. S.W. 20 August 1912. Sites, 588. Fort.—The fort on the hill, J mile north-west of Annfield, is a mere site. xli. S.E. 2 May 1913. The O.S. maps also indicate sites as under :— 589. Fort, near Slacks, xlii. S.W. TORTHORWALD. Castellated and Domestic Structure. 590. Torthorwald Castle.—A mere shell, with one corner quite gone and in parts buttressed and repaired with modern building, is what re¬ mains of the tower of Torthorwald (fig. 145) about four miles north-east of Dumfries. The site is an eminence 250 feet above ordnance datum, steeply scarped on the north and east. The tower, which is probably of the late 14th century, stood on the line of a curtain wall forming an enciente, and the whole was en¬ compassed by outworks of ditch and rampart. On the west and north-west these outworks have disappeared under cultivation, but the remaining portions are comparatively com¬ plete. The tower is placed with its major axis almost north and south and slightly south of 200 Ancient and Historical Monnnients--Dumfries To face p. 200 . Flc. 145.—Torthorwald Castle (No. 590). TORTHORWALD.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [torthorwald. the mid point of the summit ; from the south- around the perimeter of the summit, until it western angle a curtain wall has returned abutted on or was incorporated in the wall I'lG. 146.—Torthorwald Castle (Xo. 590). southwards to the extremity ot the enceinte, of a quadrangular enclosure north-west of the where the configuration of the ground suggests tower. At this point evidence for the course that there has been a circled tower ; from of the wall disappears, but on the west wall this the curtain has turned eastwards and of the tower the tusking of the curtain is 201 TORTHORWALD.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [torthorwald. seen about 9 feet from the south-western angle, whence it may be conjectured that the curtain ran west and then north towards the enclosure aforesaid. The enceinte so formed is roughly L-shaped and measures from north to south 183 feet b}' 130 feet at most from east to west. The enclosure has been some 60 feet by 50 feet. The enceinte is girt by a great ditch and outer rampart on the north, where the dimensions are 70 feet from crest to crest udth a depth of 15 to 20 feet. On the south and east the ditch lessens considerably; beyond it on the east there is an irregularly shaped enclosure flanked by a wet ditch, through which flows a burn apparently once spanned by a small bridge. Still farther east¬ ward and some 200 feet out from the enciente there is a rampart, which returns north for a distance of 380 feet and then westward for 120 feet, where it abuts on the escarp of the ditch. The entrance to the enceinte was probably through the south curtain. The tower measures 40 feet 8 inches from north to south by 20 feet 4 inches from east to west within walls some 9 feet 3 inches thick, which still stand to a height of 60 feet in parts and contain two vaulted storeys. The entire north-western angle of the building has dis¬ appeared except a fragment, which does not appear to be in situ. The walling shows three types of masonry: whinstone rubble, a reddish- coloured ashlar, and whinstone rubble with freestone dressings, the last mentioned prob¬ ably being the earliest. The north gable, the vaulting, and the upper portion of the east and south walls are of ashlar, resembling the work of the 15th century, and suggest that a re¬ construction was carried out at that time. During this reconstruction the building was extended northwards to its present dimensions ; the junction between the two portions is seen in the straight joint on the interior of the east wall above the lower vault and on the soffit of the upper vault. The ground or basement floor comprised two unequal chambers probably intercom¬ municating. The southern and larger has been vaulted with a semicircular barrel-vault of ashlar springing from the cast and west walls, the northern with a segmetal vault also of ashlar springing from the north gable, and a partition on the south which has disappeared. These chambers do not appear to have held an intermediate floor below their vaults. The vaulting is almost entirely destroyed. The entrance to a circular wheel-stair, leading from the ground to the first floor, is in the north-east angle of the north chamber. At the western end of this chamber there is a circular depression which may indicate a well. The entrance to the tower was probably placed in that portion of the walling at the north-west angle which has disappeared, for it cannot be traced in the walls remaining ; but on the exterior of the east wall are seen the remains of a former doorway, which was evidently built before the lower vault was in¬ serted, as the head is above the vault and the sill about the springing level. In its infilling a small window is formed above the springing level, corresponding with another opposite, which windows lighted the upper part of the semicircular vault. These windows have stepped soles ; the western is complete and has a very flat cavetto wrought on lintel and jambs. At a lower level there is a ruinous window in each of the lateral walls. To the north of the eastern window is the entrance of a mural chamber which is now ruinous ; between these openings there is a cupboard recess. The south wall has a central recess about 7 feet wide, from which a straight stair leads from the level of the springing of the vault to a wheel-stair in the south-western angle giving access from the first-floor level to the upper floor or floors. The first floor is ceiled with a lofty barrel- vault, which is fairly complete at the southern end. There have been fair-sized windows to the north, east, and west. The south wall contains a very ruinous fireplace. Above this vault the structure is so ruinous that the arrangement cannot be defined. On the third course beneath the corbelling in the entrance of the north-eastern staircase there is an incised mark consisting of a perpen¬ dicular terminating in parallel arms with another crossing midway. The upright line is 5-|- inches high, and the arms are 3^ inches wide. The junction of the mid-arm with the shaft is lozenge-shaped. Immediately adjoining is an acutely sloped line of the same height as the crossed figure. In the charter of 1321, granting the land of Torthorwald to Humphrey de Kirkpatrick, the “ pertinents ” are specified without any 202 TORTHORWALD.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [tundergarth. reference to a castle, which apparently there¬ fore did not exist at that date. {Reg. Mag. Sig. (1912), p. 457; Annandale Family Book, i. p. 8). In 1585, during the Maxwell- Johnstone fei’d over the Wardenship, Lord Maxwell was proposing to place a garrison in “ Tortarrel ” {Border Papers, i, p. 191). See further on Torthorwald, Introd., p. xxvii. 1 . S.W. 22 July 1919. Defensive Construction. 591. Fort, Camp Hill.—The Camp Hill is a commanding eminence that rises north-east of Torthorwald to an altitude of 612 feet above sea-level, and from which a most extensive view is obtained over Nithsdale. Its upper eminence has been converted into a fort, with probably a double rampart and inter¬ vening ditch, supplemented over a projecting shoulder to the west by an additional ram¬ part, but from long cultivation the defences are now low and inconspicuous. The enceinte appears to have been an ellipse measuring some 400 by 350 feet. 1 . S.W. 22 August 1912. Miscellaneous. 592. Inscribed Slabs, Torthorwald Church. —Set in the outer face of the west wall of the vestry is a slab of red sandstone, imperfect at one end and reduced along the sides, meas¬ uring 5 feet 4 inches by i foot 5 inches, on which is carved, in relief, a floriated cross, from which the base has been broken off. On the sinister side of the cross-shaft is a double- edged sword, with a segmented pommell and depressed quillons, surmounted by a shield bearing a saltire and a chief. On the dexter side of the cross is part of an inscription, with the date 1450. The inscription has been rendered unintelligible apparently by re¬ cutting. Built into the north side of the vestry is a M small stone, inscribed with the letters W H and the date 1644. 1 . S.W. 22 August 1912. Sites. 593 - Fort, Garrs Hill. — Around the summit of the Garrs Hill, some f mile west-north-west of the parish church, is the faintly discernible outline of a low rampart containing an oval enceinte. 1 . S.W. 22 August 1912. The O.S. map indicates a site as under :— 594. Church, about 300 yards east of Tor¬ thorwald Castle. 1 . S.W. TUNDERGARTH. Defensive Constructions. 595. Fort, Crawthat Cottage.—On a project¬ ing plateau at the base of a hill in rear of Crawthat Cottage, and on the side of the road from Lockerbie to Langholm, is a fort. The plateau has a general elevation of from 30 feet to 40 feet above the wet meadows to east and west of it and the roadway over what has also been marshy ground on the south ; Fig. 147.—Fort, Crawthat Cottage (No. 595). to the north the hillside slopes upward by an easy gradient. The fort, which is approxi¬ mately oval, measux'es interiorly 190 feet by 170 feet, its longest diameter being from north¬ east to south-west. It is divided into two distinct divisions by a trench, 18 feet wide, dug across the interior from north-east to south-west and cutting off about one-third to¬ wards the south-east, which forms a separ¬ ately defensible outer ward. The main area is surrounded to the outside by a rampart some 4 feet high at highest on the interior, with a trench in front, 24 feet in width and 6 feet deep below scarp and counterscarp, both sharply cut to a floor now 6 feet broad. 203 TUXDERGARTH.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [tundergarth. On the south side, facing the lower ground, there is a mound on the countei'scarp some 15 feet wide at base. The trench, which separates this area from the forecourt, is slightl}’ constricted at each end and is cai'ried into the trench of the main defence. The outer ward is surrounded by a rampart rising above the sectional trench across the interior and carried round the outside, with traces of a trench outside it. Somewhat to the north of the centre of the east-south-east arc is an entrance, which appears to have been some 7 feet wide, through which a roadway, distinctly traceable, leading up the bank from the eastward, passes diagonally across to a gap in the cross-trench and rampart opening to the inner ward. As far as ascer¬ tainable, this inner entrance has measured some 12 feet across. On the north side of the inner area, partially set on the rampart, is a hut-circle, measuring some 25 feet in dia¬ meter interiorly and surrounded by an earthen bank some 18 inches high and 6 feet broad, the entrance through which appears to have been from the west-south-west, where there is a slight depression, not, however, carried to the floor level. At the entrance to the outer ward the rampart on either side projects out¬ ward, and has probably been originally carried forward to the counterscarp of the trench, now cut off by a modern roadway from the cottage. xliv. S.W. 31 July 1912. 596. Fort, Haggy Hill.—This fort occupies the extreme north end of Haggy Hill, a long grassy ridge which extends northward at an elevation of 1000 feet over sea-level from the Doe’s Hill overlooking Winterhope Head. To west and north the ground declines rapidly, reaching away in the former direction to the glen of the Capel Burn. The fort is oval, with its main axis north and south, and measures interiorly 204 feet by 163 feet. It is surrounded by a stony rampart, low towards the interior but with a steep scarp falling some 12 feet to a trench cut through rock, varying in breadth from 29 feet to 35 feet and having a counterscarp from 6 feet to 7 feet high on the upper side of the fort and 3 feet on the lower. The rampart has been sur¬ mounted by a stone parapet, the founda¬ tions of which, visible at many points, measure 5 feet in breadth. The entrance has been from the west, some 6 feet wide, and on either side of it there has evidently been stonework, which was carried forward to the ends of the trench. In the south-east part of the area there is a depres¬ sion, measuring some 36 feet by 60 feet, in which rushes are growing thickly, and which appears to have been excavated. The view from this fort is most extensive, comprising all the country which lies within a girdle of mountains from Criffel to Ettrick Pen. xliv. S.E. 31 July 1912. 597 - Fort (remains of), Mosshead Hill.— To the south of the summit of Mosshead Hill, and at a slightly lower elevation, are the slight remains of a sub-oval fort, which has measured some 250 feet by 210 feet. The hill slopes away towards the south. On the upper side the defences are represented by a scarp, and only on the west and south do portions of the rampart remain, while on the west there is also about 120 feet of a trench outside, measuring 44 feet in width from crest to crest, 7 feet in depth below the scarp, 4 feet below the counterscarp, and 14 feet across the flat bottom. The entrance, some 13 feet in width, has been from the south¬ west. The fort faces Birrenswark Hill, which lies about miles to the south-south-east. li. N.E. I August 1912. 598. Fort, Whitstone Hill. — This fort occupies a commanding position on the crest of Whitstone Hill, rising in a bend of the Water of Milk to a height of some 250 feet above the stream and of over 600 feet above sea-level, commanding a prospect of many miles down the vale westward to Annandale and north-east to the head of Eskdale. The construction is elliptical in form, and is sur¬ rounded by a massive rampart rising some 4 to 5 feet above the interior. To the exterior the scarp is maintained at a fairly regular height of from 6 feet to 8 feet, and, where the ground beyond is high enough to n' cessitate its being cut through, there is a trench at its base some 30 feet in width, except beside the entrance, where it has a width of 35 feet. The entrance is from the west and is some 12 feet wide, crossing above the trench on unexcavated ground but sloping downwards from the level of the crest of the counterscarp. 204 TUNDERGARTH.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [tundergarth. The top of the rampart is stony, as if there had been a pitching of flat stones upon it, but there is no sign of a wall. The trench, for the most part, is cut through rock. xliii. S.E. I August 1912. 599. Forts, Newland Hill.—The first fort to be noted on this hill is situated on the western end of the summit at an elevation of 1004 feet over sea-level, and commands a most exten¬ sive view over Annandale to the hills that girdle the horizon from Criffel to Ettrick Pen. In form it is oval, lying with its main axis north- north-west and south-south-east, measures 225 feet by 157 feet and is surrounded by a stony rampart formed in great measure of shivers of rock produced by the excavation of the trench which lies at the foot of it. On the interior the rampart has a height of some 3 feet and is stony on its crest—a condition due either to a pitching of rather large stones or to the founda¬ tion of a parapet wall. The trench measures some 30 feet in width from crest to crest, and has a depth varying from 7 to 12 feet below the crest of the scarp and 3 feet 6 inches or thereby below that of the mound which forms the counterscarp. There are two entrances into the interior, both opening on excavated hollows, one from the north-east, as far as ascertainable about 7 feet wide, and the other from the south¬ west, of similar width. From the right side of the latter, on entering, a wall crosses the interior, cutting off about one-fourth of the area behind the south arc. The interior has not been levelled, nor, except perhaps in front of the entrances, has it been hollowed by excavation : it inclines gradually towards the east to an eminence apparently of rock. To the left of the south-west entrance and 25 feet distant from it, in the end of the trench, there appears to be a well. The trench, which broadens out by the well, contracts at 66 feet beyond it, where the rampart is much more massive than elsewhere, and the counterscarp has a height of from 5 feet to 6 feet. The second fort on Newland Hill, some 450 yards to the eastward, is adjacent to the actual summit of the hill, at an elevation of 1000 feet over sea-level. It consists merely of a ramp, some 6 feet to 8 feet in height on the south¬ west quadrant of a circle, with a slight hollow at its base. There is no indication that the construction has even been proceeded with further, as the fragment terminates abruptly and the rest of the supposed periphery is not naturally protected. xliv. N.W. 13 August igi2. 600. Fort, Craighousesteads Hill.—This fort occupies the summit of Craighousesteads Hill at an altitude of 850 feet over sea-level. The hill lies in the midst of an undulating ex¬ panse of round grassy hills, furrowed by the glens of numerous burns flowing towards Annandale. The only lowland prospect into the dale is down the valley of the Water of Milk, which flows by at the base of the hill on the west. The fort is oval on plan, lying with its longest axis north by west and south by east, measures 183 feet by 155 feet, and is surrounded by a rampart, once seemingly surmounted by a wall but now overgrown with grass. The rampart is low on the inner side but presents a scarp to the exterior, varying from 4 to 6 feet in height. Where the ground outside has risen towards the fort it has been cut through by a trench, and on the south-east such a cutting is through rock. The entrance has been from the west at the lowest point of the enceinte, but the rampart on either side has been too much worn down to admit of approximate measurement. xliv. S.W. 16 August 1912. 601. Forts, Newhall Hill.—Two forts lie on the summit of Newhall Hill, an eminence which rises to a height of 889 feet over sea-level and occupies an important strategic position at the divergence of two passes, the one by the head waters of the IMilk into Upper Eskdale and the other via Wauchopedale into Langholm. The forts lie within 40 feet of each other, the one on the actual summit of the hill, with a view in all directions, and the other to the east on the brow. The upper fort is oval in form, lying across the strike of the hill with its longest axis east and west, and measures 198 by 160 feet. It is surrounded by a rampart 22 feet wide at base, formed of the splintered rock and earth of the trench, which lies in front with a mound above the counterscarp. The rampart rises but slightly on the interior and shows the usual indication of stonework on 205 TUXDERGARTH.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [tundergarth. its crest ; and where the stones are exposed, as on the south arc, a number of them appear to be inserted not on the bed but on end. The scarp of the mound has a general height of some 5 feet ; the trench is 22 feet wide, and at deepest 4 feet below the crest of the counterscarp, but on the north-west is barely discernible. There are two entrances, one from the east and the other from the south¬ west, each some 9 feet to 10 feet in width. The interior is not level but rises towards a point on the north immediately behind the defences, and is generally uneven over the surface. To the east of the highest point, at the back of the rampart, there is an indication of a foundation, seemingly oval, but too in¬ definite for measurement. The trench has in places been cut through rock. The lower fort, at the nearest point of the counterscarp of its trench, is only some 40 feet distant, and, though of similar character to its neighbour, is at all points more formid¬ able. In plan it is oblong, with each side slightly convex, lying with its longest axis north-east and south-west and measuring 215 feet by 183 feet. Its defences consist of a massive rampart and a trench without, vary¬ ing in width from 25 feet to 30 feet and of an average depth of 6 feet below the crest of the scarp and of 4 feet below that of the counter¬ scarp, except on the south-east, to the north of the entrance, where it is almost eliminated. The interior has to some extent been ex¬ cavated, giving prominence to the rampart which, from a height of 5 feet to 6 feet at its crest, merges gradually into the slope of the interior area. The entrance, which faces the south-east, measures 8 feet in width and opens on an excavated hollow running back some distance into the interior. The rampart is formed of the splintered rock from the trench and shows the same indication of stonework on the crest as that of the adjacent fort, though the greater part of the stonework has probably been torn out to supply material for an old wall that crosses the hill top. At one point on the south, where the turf has been broken by sheep, there is a suggestion of stone pitch¬ ing on the front ; but excavation is really necessary to demonstrate this fact. Both forts overlook that above Crawthat Cottage to the south-east, and hold in view to the northward the forts on Craighousesteads (No. 600) and at Whitcastles (No. 298). xliv. S.W. 16 August 1912. 602. Fort, Crawthat Hill.—On the summit of the eminence which rises to the north of the hamlet of Dalbate and overlooks the road from Lockerbie to Langholm, are the remains of an oval fort, lying with its longest axis north and south and measuring about 170 feet by 115 feet. From the fort the ground declines steeply all around but especially so on the north-east, in which direction the defences were probably slighter and have now conseqiiently disappeared, while towards the south there is a gap of some 70 feet in the periphery. A rampart of earth and stone has contained ‘the enceinte, rising only a little above the interior level but with a scarp some 5 to 6 feet in height. There are indications of stonework on the crest, and, where a break has been made by sheep, there is a suggestion of stone pitching. A large natural mound, some 6 feet in elevation, rises in the centre of the enceinte. xliv. S.W. (unnoted). 16 August 1912. Stone Circle. 603. Stone Circle, Whiteholm Rig.—At the eastern end of Whiteholm Rig, a broad grassy ridge, rising to an elevation of some 600 feet above sea-level and about J mile east-north¬ east of Staijdburn Cottages, on the road from Lockerbie to Langholm, are the remains of a stone circle known as the “ Seven Brethren.” It appears to have had a diameter of about 62 feet, and to have consisted of at least twelve stones, of which seven and possibly eight remain. Of the seven stones four are upright and in situ — slabs of whinstone, varying from 8 inches to i foot 8 inches in thickness, from 2 feet to 3 feet 4 inches in breadth and from i foot 2 inches to 2 feet 4 inches in height above ground. The possible eighth stone is indicated by a mere point protruding through the turf to the north-east of the centre of the circle. The largest slab lies much tilted over towards the east-south¬ east, and is exposed for a length of 4 feet 7 inches with a breadth of 3 feet. Due north of the centre is a large slab, 3 feet 3 inches broad, 10 inches thick, but broken off at 14 inches TUNDERGARTH.J INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [tynron above the ground level ; and at 28 feet beyond it, also due north of the centre and outside the periphery, lies a block of stone 3 teet 5 inches in length by i foot 6 inches In breadth. It is impossible to say whether this stone lies fallen in its original position or has been re¬ moved from the circle. From the site of the circle there is an extensive prospect up the glen of the Water of Milk to the distant hills of Eskdale. xliv. S.Y". I August 1912. Hut-Circles. 604. Hut-Circles, Linnhall.—On the moor¬ land, close by the hill road from Linnhall, near the Back Burn, and almost mile south of the former place, are three hut-circles with low banks of earth or turf—one measuring 30 feet by 28 feet interiorly ; the second, some 60 yards further down the burn, 26 feet ; and the third, some 30 yards away from the last, 24 feet. Opposite the second, on the opposite side of the road, is an oblong, flat-topped mound, with an elevation of about 2 feet, evidently the ruins of a clay or turf hut. lii. N.W. 17 August 1912. Miscellaneous. 605. Enclosure, etc., Dalbate.—The O.S. map marks a fort 250 yards north-north-west of Dalbate and less than 120 yards due west of Dalbate burn. It is not a fortified site but a large enclosure surrounded by a turf bank and containing a smaller one within it. Within the larger, high up on the east side, is a hut-circle, measuring 24 feet in interior diameter and surrounded by a turf bank some 18 inches high ; and at the base of the hill is another, measuring 21 feet by 19 feet. In neither case is the entrance apparent. xliv. S.W. 31 July 1912. 606. Memorial Slabs.—To the south of the cast end of the old church stands a freestone stone, inscribed on the back :— HERE LYES JAMES JOHNSTON OF NETHER C.\STLEHILL WHO DIED OCTOBER 8 I711 AGED 39. On the front of the stone is carved the Johnstone arms : A saltire, on a chief three cushions, with a richly-carved mantling. Be¬ neath, Adam and Eve standing on either side of the tree, around which the serpent is coiled. Around the lower part of the stone is inscribed :— THE SERPENT AND THE TREE WERE FATAL TO ADAM’S POSTERITY. On a shoulder at each side is carved a human head. The whole treatment of this stone is more artistic than that usually meted out to gravestones of its class : the quality is re¬ markable, the lettering and carving being as sharp as the day on which they were cut. There are in the churchyard a number of other slabs of the end of the 17th or com¬ mencement of the i8th century, many of them bearing Johnstone arms and commemorating individuals usually described as having lived in credit and reputation and died in the Lord. li. N.E. I August 1912. Sites. The O.S. maps indicate sites as under :— 607. Fort, Southburn. li. N.E. 608. Castle, Tundergarth. li. N.E. TYNRON. Defensive Construction. 609. Tynron Doon.—Tynron Doon is a lofty, conical peak, springing from an eastern spur of Auchengibbert Hill, to which it is linked by a broad col lying 150 feet or thereby below its summit. On all other sides the Doon rises by a very steep inclination to an altitude of 946 feet above sea-level and some 600 feet above its base. On the top of this hill is a fort, which has given its name to the eminence, and which is remarkable for the massive nature of its defences. These, by the great natural advantages which the defence possessed on the other sides, have practically been confined to the west and south-west; and to the enclosure, by means of a bold stony parapet mound of a terrace, which lies some 30 feet below the summit on the north¬ east. The summit is sub-oval in form, with its longest axis north-east to south-west, and measures 150 feet by about 130 feet. Encircling it there appears to have been a 207 TVXRON.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. [tynron. wall built of boulders, man}- of them water- worn, which have been brought up from the vaUey below and are here and there to be seen protruding from the turf overgrowing the fort. To the east of the entrance, towards the south-west, as shown on the plan, some of the stones of this wall are coated with lime, and pieces of that material are to be picked up along its course for a short distance and on this section only. Below this summit wall lie three distinct segmental ramparts, with very bold scarps, rising to a height of 14, 14, and 19 feet respectively, with deep intervening hollows ; while at the lowest level to the outside is a slight trench, from which the lowest scarp rises, and which is cut through rock for almost its whole length, the rock face being exposed on the counterscarp. The termination of these defences to the northward is against the steep slope of the hill, the second or lower hollow being closed by an outcropping rock. The ramparts are formed of earth and stone— the excavated material from the intervening hollows ; and are acutely pointed to an unusual extent; the splinters of rock of which they are largely composed having evidently maintained their position well. From the westward, beyond the lines, a roadway approaches to the south-west arc of the lower rampart and passes through it. Within, a flanking defence is carried across the hollow between the outer and second ram¬ parts, the road being evidently kept to the east away from the main defences, and, ere it reaches the actual summit, it passes on the east a large circular depression, measuring 15 feet in diameter, which is seemingly the site of a hut-circle. On the west edge of the summit plateau are several other smaller depressions, which also seem to have been the sites of huts. At the end of the hollow that lies between the first and second ramparts, and just to the side of the steep track which at this point carries the roadway to the interior, is a well. In all directions, except the west and north-west, where Auchengibbert Hill intrudes, an immense prospect over hill and valley is commanded from the Boon, which itself is a very conspicuous object in the landscape for miles around. XXX. N.E. 9 June 1912. Cairns. 610. Cairn (remains of), Craigencoon Hill.— On the summit of the hill at Craigencoon are the remains of a large circular cairn, now re¬ duced almost to ground level, which has measured in diameter some 82 feet. Several large stones still stand on the periphery, sug¬ gesting that its outline has been marked in this manner. xxi. S.W. (unnoted). 5 May 1913. 611. Cairn, Craigencoon.—Some 300 yards directly in front of Craigencoon, and some 75 yards distant from the road, are the remains of a large cairn, which has measured in diameter some 52 feet. There has evidently been a cist in the centre, which has been rifled. xxi. S.W. (unnoted). 5 May 1913. 612. Cairn (remains of), Gledbrae, Holm- house.—On the east side of the loaning, the old road from Holmhouse to Moniaive bearing the name of the Gledbrae, are the remains of a very large cairn, greatly dilapidated, which has measured in diameter some 93 feet by 87 feet. Its periphery has been marked by large boulders, a number of which still remain in situ. In its present condition it is not possible to tell whether this cairn has contained a cist or chamber. Two large stones stand a little to the north of the centre, protruding above the debris and seemingly in situ, which suggest a chamber, but without 208 TYNRON.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [wamphray. excavation it is impossible to arrive at a definite conclusion. XXX. N.E. (unnoted). 5 May 1913. 613. Small Cairns, Old Auchenbrack.—On the south-west face of Cairney Knowe, a grassy brae that lies between the shepherd’s house at Old Auchenbrack and the Appin Lodge bridge, is a group of small cairns, varying in diameter from 12 feet to 18 feet and lying at an elevation of from 800 feet to 900 feet over sea-level. There is also the foundation of a larger cairn, measuring some 26 feet in diameter. xxi. S.W. (unnoted). 5 May 1913. Miscellaneous. 614. Martyr’s Stone.—Within thechurchyard of Tynron Kirk is a flat table-stone, 5 inches thick, 6 feet inches long, and 2 feet 9 inches broad, bearing at the top the inscription, across the breadth of the stone :— HERE LYETH WILLIAM SMITH SON TO WILLIAM SMITH IN HILL WHO FOR HIS ADHERING TO THE COVE NANTED WORK OF REFOR MATION WAS SHOT AT MINNYHIVE MOSS THE 29TH DAY OF MARCH 1685 HIS AGE 19 YEARS. THIS DEED WAS NOT DONE BY A COUNCIL OF WAR BUT BY COUNTRYMEN WITHOUT SYSE At the other end, and reading lengthwise :— I WILLIAM SMITH NOW HERE DO LY ONCE MARTYR FOR CHRIST’S VERITY DOUGLAS OF STENHOUSE LAURIE OF MAX- WELTON CAUSED CORNET BAILIE GIVE ME MARTYR¬ DOM WHAT CRUELTY TO MY CORPS THEN USED LIVING MAY JUDGE ME BURIAL THEY REFUSED. On another stone, 4 inches thick, 6 feet 10 inches long, and i foot 9I inches broad, are the inscriptions, separated by an hour-glass and skull and cross-bones, over which is the legend memento mori :— HOC • EST • SEPUL CHRUM • lOANNIS DOUGLAS • DE • STEN -HOUS 1683 AND MAJOR • JOHN • KIRKPATRICK WAS BURIED HERE 3 JAN 1634. XXX. N.E. 10 June 1912. 615. Killywarren, Tynron. — Killywarren farm-house is a small 17th-century dwelling presenting no special features of interest. Over the doorway of a modern wing, to the north, are two stones removed from the old house ; the upper bears a shield with arms: Quarterly, first and fourth, a heart, three stars in chief; second and third, a bend between six cross-crosslets fitchy; in the upper dexter quarter are initials, thus yj g and Q j. On the lower stone is the date 1617 and the initials I D and E D. These are Douglas arms. XXX. N.E. 10 June 1912. Sites. The O.S. maps indicate sites as under ;— 616. St Connel’s Chapel and Connel’s Well, near Kirkconnel. xxx. N.W. WAMPHRAY. Defensive Constructions. 617. “Mote,” Wamphray Place.—In the angle formed by the junction of the Leithen- hall Burn with the Wamphray Water, a number of low artificial banks and foundations of rectangular structures mark the site of Wamphray Place. The site is a rectangular plateau, flanked on the east and south by the deep glens of the two streams which flow some 40 to 50 feet below. At the edge of the eastern glen, overlooking the Leithenhall Burn, there rises to a height varying from ii to 13 feet a mound of clay and pebbles, flat on the summit and steeply scarped all round, evidently the remains of a mote hill. From the actual edge of the bank, by some secondary operation, it has been cut back for a distance of some 10 feet at base and 209 14 WAMPHRAY.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [wamphray. correspondingly upwards, so that the summit now forms a little more than a semicircle, with its longest axis north and south, which measures 32 feet by 20 feet. To north and south a trench cuts off the respective ends of the ridge on which the mote has been con¬ structed. On the west the foundations of an oblong rectangular building of late date have obliterated all indications of a trench that ma}^ previously have existed in this direction. For a sundial from Wamphray Place see No. 495 - M amphray of old belonged to the Corries, but in 1357 John of Corrie gitted the lands to Roger Kirkpatrick [Annandale Family Book, vol. i. p. ii). In 1450 Sir Robert Crichton of Sanquhar had a charter of the five merk- lands of Wamphray [Reg. Mag. Sig., s.d., ^0- 333 )- By November 22, 1476, James Johnstone was in possession of these lands and granted them to his son John [Annandale Family Book, vol. i. pp. 14-15). They passed from the Johnstones in 1747 by sale to the Earl of Hopetoun (Paterson’s Wamphray, P- 39 )- XXV. S.W. II September 1912. 618, Castle Knowe, Saughtrees.—This is a fragmentary site, situated by the side of a farm road some 300 yards to the north of Saugh¬ trees farm, and consists of a knoll or slight promontory, measuring some 27 feet across at base, which projects some 17 feet out from the face of a steep bank. A trench around it, 16 to 18 feet below the summit, measures some 12 feet wide at the level of the counter¬ scarp and 3 feet in height. The construction appears to be of earth and stone, but is within a wood overgrown with a forest of nettles and no doubt has been considerably interfered with by the roadway. xxxiv. N.W. 12 August 1912. 619. Fort, Dundoran.—In the north end of Dundoran Plantation, which crowns the southern termination of the long hog-backed ridge of Dundoran, are the remains of an elliptical fort, at an elevation of 800 feet above sea-level, now represented by two segmental ramparts, seemingly concentric, crossing the crest of the ridge at the north end, with a trench in front of the outer rampart. The two ramparts, which are some 64 feet apart. appear to be composed of small angular frag¬ ments of stone : the inner has an elevation of about 4 feet ; the outer, on the inner side, 5 feet and from the bottom of the trench in front, 6 feet. The trench, which has been to some extent cut through rock, measures 28 feet in width at the centre but gradually diminishes as it passes to either side. The fort is shown on the 25-inch O.S. map as an ellipse measuring some 260 feet by 185 feet, but, except as noted above, its outline has almost disappeared. The site is very defen¬ sible and commands the most extensive view over Annandale. XXV. S.W. II September 1912. 620, Fort, Catherine’s Hill, Poldean.—The site of this fort is the summit of a lofty hillock, which rises with a very steep bank to a height of some 60 feet to 80 feet above the Annan, which flows round its base on the west, I mile to the south of Poldean farm. The hill presents a steep slope in all directions and is fairly level on the summit. The fort, the outline of which is not now very distinct, appears to have been in form about two-thirds of an oval, the chord lying ob¬ liquely along the edge of the bank overlook¬ ing the river. The longest axis is north- north-west and south-south-east, and the interior measurements are 219 by 134 feet. The defences have consisted of a single stony rampart, now of slight elevation. The entrance appears to have been into the lowest part of the area from the east and through the centre of the curve. Near the centre is a low stony mound suggestive of a small cairn. xxiv. N.E. 20 September 1912. 621. Fort, The Dod.—This fort crowns the summit of a round-topped hill, which swells upward from the western side of Annandale to an elevation of 762 feet over sea-level, about I mile to the north-east of Poldean. To the north and north-east the hill descends steeply for 100 feet or thereby towards the glen of the Beldcraig Burn, while southward it falls away by a long easy gradient. The view it commands up and down Annandale is ex¬ tensive. Except along the steep slopes on north and north-east, the summit has been encircled by a rampart some 18 feet in thick¬ ness at base, rising where highest to a height 210 WAMPHRAY.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. [wamphray. of some 5 feet on the exterior and 2 to 3 feet on the interior, with a slight trench to the outside some 24 feet in width. The ram¬ part is formed of earth and stone, the upcast of the trench. As it approaches the steep slopes it diminishes in height, and on the north¬ west it is barely traceable. The interior area, which is approximately circular, measures some 280 feet in diameter. The entrance has probably been from the east, not far from where the gradient takes a steeper pitch to northward. xxiv. N.E. 20 September 1912. 622, Fort, Laverhay Height.—On the western slope of Laverhay Height, and on the south side of a hollow down which flows the Thor Beck to join the Wamphray Water, lies a small rectangular fort, incomplete and seem¬ ingly never finished. It is distant about J mile north from Laverhay and lies at an elevation of some 700 feet above sea-level. It is on falling ground and is entirely over¬ looked from the higher ground to the east, there being a fall across the interior area of i in 7. The enclosure, measuring some 109 feet from north to south by 106 feet from east to west, faces the glen of the Thor Beck on the north, from which it is distant some 50 feet at the north-east angle and 20 feet at the north-west, and has been surrounded, except as after noted, by a flat-bottomed trench, 24 feet wide from crest to crest, 3 feet deep ^t most below scarp and counterscarp, with a mound on both sides, that on the scarp some 3 feet high above the interior and that on the counterscarp 18 inches to 2 feet high. On the east the outer mound, above a rather long counterscarp, is evidently secondary, as it crosses the trench at the north-east corner and similarly traverses it at the north-west angle ; it has probably been a feal dyke, connected with a number of others crossing the slope to the south and passing up and down the hill. The trench does not appear to have been fully dug out on the east side, and on the west the defences are only con¬ tinued for a distance of 40 feet from the north¬ west angle, beyond which the ground maintains its natural gradient and condition. XXV. N.W. (“Earthwork”). 27 September 1912. Enclosures. 623. Enclosure, Alais Knowe, Elbeckhill.— This enclosure is situated on the south-west side of the moorland road which leads from Berryscaur to Wamphray, at an elevation of some 700 feet above sea-level and about | mile north by east of Hazelbank farm. It occupies the summit of a knoll forming the south end of a grassy ridge and is suboval in plan, the north end being considerably flat¬ tened in its curve. The main axis is north by west and south by east, and the interior dimensions are 260 feet by 270 feet or thereby. The knoll has an altitude of some 15 feet above a hollow on the west, tails away gradu¬ ally from the base of an 8-feet scarp on the south, and drops for some 30 feet to the bed of a burn on the east. Only at the north end, where the fort is faced by the continua¬ tion of the ridge, do its defences remain, and these consist of a rampart which slopes away gradually from its crest on the interior, with a fall of some 20 feet to the south end, and which in front forms a scarp, 6 feet in height, to a trench, 28 feet to 33 feet in width, cut through rock, with a counterscarp some 4 feet in height. Around the rest of the fort no rampart sur¬ mounts the slopes, and at the south end only for a short distance towards the south-west is a trench cut through opposing ground. The outline of the south end is very ii'regular. In the interior there are the foundations of three rectangular enclosures with banks of earth and stone, evidently sheepfolds and probably secondary, while along the east side are less definite remains of smaller enclosures. The entrance is in the centre of the south end and opens into an artificial-looking hollow some 30 feet in length by 9 feet in breadth. There is a considerable view from the site over the upper reaches of Annandale. xxxiv. N.W. (“ Fort ”). 12 August 1912. 624. Enclosures, Leithenhall. — Somewhat less than | mile north by east of the farm of Leithenhall, and on the southern end of a ridge looking down Annandale, are the remains of a circular enclosure which has suffered greatly from the action of the plough. It appears to have been surrounded by a single rampart, now merely a slight swelling on the surface, and has been hollowed by 211 WAMPHRAY.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [wamphray. excavation in the interior to a depth of some 4 feet to 5 feet on the north or higher side and about 2 feet on the lower. It appears to have been approximately circular, measuring interiorly 170 feet b}^ 159 feet. About I: mile due north of the last, and now entirely planted over with trees, are the remains of a similar enclosure, which has like- vise been considerably dug out in the interior. Though not now measurable, it is shown on the 25-inch O.S. map as an oval, surrounded by a single rampart and measuring 150 by 112 feet. XXV. S.W. (“ Forts ”). ii September 1912. Stone Circle. 625. Stone Circle, Kirkhill.—On the moor, in an angle between two feeders of a burn and i mile east-south-east of Kirkhill, are the remains of a stone circle, consisting of seven large blocks of whinstone not more than 18 inches above ground, of which pro¬ bably only two are now in their original up¬ right position. The stones are not large, the average length being about 3 feet 6 inches. The diameter of the circle has been approxi¬ mately 50 feet, and a number of the stones have evidently been removed. XXV. S.W. 12 August 1912. Miscellaneous. 626. Earthwork, Laverhay. — This earth¬ work, on the moorland about J mile to the north of Laverhay, seems to be the turf dyke of an old enclosure of some considerable extent and not defensive. It is significant that within it the ground is devoid of heather, which luxuriates all round. Along the north side there is an outer bank running down from the north-east corner, some 15 feet out, rather slighter. Probably it has formed, with the inner mound, a bught or pen. XXV. N.W. II September 1912. 627. Standing Stone, Skipknowe.—In a field on the east side of the road from Lockerbie to Wamphray, and | mile south of Wamphray station, stands a massive whinstone boulder, 5 feet 7 inches in height above ground, oval in section, measuring in circumference at base 12 feet 4 inches and set with its longest axis east and west. At 18 feet distant another boulder protrudes from the bank at the road¬ side ; but, without removal of the soil heaped about it, it is not possible to conclude whether it has any relation to that in the field. Should it, however, be found to be deeply sunk and firmly embedded in the ground, it is possible that these are the remains of a stone circle. xxxiii. N.E. 14 August 1912. 628. Sculptured Stone, Wamphray Church. —Built into the wall of the church, above the doorway, is a panel 4 feet 2 inches in length, divided into two oblong compartments, each surrounded by a rope-mould¬ ing. The smaller of the two com¬ partments, to¬ wards the north, contains a cir¬ cular foliaceous design with fleur-de-lis pro¬ jecting from it into the angles, while the larger compart ment contains a dra- gonesque crea¬ ture whose body breaks away in numerous s p i - rals. (This stone is illustrated and described in the Early Christ. Mon.) A slab set against the east sideof arailed-in burial enclosure belonging to a family of the Fig. 149.—Sculptured Stone, name of Car- Wamphray Church (No. 628). ruthers, some 50 feet south of the east end of the church, bears on the front, in the centre, a shield charged with a saltire, in chief two cushions tasselled, in base a heart (for Johnstone), beneath, the date 1697. Along one side and the top is incised the inscription :— 212 WAMPHRAY.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [westerkirk. EVINE THIS OUR ANCHENTS HAVE US TOLD THAT VE OUR BURIAL PLACE SHOULD HOLD IN ME. And on the back of the stone, with the first word mutilated, is the further inscription :— ALL YE CHRISTIANS WHO DOE LIVE/AND ONE THE EARTHE IN SIN/DOE MOVE BEG PARDEN FROM EVERY ONE SO LONG AS TIME MAY NOT BE GONE/THE WORD OF GOD HATH BENE REVEALD OF OLD/tO PROPHETS AND THAY HAVE US TOLD/NOW TIME IS GONE AND SHALL NO MOR/ETERNITY IS THEE BEFOR WH. 4T MAN IS THAT LIVETHE HEAR AND UNTO DEATH WIL NOT COMPEAR. XXV. S.W. II September 1912. 629. Standing-Stone, Newbigging.—On the east side of the road from Wamphray Church to Moffat, between 600 yards and 700 yards to the north of the farm of Newbigging, there lies a fallen standing-stone of the native whin, 6 feet 3 inches in length, 2 feet 8 inches in breadth at base, and tapering to a point. xxiv. N.E. (“ Chapman’s Stone ”). ii Sep¬ tember 1912. 630. Standing-Stone, Poldean.—In a low- lying meadow on the west side of the road, and some 200 yards south of Poldean farm, is a standing-stone of whin, 4 feet 7 inches in height above ground, and triangular in section. This stone is said to have been erected to mark the spot where Prince Charlie’s troops bivouacked on their march into England in the ’45. xxiv. N.E. II September 1912. 631. Carved Armorial Panel, Poldean.—Built into the front of the farm-house of Poldean, in Wamphray parish, beneath a bow window on the south side of the doorway, is a panel, measuring i foot 9 inches by i foot 5 inches, on which are carved tw’o ornamental shields bearing arms :—dexter shield A saltire, a chief charged with three cushions, a heart in base; over the shield the initials A I ; sinister shield Three piles, three mullets in chief, a heart in base; above, the letteis N.D. The dexter shield and initials appear to be those of Ambrose Johnstone, wiio owmed Poldean (Powdene) in the reign of Charles II. (Pater¬ son’s Wamphray, p. 47). The other set would represent his wife, a Douglas. The stone is said to have come from a former house of Poldean. xxiv. N.E. 27 September 1912. Site. The O.S. map indicates a site as under :— 632. Chapel, Chapel Lea, Wamphray Water. XXV. N.W. WESTERKIRK. Defensive Constructions. 633. Fort, Bankburnfoot. — This fort is situated at the very extremity of the south slope down from the Shaw Rig, close to the left bank of the Bankhead Burn and some 150 yards from its confluence with the River Esk. It is almost circular, measuring interiorly 182 feet from north to south by 184 feet from east to w'est, and is surrounded by a bold ram¬ part, most prominent on the north towards the higher ground and declining southward, where, how^ever, a natural defence is furnished by the steep scarp of a bank falling to the haugh-land below. On the interior the rampart on the north rises to an elevation of about 9 feet and on the east to 8 feet ; it has been formed of earth and stone ; and along its crest, which is flat and narrow, there is a suggestion of stone- w'ork. On the w'est flank a trench has been excavated, not traceable around the north, some 26 feet in width from crest to crest, 10 feet in depth from the top of the rampart, and 5 feet from the counterscarp ; and along the east side runs a trench-like hollow, possibly natural. Some 30 feet out from the counter¬ scarp of the trench, on the w'est side, is an outer trench, some 25 feet wide and 6 feet deep, w'hich is carried on to the edge of the burn. The double trench on this flank ap¬ pears to have been intended to counter any attack on the fort by means of the hollow' dow'n which flows the burn, some 200 feet away. The entrance has apparently been from the south-w'est into a slight circular depression, which is faced with a distinct scarp passing segmentally from west to south and sw’ells into a low stony mound wiiere it merges 213 WESTERKIRK.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [westerkirk. into the rampart on the west. Within the hollow is a water hole, possibly a well. In the \'icinity of the entrance the original ram¬ part of the fort has probably been interfered with, and against the low bank which now represents it rests a triangular enclosure, seemingly of secondary construction; while on the east side, where the rampart also appears to have been reduced, there is another small enclosure, also secondary, but ap¬ parently set within an artificial hollow of earlier formation, which is visible on the north side of it. Enxlosures. —On slightly higher ground, about 100 yards north-east of the fort, is a small oval turf enclosure, lying with its longest axis north-north-west to south-south-east and broadening somewhat to the latter direction. It measures interiorly some 43 feet by 31 feet. The rampait has a breadth of 7 feet and a height of about 2 feet. There is no definite break in the periphery to indicate an entrance. Immediately in front of the rampart, at the north side of the fort, is another similar en¬ closure of smaller dimensions. The entrance is not well defined but appears to have been from the south-east. XXXV. N.E. I July 1912. 634. Fort, Bankhead Hill.—This fort lies on the eastern flank of Bankhead Hill, at an eleva¬ tion of some 700 feet above sea-level, and 200 feet above the fort last described (No. 633), which lies about J mile to the south-east. It is pear-shaped in form with its main axis north and south across the slope of the ground, which is from west to east, and measures interiorly some 245 by 235 feet. Around the enceinte runs a bold rampart of earth and stone with a suggestion of stonework on its sharp narrow crest. In front of it is a trench, varying in width from 24 feet on the east to 40 feet on the south, which in places is cut through rock. It is sharply cut to a V-shape on the west, but elsewhere it is now flat bottomed. On the west and south the scarp rises to a height of 7 feet or 8 feet, while the counter¬ scarp varies on these sides from 4 feet to 6 feet; elsewhere both sides are lower. Except on the north, where the foreground is flat and wet, there is an outer mound above the counter¬ scarp. The main entrance has been in the north¬ east angle, 8 feet wide, opening into a circular hollow, around which the ground rises to a height of from 4 to 6 feet. There is a break in the periphery towards the south¬ east ; but on the adjacent sides of the rampart much material has been thrown up, evidently from the widening or formation of the gap, the original character of which is thus open to doubt. The interior is not level, and the rock, which is near the surface in many places, outcrops. XXXV. N.E. I July 1912. 635. Fort, Shiel Burn.—This fort, which is an earthwork, is situated at the edge of the steep right bank of the Shiel Burn flowing 30 feet below and some 100 yards above the road bridge. Its elevation above sea-level is about 450 feet. In form it is a semicircle, with its chord, somewhat irregular in line, formed by the edge of the bank, and measures 240 feet by 150 feet. It is enclosed by a massive rampart, rising at highest 7 to 8 feet on the inner side and dropping some 5 feet to a trench in front, now very shallow, but measuring 30 feet in width from crest to crest. The rampart, with a flat top some 4 feet 6 inches broad, has been sur¬ mounted by a stony parapet; the large stones which have formed the outer margin of its base are visible on the south-east arc, and a regularly laid foundation of cobbles is ex¬ posed where a section is broken above the bank. The entrance has been from the south-west, and has passed through an opening in the trench, the ends of which have been closed on either side. In the interior it gives on a hollow way, flanked for some 60 feet on the left by a slight bank, which appears to be stony where it merges with the rampart. Except in rear of the north-west arc of the rampart, the interior is very wet. XXXV. N.E. I July 1912. 636. Fort, Shiel Burn.—This fort is situated on declining ground on the south-east extremity of Shiel Hill, some 500 feet above sea-level and about J mile from where the burn de¬ bouches on the Esk. It is only about 200 yards distant from the last described (No. 214 Aih iiiii/ ij/td flis/orical Monuments — Dumfries. Fig. 150.—Fort, “ Bogle Walls ” (No. 638). Fig. 151.—Cist, “ King .Schaw's Grave ” (No. 648). To face p. 215 . WESTERKIRK.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [westerkirk. 635), which lies in full view less than 100 feet lower down the hillside. It is four-sided, but not truly rectilinear ; only one side, the east, is approximately straight, and the angles are rounded. It lies four square to the cardinal points of the compass and measures 160 feet by 167 feet. A massive rampart surrounds it, rising as much as g feet on the upper or west side above the ground immediately in rear of it, and imperceptibly blends into the eastward declination of the interior. In front of the rampart is a trench, which measures from crest to crest from 38 to 40 feet in width, with a scarp varying in height from 6 to 10 feet and a counterscarp from 4 to 7 feet. The trench is lost at the north-east angle, where the ground falls sharply away, but reappears to the south of the entrance, which occurs towards the south end of the east side. At its termination the counterscarp makes a return to the rampart. The entrance is some 10 feet wide and opens on to the lowest level of the interior area, with higher ground encircling it. The interior surface is very uneven, numerous hummocks of rock obtruding over it. XXXV. N.E. I July igi 2 . 637. Fort, Craig Hill.—This fort is situated on a northern spur of the Craig Hill, overlook¬ ing Burntoot, at an elevation of 800 feet over sea-level. The hill projects far into the dale on the north, forcing the Esk at this point to make a considerable detour, and the fort consequently commands a wide prospect up and down the valley. The enceinte is oval in form and is encircled by a stone wall of dry masonry, now a structureless mass of debris, about 2 feet high and measuring from 12 to 14 feet across, but there are indications pointing to the original breadth having been about 9 feet. It lies with its longest axis north and south and measures interiorly 142 feet by in feet. The entrance, some 4 to 5 feet wide, has been through the centre of the west side. Around the south arc a rampart of earth and stone curves round from in front of the entrance on its south side towards the west, at neither end now visibly connected with the wall of enceinte, and con¬ tains a somewhat level area measuring about 120 feet across near the centre. In the interior, against the wall to the south, a mass of stones suggests the ruins of one or more huts ; and within a trench-like hollow, beyond the outer rampart on the south, is a hut-circle measuring 13 feet by 10 feet, with depressions of a similar character, one on either side, measuring some 12 feet in diameter. To the southward the hill rises upward for over 200 feet to its summit, but in the other directions the ground falls at a steep gradient to the bottom of the valley. xlv. N.W. 10 July 1912. 638. Fort,“BogleWalls. ”—This fort(fig. 150) is situated at the roadside about J mile east by south of Enzieholm, on an angular spit of land formed by the debouchement of the glen of a small burn on the west on to the high bank which bounds the bed of the valley of the Esk, and some 70 yards back from the stream. The glen on the west has an average depth of some 15 feet, and the bank towards the river an altitude of about 30 feet. By constructing from bank to bank, on a quadrant of a circle, a massive rampart with a deep V-shaped ditch to the outside and a slight mound on the counterscarp, a triangular enceinte has been formed which measures along the straight sides 120 feet and 144 feet and bisectionally 132 feet. The rampart towards the interior rises to a height of from 7 feet to 8 feet ; it appears to be formed of earth and has evi¬ dently been surmounted by a stone parapet 3 feet broad, in rear of which, towards the east end, the top of the rampart forms a level platform, 6 feet wide, some 18 inches to 2 feet below the present crest, and 4 feet above the level of the interior. The trench is 42 feet wide from crest to crest and has a depth of from ii feet to 12 feet below the scarp and of 9 feet below the counterscarp : the low mound above the latter has a breadth of 12 feet at base and a height on the exterior of about 3 feet. Along the north-east side, facing the Esk, the ram¬ part appears to have been returned for some 40 feet, and a hollow seems to mark a place where it has been dug up ; beyond this, to the apex of the triangle, there has been a stone wall some 5 feet thick. On the very point there are two depressions, which probably mark the sites of huts. Along the flank above the burn there are also indications of walling. 215 WESTERKIRK.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [westerkirk. The entrance has been from the bottom of the glen past the west end of the rampart, and to right and left of it in the interior are small excavated hollows, which may be the sites of guard huts. A wet hole in the centre, where lie some detached masses of rock, is probably a secondary excavation. XXXV. S.E. 10 July 1912. 639. Fort, ‘ ‘ White Birren, ’ ’ Lyneholm Hill.— The remains of this fort are situated on the top of Lyneholm Hill, at an elevation of about 800 feet over sea-level and some 350 feet above the Esk, which flows by its base on the north. The construction has been greatly pillaged for stones wherewith to build dykes, and a large segment of the rampart on the north¬ east has been completely eradicated. The enceinte appears to have been oval to oblong in form, surrounded by a stony rampart, and, where the ground level rises to the southward, by a trench beyond. The rampart has a breadth at base of 14 feet and appears to have been surmounted by a wall ; the trench, where best preserved, shows a width of 27 feet from crest to crest and, at most, has a depth of 6 feet below the crest of the rather irregular scarp and 4 feet below that of the counter¬ scarp. At one point it has been cut through rock. The entrance has been from the south¬ east and is situated near the south end of the east side (a gap in the rampart on the opposite side, near the south end, being secondary) ; it opens into the lowest area of the interior and is faced with higher and rocky ground in front. There has been a certain amount of original excavation in the interior, where the rock lies very near the surface, probably to obtain material for the rampart and wall. The situation is a strong one and commands an extensive view down the dale. XXXV. S.E. 10 July 1912. 640. Fort, Camp Hill, Bailiehill.—This fort ranks, of the native forts in Eskdale, next in size and importance to Castle O’er, which is situated little more than i| miles away from it to the northward. It is situated on a prominent rocky hillock, which rises to a height of over 700 feet above sea-level, at a point where the White and Black Esks mingle their waters to flow onward to the Solway as the River Esk. The fort looks straight up the White Esk valley and also down the dale of the main river. The hillock rises steeply on the west for some 200 feet in elevation, above the ravine of a small burn which has its source in the Mid-Height, and it has a similar altitude, attained at a somewhat easier gradient, from the Esk on the north. To the southward the slope away to the moorland is slight, and from the east also the inclination is not great. The fort has its main axis from north-east to south¬ west, being controlled by the actual summit of the hill, and measures in extreme length over its outer defences some 550 feet. Its original breadth is not obtainable, as the culti¬ vated land on the east has encroached on the periphery, so that in that direction the outer defences have been eliminated. The central area is an oval enclosure measur¬ ing some 160 feet by 140 feet, which is sur¬ rounded by a stony rampart where the ground is low to the outside, and, where the level rises by a steep gradient towards it on the northern half, by a rampart slighter in degree, with a considerable excavation forming a scarp on the interior. Where best preserved, this rampart has a breadth at base of 19 feet and a height of about 5 feet. The interior level rises towards a rocky eminence on the west, along the outer edge of which the ram¬ part is continued. There has been a wide entrance into this enceinte from the south— the gap now measuring 20 feet across—open¬ ing on to the lowest part of the interior area. The top of the rocky eminence has been hollowed out to a depth of from 4 to 5 feet, so that its summit is a concavity which measures some 56 feet by 46 feet in diameter, open towards the lower level of the interior on the south. Within this hollow are several ridges on the turf, which mark the founda¬ tions of structures and were produced pro¬ bably by the cutting away of the rock on either side, as at Castle O’er. Enclosing this inner enCeinte is an outer line of defence, con¬ sisting of a deep V-shaped trench, the scarps of which are very sharply cut and smooth, varying from 17 feet to 26 feet in width from crest to crest and from 6 feet to 8 feet in depth, with a mound above both scarp and counter¬ scarp, some 20 feet broad at base, which is formed of earth and stone. This vallum on the 216 WESTERKIRK.] INVENTORY OF MONUMENTS IN DUMFRIESSHIRE, [westerkirk. north-west passes immediately below the rocky eminence, and extends out at either end some 6o feet and 130 feet beyond the rampart of the inner enceinte. There is an opening through it, 9 feet wide, on the south, which may have been an entrance, but it is doubtful. XXXV. S.W. 10 July 1912. Enclosures. 641. Enclosure, Burnfoot.—At the top of the high bank overlooking Burnfoot House from the north, and 150 yards to the eastward of the glen of the burn, are the remains of a circular defensive construction, which has been surrounded by a broad rampart rising some 5 feet to 6 feet above the level of the interior. Much of the area is under crop, the periphery in places is destroyed, and a gardener’s cottage stands within it. The approximate diameter is 200 feet. xxxvi. S.W. (“Birren”). i July 1912. 642. Enclosure, Cuil Plantation, Westerhall. —This enclosure is situated just across the glen of the Rig Burn opposite the mill dam, and occupies an angular promontory formed by the meeting of the high west bank of the burn with a steep scarp facing the low- lying ground which intervenes between it and the left bank of the Esk. The construction is four-sided and is protected on three sides by a stony rampart, some 4 feet in height to the exterior and 5 feet to 6 feet above the ground level of the interior, which declines towards the south, and 17 feet broad at base. Along the fourth side, that overlooking the burn, there is now no artificial defence. The interior is very uneven and has in places probably been dug out ; but, the area being within a thick wood and much over¬ grown, artificial features are difficult to dis¬ cern. Along the inner side of the rampart, on the north, a plateau, some 30 to 40 feet broad, extends from the edge of the bank on the north-east for about half its length. The position of the entrance is obscure ; but it has probably been in the west angle, where there is a gap in the rampart opening into a hollow in the interior and carried forward past the end of the plateau. Near the centre of the enclosure there appear to be the founda¬ tions of a small rectangular structure, and over the general area odd heaps of stone ai’e discernible through the rank vegetation which covers it. xxxvi. S.W. (“ Fort ”). i July 1912. 643. Enclosure, Hizzie Birren, Westerhall.— This enclosure has been cut through by the formation of the north-west avenue to Wester¬ hall and is about | mile distant from the mansion-house. It has been an irregular ellipse in form, with its longest axis north-north-west to south-south-east, measuring 150 by 120 feet; and its west flank rests on the edge of a steep bank overlooking the Esk. It has been surrounded by a massive stony ram¬ part or possibly a wall, now considerably dilapidated and overgrown, which measures 10 to 12 feet in width at base and some 3 feet in height. The ground falls away to a hollow in the interior. There is a gap through the rampart at the north end, but it is doubtful if this is the original entrance. An arm pro¬ jects from the main rampart down a natural ridge towards the south-east, covering a natural hollow, which may possibly have been en¬ closed. xxxvi. S.W. (“ Fort ”). i July 1912. 644. Enclosure, Crooks.—This enclosure is situated on the east side of the Meggat Glen, somew'hat less than J mile to the north of Crooks farm, in the south angle formed by the junction of the Stennies Water with the Meggat, and is an irregular saucer-shaped hollow excavated at the foot of the north-west slope of the Fell Hill, above a bank some 20 feet in height, which falls steeply to the meadow- land below. It is circular, measuring some 132 feet in diameter, with a depth of as much as 5 feet at the lowest point below the adjacent level and is overlooked by the rapidly rising face of the hill on the east. Only on the north-east arc is there a mound above the interior scarp, some 18 inches in height on the exterior, while the interior level behind it drops to 4 feet below its crest. The interior is divided into two distinct levels ; a circular hollow some 66 feet in diameter at the edge of the bank on the west, into which the en¬ trance, some 4 feet wide, opens, and higher platforms around it on the east and north¬ east. XXXV. N.E. (“ Fort ”). i July 1912. 217 WESTERKIRK.] HISTORICAL MONUMENTS (SCOTLAND) COMMISSION, [westerkirk. 645. Enclosure, Boonies.—This enclosure is situated between the high road and the right bank ot the Esk, immediately to the west of Boonies and on the crest of a steep 30-foot bank, which drops abruptly to the narrow strip of meadow l3dng by the margin of the river to the northward. On the east the deep glen of a small burn, some 60 feet to 80 feet distant, forms a natural defence. To south and west the adjacent land is flat. The en¬ closure is almost circular in form, measures in diameter from crest to crest 126 feet by 124 feet, and is surrounded by a massive rampart some 6 feet in height above the ground on the exterior and from 6 feet to 7 feet 6 inches on the inner face, with a breadth at base of some 30 feet. The entrance is from the east, at the top of the bank near the river, 4 feet in breadth, and directly overlooks a hollow on the face of the bank carried to the higher level in front of the entrance. The rampart appears to be of earth, and, as there is no surrounding trench, it has probably been formed of material dug out of the interior, which is rather damp. XXXV. S.E. (“Fort ”). 10 July 1912. 646. Enclosure, Bentpath.—This enclosure is situated on the moorland, just outside the southern end of the fir wood which runs south for I mile from Bentpath. It lies just over the crest of the watershed, facing south and overlooking the glen ot the Boyken Burn, at an elevation of some 700 feet over sea- level. In form it is circular, measuring some 124 feet in diameter, and is surrounded by a bank of earth and stone 18 feet in breadth at base, with a trench to the outside, 18 feet wide and 3 feet 6 inches deep below scarp and counterscarp, where best preserved, which is on the upper or north-west side. On the east, adjoining the wood, the bank has been almost entirely removed. The entrance has been from the west, over the trench and through the bank, and does not, as is usual, open on the lowest part of the interior. Any details in the interior are obscured by a dense growth of bracken. XXXV. S.E. (“ Fort ”). 12 July 1912. 647. Enclosure, Cauldkinefoot.—This enclo¬ sure is situated on the north-east face of Shaw Hill below Cauldkine Rig, as it dips to the glen of the burn overlooking Cauldkinefoot, and at an elevation of some 850 feet over sea-level. It is oval in form, with its longest axis north¬ west and south-east, measures 240 feet by 183 feet, and is surrounded by a stony mound with a well-formed concentric trench beyond which has an outer mound above the counter¬ scarp on the lower side. As usual, the interior has been hollowed by excavation, so that on the upper side the base of the ramp on the inside lies at a level some 10 feet below that of the ground on the outside. The trench is deepest on the upper side, where the bottom lies some 6 feet below the crest of the scarp and 10 feet below that of the counterscarp. A quadrant of the interior has been formed into an inner enclosure by a stony mound curving from west to north. There are at least three entrances—a wide one into the smaller enclosure from the north ; another, 10 feet wide, into the main part of the interior from the lower or east side ; a narrowei one from the south ; and possibly a similarly narrow one from the west. xliv. N.E. 15 July 1912. Sepulchral Construction. 648. Cist, “King Schaw’s Grave.’’—Situ¬ ated on ground which rises slightly above the marshy area on the top of Bankhead Hill, known as Airswood Moss, is a short cist, com¬ plete except for the slab forming one end (fig. 151). It lies with its longest axis north-east and south-west, and is formed with two slabs on one side, a single slab on the other, a slab between them^ at the south-west end, and a cover. The length of the side slab is 4 feet ; the width between the sides, 2 feet ; the height exposed between the cover and the ground, 2 feet 6 inches ; the superficial area of the cover, 4 feet 3 inches by 2 feet 10 inches, and its thickness, 8 inches. This cist, which is known as “ King Schaw’s Grave,’’ formerly stood within a large circular cairn, traces of which are still observable around it and suggest a diameter of about 46 feet. It is said that the cairn was demolished in 1828, when 150 cart¬ loads of stone were removed from it to build the adjacent march dyke. XXXV. N.W. I July 1912. Site. The O.S. map indicates a site as under :— 649. Church, Kirktonhill. xxxvi. S.W. # Ant uni and d/is/orua/ Motiumiu! —Dumfries Photograph by Mr J. Montgomerie, Stair, Ayrshire. Fig. 152. — The Ruth well Cross as now set up in the Parish Church. 'J’oface p. 219. APPENDIX. REPORT ON THE RUTHWELL CROSS. With some references to that at Bewcastle in Cumberland, Character and Artistic Value of the Monuments. Within the church at Ruthwell is preserved the most important monument of its kind in Scotland, the Ruthwell Cross. This cross, and its sister monument, from which it cannot be properly separated, the cross or cross shaft at Bewcastle in Cumberland, represent the finest existing examples of those carved stone monuments of Early Christian or early mediaeval date, that are numerous in Scotland, Ireland, Wales, Cornwall, the Isle of Man, and some of the Anglian districts of England, while they occur also, though more sparingly, in the regions occupied by the Saxons. The two crosses must be nearly contem¬ porary, for, though there are notable differ¬ ences between them, the resemblances are so striking that no other conclusion is possible. They mutually interpret each other, so that it has been found advisable to include in the illustrations accompany¬ ing this Report, for purposes of comparison, a photograph of the Bewcastle shaft. See figs. 154 and 155. The beauty and interest of these ancient Northumbrian monuments explain the fact that a considerable literature has grown up around them, and that this literature is to a large extent controversial is due to the fact that one of the chief points of interest con¬ nected with the objects is the apparent impossibility of securing any consensus of opinion about their date. Antiquaries and philologists differ in their chronological attri¬ butions by five centuries, some assigning the works to the latter part of the seventh, others to the twelfth century, while various inter¬ mediate dates are also favoured. It follows too that there are the same divergencies of view as regards the provenance of the decora¬ tion and the styles or schools to which the figure work or the ornament is to be ascribed. In the case of the inscriptions which form an important element in the decoration the philologist must have his say about the linguistic forms employed, and the student of Latin epigraphy or of runes will discuss the chronology of this or that character or form of capital or minuscule. It is no part of the purpose of this Report to enter into controversial discussions. Its primary object is to describe the monument in all its details with fulness and accuracy, but it is believed that the Scottish public will expect something more than a mere colourless statement in each case of the bare facts. If a figure be described it is impossible to avoid employing attributes that indicate style and aesthetic quality, and in this way introducing con¬ siderations of school and country and period. A special form of letter or runic character is not properly brought before the reader by a mere notice or figure of its actual shape. It can only be intelligently understood when it is ‘ placed ’ in its historical or geographical position, and this necessarily involves some indication of its date. It may be said at the outset that the two crosses regarded artistically need not fear comparison with any European monuments of a similar kind belonging to the Early Christian or early mediaeval periods. This applies to both the figure sculpture and the ornament. There are qualities of proportion, grace, elasticity, expression, in some of the figures on the crosses that we look for in vain in European sculpture between the Early 219 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. Christian ivories or stone carvings of the Hellenistic East, and the French Gothic sculpture of the middle of the twelfth century. The vine scrolls with animals on the two crosses are admirable of their kind though parallels to them might be found, but the panels of pure foliage ornament on the Bew- castle shaft, in a style represented nowhere else in European sculpture, are of an aesthetic quality quite distinctive, and the same may be said of the treatment of the bird form at Ruthwell, where there is evinced a feeling for nature and an accent in the carving not to be found elsewhere save in good classical work or in that of the Gothic and Renaissance periods. The panels of interlacing work at Bewcastle are better executed than on any other of the innumerable British monuments on which this motive is in evidence. This estimate may seem a high one, but the illustra¬ tions shown will justify it, and it may be noticed that the questions connected with the crosses are not of merely local or even national interest. Continental scholars fully recognise the archaeological and artistic value of the monuments, and speculation as to their date and provenance occupies the attention of all professed students of the Early Christian and early mediaeval periods. It may be said here that the leading French authorities of the day agree in ascribing the style of the works to the seventh century, whereas some of the best-known experts in Germany have favoured a date in the tenth. Present Condition of the Two Crosses. The Bewcastle monument is still in its original position in the burying ground at¬ tached to the church of St. Cuthbert within the bounds of a Roman station at Shopford in Cumberland some ten miles south of the Scottish border. The shaft, from which the cross-head or other terminal, that was in a separate piece, has been broken away, is in its present truncated condition 14 feet 6 inches high above the ground and in width measures below i foot loj inches on the broader face by i foot 9 inches on the narrower, tapering to corresponding dimensions at the top of I foot 2 inches by i foot i inch. The cross section accordingly is not square but oblong {set postea, p. 246), and this is enough to show that it is a cross-shaft, and not an ‘ obelisk ’ as it is sometimes improperly called. The shaft is sunk about ii inches into a massive socket-stone. It is of local material, a hard, rather coarse-grained, grey sandstone, and a block of exactly the same stone, evidently intended for a fellow monu¬ ment, hewn away from its bed but never brought down to the lower ground, is still to be seen on the elevated ridge of the moor about five miles from the church.^ In contrast with this unbroken record the history of the Ruthwell monument is a chequered one. It now stands in an apsidal projection built out on the north side of Ruthwell church, where it was set up in 1887 and scheduled as under Government protection in accordance with the Ancient Monuments Act of 1882. Owing to its height it was found necessary to sink it below the floor of the building, and it is located, with plenty of room about it, in a sort of pit, some 4 feet deep, surrounded above with a balustrade and provided with steps by which it is possible to descend and examine closely the lower part of the shaft. From the floor of the church the spectator is well able to study the work, for the eye is about level with a point half way up the shaft. On the other hand the position of the stone makes it impossible to obtain complete photographs on a single plate of the whole of any one side, and the repre¬ sentations on fig. 155 are made up of a number of sectional photographs carefully pieced together. The shaft is bedded to a depth of 9 inches in the concrete at the bottom of the sinking, and rises to the height of 17 feet 4 inches above this level. As is usually the case with monuments of the kind, the shaft is of an oblong not a square section and it has two broad and two narrow faces, the former bearing the figure sculpture. On the broad faces there is a plinth at the bottom of the height of about 3 feet 8 inches, but it has been cut away on the eastern side so that the projection is now only in evidence on the west. The present breadth of the plinth is about 1 A suggestion has been made in print that the stone on the moor has been split away naturally from its bed. The even surfaces on the east face of the block and the west face of the bed with the presence of abundant pickmarks are proof positive of human manipulation. 220 REPORT ON THE RUTHWELL CROSS. 2 feet 4 inches. Above this the shaft is re¬ duced in width by curved sets-off, one being cut away, and it tapers from a breadth of about I foot 9 inches to one of about i foot I inch at the base of the cross-head. The narrow faces of the shaft taper from about I foot 5 or 6 inches at the ground level—the plinth does not project laterally—to about 9| inches at the base of the cross-head. The exact dimensions of the latter cannot be ascertained, but as now set up the height of the head is 3 feet i inch. This head was not, like the head at Bewcastle, in a separate piece, but was in the same piece as the upper part of the shaft. There was originally a join in the shaft lower down, so that the whole cross was in two portions. The division between these two original sections of the shaft comes above the panels with the hgure of Christ on the north and south faces. The cross is now in six pieces, and beside these six which are original there are also modern additions to make up portions that have been lost under conditions which the history of the monument will explain. History of the Ruthwell Cross. Prior to 1887 this history can be followed back in authentic records to the middle of the seventeenth century, before which time only tradition is available. Local tradition, placed on record as early as the beginning of the eighteenth century, affirmed that previous to the Reformation epoch it stood within the church. The present condition of the monu¬ ment lends to this credibility. When the surface of the stone is compared wdth that of the Bewcastle shaft it is seen that the latter, which has always stood in the open church¬ yard, is considerably weathered all over, whereas, though the Ruthwell stone is sadly bruised and scarred, those parts where the surface has not been exposed to definite injury are very well preserved—this applies specially to the incised runic lettering—and the inference is that for many centuries the monument stood under cover. Even after the actual Reformation the same protection was afforded to it, and it did not suffer from the iconoclastic zeal of the extreme party in the Scottish Church till nearly the middle of the seventeenth century. Dr. Duncan states that it was preserved from demolition till this period ‘ probably by the influence of the Murrays of Cockpool, the ancestors of the Earl of Mansfield, who were the chief pro¬ prietors as well as the patrons of the parish, and who had espoused the cause of the Episcopal party, in opposition to that of the Presbyterians Its time however came, for a couple of years after the passing by the General Assembly at Aberdeen of the Act of 1640 ‘ anent the demolishing of Idolatrous Monuments ’, official cognizance was taken of it by name. The following is from the Rev. James M'Earlan’s account. ‘ 1642. In \.\\Q Index of the Principal Acts of the Assembly holden at S. Andrews, 27 July, Not Printed, the Sixth Item is as follows ;—“ Act anent Idolatrous Monuments in Ruthwall ” ’, and he goes on ‘ In obedience to this Act it was probably thrown down during the ecclesi¬ astical troubles of Charles the First’s reign, in 1642 or 1644, during the ministry of the Rev. Gavin Young, who allowed its broken pieces to remain within the church, where they found protection for 130 years or more, in the old-fashioned earthen floor Here the fragments were seen and commented on before the end of the century by more than one observer,® and we possess two interesting notices from the pen of William Nicolson, the zealous ecclesiastical antiquary who in 1702 became Bishop of Carlisle. In a letter of May 24, 1697 he writes of ‘ a most ravishing Runic monument ... on a square stone- cross in Revel church ’, and in an earlier letter of April 22, 1679 notices with surprise the legibility of the inscriptions ‘ on all its four sides ’, ^ while in his Diary he gives an account of a further examination of the monument with a notice of the local tradition concerning its origin and earliest history. This is quoted later on. The cross had been broken when it was ^ ‘ An Account of the Remarkable ^lonument in the shape of a Cross, etc.’, in ArchcBologia Scotica, vol. IV, p. 317. The paper was read in 1832, when Dr. Duncan was minister of Ruthwell. - The Ruthwell Cross, J. INIaxwell & Son, Dumfries, 1896, p. 19. ^ Rev. Dr. Hewison, The Runic Roods of Ruthwell and Bewcastle, Glasgow, John Smith & Son, 1914. * Letters on various Subjects, etc., to and from William Nicolson, D.D., edited by John Nichols, London, 1808, pp. 63, 62. 221 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. thrown down, and the great block forming the undermost and larger portion of the shaft was in two pieces, of which the lower, about 6 feet 6 inches long, was too heavy to move without crowbars, while the upper, some 5 feet long, seems to have been more manage¬ able, for Nicolson obtained a transcript of the inscriptions on all four sides of it. On the heavy lower block were lying ‘ some little fragments’, and ‘some lesser pieces ... we found throun under Through-stones ^ in ye Church-yard.’ Besides the two portions of the fractured original lower section of the cross noticed here by Nicolson a third piece is mentioned by Alexander Gordon,- and in Pennant’s Tour in Scotland^ ‘some fragments of the capital ’ with carved eagles, clearly the present uppermost arm of the cross, are re¬ ferred to, but there is no direct mention of the cross-head as a whcJe, though had it been preserved it would certainly owing to its size and the conspicuous transom have attracted attention. An examination of the irregular fractures of the original upper and lower arms of the cross where they would join the transom shows that the whole cross-head including the transom was in one piece with the upper section of the shaft. It is probable accord¬ ingly that the transverse arm of the cross, in tire centre of which no doubt appeared some symbol of special sacredness, had attracted to itself iconoclastic hostility and was put out of existence after 1642, the top piece being fortunately spared. We turn with interest to the well-known Statistical Account of Scotland, ^ compiled by Sir John Sinclair towards the end of the eighteenth century. In the account of Ruth- well parish ® it is stated that the monument ‘ broken into two or three fragments ’ was in the churchyard, and Dr. Duncan ascribed this removal from the church to ‘ the increasing population, and the improved taste of the times ’ which ‘ rendered necessary better ac- 1 Table-tombstones supported above the ground by small pillars. 2 Itinerarium Septentrionale, Lond., 1726, pt. ii, p. 161. ‘ This obelisk some think was originally of one entire stone, but is now broke into three parts.’ This third part would be the original upper section of the monument, or, rather, a portion of this. 3 Lond., 1776, II, p. g6 f. ‘ Edinburgh, 1791-9- ‘ Vol. X, p. 220 (1794)* commodation to the congregation ’.^ In the churchyard which was not properly enclosed ‘ the prostrate column became more exposed to injury ’, ^ and he tells us that when he acquired the living ‘ he found it undergoing such rapid demolition, that he resolved to preserve it, by transferring it to a place of greater security. This resolution was carried into effect in the summer of 1802, when it was erected in a garden, which he had newly formed in the immediate neighbourhood of the church-yard ’. ^ Some twenty years later Dr. Duncan undertook a drastic work of restoration, though without working over or injuring in any way the actual original frag¬ ments, save only by a rather too lavish use of cement in the neighbourhood of the joins. Either then, or perhaps already in 1802, portions of new stone were fitted in at the upper part of the shaft to supply the place of sections of the original that were wanting, and the important addition was made in 1823 of a new transverse arm. The form of this, he says, was arrived at by comparison with the still surviving vertical arms of the cross, but he tells us nothing about the ornamenta¬ tion placed upon it. This is purely arbitrary and has no authority or archaeological value. The work was done ‘ by the aid of a country mason ’. The appearance of the monument as it stood in the ground attached to the Manse of Ruthwell, on a spot now marked by a tree surrounded with an iron fence, is well known from the photograph taken of it in a fortunate moment by Mr. J. Rutherford of Dumfries (fig. 153). Here it remained till the latter part of the nineteenth century, when (under the ministry of the Rev. James M‘Farlan, from whose pamphlet in 1884 the following is quoted) ‘ the Rev. Archibald M‘Ewen, M.A., Dum¬ fries, and the Rev. Mr. Lukis, Wath Rectory, Ripon, Fellows of the Royal Society of Anti¬ quaries, visit the Cross; and seeing how weather and climate are telling on it, urge its being taken again, and that speedily, into the shelter of the Kirk’. The result of this representation was the final installation of the monument in the apse within the Parish church, as already described on page 220. In the wall of the sinking in which it 1 Arch. Scot, iv., p. 318. ^ Ibid. ^ Ibid. 222 Ancient and Historical Monuments — Dumfries’. P/wtograpti hy Mr J. Rutherford^ Dumfries. I'lG. 153.—The Cross as it stood before its removal to the Parish Church in 18S7. To face p. REPORT ON THE RUTHWELL CROSS. stands is a brass plate with the following inscription :— THE RUTHWELL CROSS : DATES FROM ANGLO-SAXON TIMES : DE¬ STROYED DURING THE CONFLICTS WHICH FOLLOWED THE REFORMATION ; LAY IN THE EARTHEN FLOOR OF THIS CHURCH FROM 1642 to 1790 : ERECTED IN THE MANSE GARDEN in 1823 ; SHELTERED HERE AND DECLARED A MONUMENT UNDER THE ANCIENT MONU¬ MENTS ACT IN 1887. Material of the Cross, with Local Traditions of its Origin. It has been seen that the cross was origin¬ ally in two pieces, one forming the lower and larger portion of the shaft, the other the upper portion of the shaft and the head. The two pieces of stone differ in some degree in colour, the upper having a tinge of deep red owing to the presence of a large quantity of iron, the lower being of a warm grey. Both belong to the geological formation known as the New Red Sandstone, which occurs in the Niths- dale district, where Ruthwell is situated, as well as on the other side of the Solway in Cumberland. The question of the proven¬ ance of the stone or rather stones of the cross at once presents itself, and in this connection reference must be made to the traditions of the early history of the monument already noticed. An entry in the Diary of William Nicolson^ of the date July 5, 1704, contains the following :—‘ The common Tradition of y® Original of this stone is this : It was found, letter’d and entire, in a Stone-Quarry on this Shore ’—the entry refers to ‘ Revel ’—‘ (a good way within y® Sea-mark) called Rough- Scarr. Here it had lain long admir’d when (in a Dream) a neighbouring Labourer was directed to yoke four Heifers of a certain Widow y‘ lived near him ; and, where they stop’d with y'"^ Burden, there to slack his Team, erect y® Cross & build a Church over it: All which was done accordingly ’. Later in the eighteenth century the notice com¬ municated to the Old Statistical Account by * Printed in the Transactions of the Cumberland and Westmorland Antiquarian Society, New Series, vol. II, p. ig6. John Craig, who was minister of Ruthwell from 1783 to 1798, contains the additional information, ‘ Tradition says, that this obelisk, in remote times, was set up at a place called Priestwoodside near the sea, in order to assist the vulgar, by sensible images, to form some notions of religion, but was drawn from thence by a team of oxen belonging to a widow and placed in the churchyard, where it remained till the reformation . . .’, and Dr. Duncan in his Account of 1832 says that this tradition was still common at that time in the parish. It needs hardly to be said that in relation to monuments of this kind no genuine local tradition should be ignored. On the basis of what has been quoted there might be founded a plausible theory that the cross was origin¬ ally set up near the sea and used, as we know such crosses were actually employed, as a preaching station, from which fact might be explained the local name containing the word ‘ priest ’. The removal of the monu¬ ment to the church is also quite possible, though the manner of the transfer reminds us of the familiar legend about Durham. Its erection, as Craig reports, in the churchyard is more likely than its admission into the actual building, though as is noticed above the con¬ dition of the stone suggests that it had stood for a long time under cover, and the undoubted fact that the fragments of it were preserved within the sacred edihce is additional proof that it had stood there before it was thrown down and broken. On this same legend however a theory has been founded that the cross is not of local origin but was brought from over the sea and set up near the shore, and it has even been suggested that its place of origin was Cumber¬ land, where there stands at Gosforth the superb monolithic cross of sandstone which makes a third with the two at Bewcastle and Ruthwell. In connection with this the ques¬ tion at once arises whether or not the material of the Ruthwell cross is a product of a local quarry. Bewcastle is of local stone, and there is good reason to assume the same in the case of Gosforth, though the exact provenance of the rather peculiar but excessiv'ely hard and lasting stone of the Gosforth cross has never been ascertained. In regard to Ruthwell, though sandstones of the same kind might be found in Cumberland, yet there is every reason 223 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. to believe its material to be of local origin, and the idea that the monument is imported is not one to be taken seriously. There is now no actual sandstone quarry worked in the vicinity of Ruthwell, but there is plenty of the stone available, and it is used for the houses and walls of the locality. A single sound block the whole length of the monument and of a breadth at one end sufficient for the transom would not be easy to find save in a quarry of exceptional excellence, and it is not surprising that two pieces have been used. That these are not exactly alike does not prove that they were not both from the same site, for Mr. Postlethwaite, the well-known authority on Cumbrian geology, in a private communica¬ tion notices that in these sandstones ‘ local variations in hardness, etc., may occur in a few feet or yards ’. When Dr. Henry Duncan put the broken cross together he employed ‘ a country mason ’ who had to add some modern pieces to make up what was found to be missing. Local stone would naturally be employed, and it is evident that pains were taken to match in the new pieces the colours of the old, a redder stone being used in one part, a greyer in another, to suit the differ¬ ences in the original materials. Now a careful comparison made with the aid of a powerful lens between the original stones and these added pieces together wdth other fragments of sandstone picked up in the vicinity makes it quite clear that the original material was obtained from the local sandstones. The size of the grains of quartz, their sharpness and polish, which are very notable, and the amount of admixture of the particles of mica, are all practically the same, and place what has been said beyond any reasonable doubt. On this point, an important one, reference may be made to a paper ^ on the subject by Mr. James Barbour, F.S.A., of Dumfries, an experienced architect and an archaeologist of established repute. He notices that ‘ the popular account of the origin of the Ruthwell Cross derived from tradition affirms that on being conveyed by sea from some distant country it was shipwrecked at a place called Priestwoodside ’, and explains this theory of a foreign origin as ‘ a common way of account¬ ing for the presence of works without a history, and possessing merit superior apparently to any effort of local skill as this is ’, going on then to demonstrate the indigenous origin of the monument from the character of its material. A comparison of this with local stones with which he had dealt professionally convinced Mr. Barbour that they all belonged to the same formations, and after giving the results of a detailed analysis he sums up in the follow¬ ing words ;—‘ The facts stated are, I submit, fairly conclusive of the stone having been obtained from a local quarry ; and it follows that in all probability the Cross was sculptured and first set up in the vicinity where it stands ’. In connection with the last sentence refer¬ ence may be made to an unfinished cross that stands in the churchyard of Kells in Ireland. This monument is erect and fully shaped, but only a small portion of carved ornament has been executed upon it. Most of the panels where this was to come are left as projections on the face of the shaft and arms ready for the sculptor’s chisel, and this may be taken to show that work of the kind was executed locally and with the material of the place, though of course it does not follow that the executants did not come from a distance. The Principal Face of the Cross. The principal subject on the cross is the figure of the glorified Christ with His feet resting on the heads of beasts. The panel containing it is the largest of any on the shaft and is a third longer than the one next in importance, while it may be noticed that the Latin inscription on the margins, as is the case too with the other inscriptions on this face, is much better wrought than the similar lettering on the other side of the cross. On the Bewcastle shaft which has never been moved from its original site, and which is figured on its four faces on fig. 154, this same figure occupies the central position on the western face, and this is the natural arrangement, for the western face, say of a church or of a chancel arch, is always the decorated face because it is turned towards the worshipper whose looks are directed to the • eastern or altar end of the building. Hence in the case of the Ruthwell cross the face with the Christ in glory, though now as 1 In The Transactions and Journal of Proceedings of the Dumfriesshire and Galloway Natural History and Antiquarian Society, vol. XVI, Session 1899-1900. 224 Ancieni and Ilistoiical Monuments — Dumjries. West. Noith. I'.ast. Fig. 154. The Bewcastle Fross, shciwing the four sides. .South. ; 224. * iirl Aiiiu'iit ana //is/ortra/ Monnnitnts — Dumfries, South. East. North. West. Fig. 155.—The Ruthwell Cross, showing the four sides. To face p. 225, REPORT ON THE RUTHWELL CROSS. a fact directed northwards, was in all prob¬ ability the western face and as the principal one may be taken first in the description. As has been already explained the topmost piece of the cross is ancient, while the trans¬ verse arm is modern and of no archaeological importance. When the monument was finally pieced together in 1823 this top piece was turned the wrong way, and as this error introduces confusion into the representations it has been corrected in the photographs reproduced on fig. 155. The subject shown on the photograph at the summit of the (present) north side is the Evangelist St. John with his eagle treated in a somewhat remark¬ able fashion. Fig. 160 (i) shows it on a proper scale. The eagle is no mere attribute but is the principal element in the composition, occupying with the sweeping wing at its back the whole height of the panel. It is the Evangelist that is the attribute, and he is in comparison miserably cramped and awkward. He has a large head, but a torso of the most meagre proportions covered with a cloak thrown over both shoulders and open at the breast. The lower part of the figure may be best explained on the supposition that the Evangelist is seated and that he holds across his knees an open book or scroll the sharp edge of which, seen on the dexter side ^ in the photograph, is very clearly indicated. The left-hand claw of the bird appears to be held in the air a little above the book or scroll, and the space between it and the latter is filled in with a shapeless mass in which must be sought the hand or hands of the Evangelist. The probability is that the eagle is regarded by the artist as the source of the inspiration for what is to be written on the scroll, and that the action of the claw is that of dictation or indication. The margins of the panel bore an inscription in Latin language and charac¬ ter, which may be read, or filled out to, in PR iNciPio ERAT VERBUM, the first two letters of which are faintly and the last two clearly visible at the lower part of the dexter and sinister margins, and are sufficient to attest the former presence of the phrase. These words are evidently being dictated by the mystical ' The words ‘ dexter ’ and ‘ sinister ’ are used in their heraldic sense and mean ‘ right ’ and ‘ left ’ looking from the object, not from the spectator’s point of view. bird, whose importance in the representation is thus accounted for. Of the artistic render¬ ing of the creature a word must be said in view of the prominence of the bird form in the decoration of the two crosses as a whole. The other claw of the eagle comes down nearly to the ground, and when carefully examined in the original reveals a feeling for nature and a skill in carving nothing short of astonishing. The anatomy of the claw with the pad, out of which issues the actual talon, makes it clear that the carver knew and loved birds of the eagle or falcon kind, and we cannot fail to remember the special ex¬ cellence in the rendering of birds evinced by the designers of the early Anglo-Saxon sceat coins of the seventh century. The leg of the creature above the claw is cut away under¬ neath so as to stand quite free of the ground, a detail that is a proof of the carver’s artistic feeling and skill. The little finger can be passed in behind. These finer details are of course lost in the casts of the cross by which it is generally judged. The comparatively large panel below the head of the cross and above the Christ con¬ tains a figure of John the Baptist holding the Lamb. He is bearded and wears long hair and has a nimbus. The head of the Lamb is also nimbed and its body is treated with some elegance ; the right fore and hind legs, which are preserved, are gracefully drawn. The two left ones, now broken away, were cut quite free from the ground. The stone has been fractured across the middle of the Baptist’s figure and there is some filling in with cement, the result being that the arms cannot be clearly made out. A heavy cloak comes down over the shoulders and falls in thick folds over the lower part of the figure and was probably caught up over the right arm leaving the right hand, of which the fingers are visible, bare, while the left arm, also apparently enfolded in the cloak, sustains the Lamb, which seems to have been stepping upon it with the left foreleg and the two hind¬ legs, the right foreleg being raised. The two feet of the Baptist are generally described as supported on round globes, but others have seen in these supposed ‘ globes ’ merely a pair of wooden ‘ sabots ’ seen in front view. The ‘ sabot ’, as will be seen, is the footgear of Mary and Elizabeth in the ‘ Msitation ’, 15 225 HISTORICAL MONUIMENTS (SCOTLAND) COMMISSION. but in this case of the Baptist the old view is really the correct one, for on a close inspection the two feet are seen fully modelled outside and on the top of the two globe-like objects, and not in them, as would be the case were they ‘ sabots The probability is that there was some miscalculation about the height of the figure, and it will be seen that on the corresponding panel on the southern side of the cross the height of the two figures of the ‘ Msitation ’ has been similarly made up with a plain block of stone under their feet. Under the cloak the Baptist may wear a tunic, but there is no sign of this above and no very clear indication of it below, for the cloak, as is shown on the dexter side, comes down to the ankles, and may be his only garment. The whole of the margins of the panel were probably inscribed in Latin characters, but the only legible lettering is that on the lower part of the dexter border where (a)doramus is quite clear. The O is a square set diamond- vise. The upper two-thirds of this margin, the top margin, and the upper half of the sinister one bear faint traces of lettering. The lower half of this border is a modern restoration, two new stones running through the whole thickness of the shaft having been here inserted. On the bottom margin letters are to be made out but form no intelligible words. T X D X . . V can be read from the dexter end of this margin, but no interpretation has been suggested. The two St. Johns and the Christ in glory, which appear in this way on the side of the Ruthwell cross now under notice, are displayed also on the western or principal face of the Bewcastle shaft,^ and they are the only figure subjects shown on that cross, whereas at Ruthwell many other figure motives are employed in the decoration. This is one ground that has been adduced for assigning a slightly later date to the Bewcastle monu¬ ment, the artist of the latter having presum¬ ably selected three important subjects from the larger repertory offered at Ruthwell. Among these additional subjects on the monu¬ ment before us, the one on the base of the cross-head just above John the Baptist is the 1 It is assumed here that the very remarkable figure of the Falconer with his bird on the lower part of the Bewcastle cross is really a St. John the Evangel¬ ist and the eagle. See postea, p. 267. only one of which the interpretation presents difficulties. Two figures, seen about half length, stand side by side. Both have long hair and that on the dexter side appears to be winged, for behind both shoulders there is visible a projection that resembles those indi¬ cating wings in the case of the Angel Gabriel in the scene of the Annunciation on the present southern face of the cross. The figure’s right arm is bent and brought across the body, the index finger of the hand being extended as if in demonstration. The other figure is long¬ haired and beardless and holds with both hands a large book or tablet. The fingers of the right hand are clearly visible under the lower dexter corner of the object, while so far as can be made out, the left hand enveloped in a cloak, like the hand of the Christ with Mary Magdalene on the opposite side of the shaft, also supports the book or tablet. The margins, especially that on the sinister side, are so worn that no lettering could have survived assuming any to have been originally present. That the inscription would be in Latin char¬ acters and not in runes may be safely assumed, for with John the Evangelist brought to this side all the inscriptions on this face will be of the former kind. There is little doubt that the two figures represent the evangelist Matthew and an angel which as his attribute might suitably find its place beside him. The motive would corre¬ spond to that of the evangelist St. John with his eagle that occurs on the topmost arm of the cross, and we could safely postulate St. Luke and St. Mark, with their accompanying symbols, on the two ends of the transom, with an Agnus-Dei or some other recognized symbol of Christ in the centre. The devices intro¬ duced by Dr. Duncan on the modern stone he inserted as a transom have a semi-masonic appearance and as we have seen possess no authority. On this account in the photograph, fig. 155, the transom has been cut short on both sides. We now come to the principal figure on the cross, the Christ in Glory with His feet resting on the heads of two beasts that have the appearance of swine. Practically the same figure occurs on the Bewcastle cross, also with feet on similar beasts, but it may be noted that the nimbus of the Ruthwell figure is cruciferous while that at Bewcastle lacks this indication. 226 1 «t .-///<■/< nt and Historical Monuments—Dumfries Fig. 156.—The Christ, on the North face of the Cross. Tojacep. 227 REPORT ON THE RUTHWELL CROSS. That the cast of the drapery is different in the two figures is a remarkable testimony to the artistic endowment of the carver or carvers responsible for them, for it shows that there was nothing mechanical in the reproduction. There is another distinction in the beasts, the legs of which are differently disposed on the two monuments. At Ruthwell the two inner fore-paws cross between and below the heads, while at Bewcastle the outer paws appear on each side above the snouts. There is a possibility of doubt as to the creature repre¬ sented, because, while swine are cloven-hoofed, one of the beasts at Ruthwell has the distinct marks of four digits on the foot. The char¬ acterization is not very good. The so-called swine here take the place of the more familiar asp and basilisk, or young lion and dragon, as representing the spirit of evil, but the swine motive is either unique or at any rate so rare that it is unnoticed in any of the standard works on Christian antiquities. The same applies to the wolf motive, which has been suggested as the origin of the representa¬ tion. Coming to the figure itself, represented on fig. 156, we find ourselves in presence of one of the very best Christ efhgies in Early Christian or early mediaeval art. It combines dignity with grace after the fashion of the early Gothic figures in the French cathedrals, and it exhibits that simplicity and freshness which were then in evidence but were so soon lost in the advancing art of the later Middle Ages and the Renaissance. It possesses at the same time classical affinities, and if we compare it on the one side with French Gothic sculpture we may find on the other side resemblances between it and some of those late Hellenistic works of the Christian Orient to which Professor .Strzygowski and others have recently directed attention. See posiea, p. 252. The proportions of the figure are those of classical sculpture of the best period for it is about seven heads high, and the cast of drapery is of the Early Christian type, the form being covered with a tunic reaching to the ankles and an ample cloak or paenula. The figure of Eusebius in the Rabula Codex of 586 in the Laurentian Library at Florence ^ is draped in a similar manner to that of the Bewcastle Christ. The left hand on both 1 Garrucci, Storia dell’ Arte Crisiiana, III, Tav. 129. crosses holds a rolled-up scroll and the right is raised to bless or to demand attention. The type of head is the same in both cases, and even in the present weathered condition of the stone, though features are not to be made out there is a suggestion of beauty in the head which is especially to be noted in the Bew¬ castle example. It is impossible to detect any sign of a beard, which, if it existed, must have lain quite close, for there is no such break in the outline of the cheeks as a beard would have caused. On the other hand the Christ of the Ruthwell Cross possesses a well- marked moustache seen most conspiciiously as it lies over the right cheek running back towards the angle of the jaw. The photo¬ graph, fig. 156, shows it clearly, and the most careful examination with a magnifying glass serves only to confirm the impression that this is no accidental mark but a deliberately carved feature. It must be noted that the cogent piece of evidence for the moustache is not the appearance of a projecting mass of that shape on the lower part of the front of the face, for this might be accidental. If we follow how¬ ever the projection down to the background we see that it is there also in relief on the cheek, and this cannot be the result of wea¬ thering. It needs hardly to be said that the occurrence of the moustache on a head of the kind unaccompanied by a beard would be a distinctively non-classical and barbarian trait. In Gaul, Italy, and the Hellenistic East such a representation would be highly improbable. The top and the two side margins of the lofty panel which measures 3 feet 6| inches in height bear a Latin inscription somewhat curiously arranged, the good execution of which has already been noticed. It begins at the dexter side of the top margin with a cross followed by the name of Christ written in the form constantly used, IhS XPS, on which something is said below. This is followed down the sinister margin by the title judex • AEQViTATis • after each of which words comes a point. The sequence is broken off here, and the continuation is found on the dexter margin beginning at the top under the cross, and runs as follows : bestiae • et • dr.^coxes ■ COgXOUERVXT • IX ' DE to the bottom of the margin. A transference is then again made to the sinister side, and with the rest 227 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. of the last word (de)serto • the inscription proceeds from the full stop after aeqvitatis • down to the bottom of this margin, ending SALVATOREM • MUNDi •. The inscription, like all the others in Latin characters, is printed here in ordinary Roman lettering. The special shapes of the letters on the cross are given in fig. 162, p. 243. The occurrence here of divisions by means of points between the words will be noticed, the points being placed in the middle of the height of the line. In this as in all the Latin inscriptions on the two broad faces of the shaft the letters on both vertical margins are cut so as to be read from a position at the dexter side of the panel, while about the IhS XPS there is the following to be said. The six letters are abbreviations of the name of Our Lord as written in Greek capitals IH 20 YE XPI 2 T 02 . Mediaeval scribes in the West, inexpert in the differences between the Greek and Latin alphabets, took the Eta (H) for the Latin aspirate (H), and when they wanted to write it in minuscules figured it as given on the cross ‘ h.’ Similarly, the Greek Rho (P) was mistaken for the Latin P (P), and the Greek Chi (X) replaced by the Latin X which it resembles in form but not in phonetic value. To make IHS stand for ‘ Jesus Hominum Salvator ’ was quite an afterthought. The subject next below the Christ gives us the figures of the two early hermit saints of Egypt, Paul and Anthony, breaking between them a loaf of bread which has been brought to them in the desert by a raven, scs • PAULUS • preceded by a Latin cross appears on the upper margin, and the letters et • a which begin to descend the sinister margin can be readily extended to antonius eremi- TAE, though the whole margin is here broken away to the bottom, and with it a considerable portion of the right-hand figure, freger t • PAXEM IN DESERTO • runs down the dexter margin of the panel. Both figures have long hair. The breaking of the loaf is shown by a vertical groove in the middle betokening a fracture. The words on the margin below the feet of the two saints, preceded by a Latin cross, MARIA • ET 10 , all that are left of the inscription belonging to the next panel, pre¬ pare us for the subject of the representation, which is the Flight into Egypt. This must have been when perfect one of the best executed of the figure panels, and it is shown on fig. 157. The ass is well rendered and possesses a magnificent right ear. The left foreleg is broken away, and the condition of the surface of the right foreleg shown in relief against the background of the panel shows that the former was cut free of the ground. The figure of Mary is slender and graceful as she rides sideways upon the creature holding the child upon her knee. A portion of the nimbus, marked with a cross, against which the head of Christ was relieved, gives the scale of the figure which was of substantial size. The right hand of the mother is brought round and seems to be holding the hands of the child while her left arm gives support at the back. A veil falls from each side of her head down over the shoulders, and a rug or horsecloth on which she is seated is visible against the background underneath the body of the ass. In the upper dexter corner of the panel appears—not the head of Joseph —but the rounded top of a tree, a detail occur¬ ring in other representations of the subject. Below the Flight there was evidently another figure subject occupying the lowest panel of the (present) northern face. The theme of the Nativity has been suggested, and this would correspond aptly with the Crucifixion, which can be plainly made out on the lowest panel of the opposite, or southern, face. There are considerable traces of figures, apparently on rather a large scale, on this lowest northern panel, and one can imagine one sees on the dexter side the figure of the mother bending over the child while other figures are inclined in adoration in the same direction from the opposite side. It is probable however that no two observers would interpret in the same way the obscure indications on the worn and bruised surface of the stone. The Back of the Cross. Passing now to the (present) southern side, formerly, as we have seen to be most probable, turned towards the east and hence the back of the cross, we find on the topmost stone, the upper arm of the cross, what is perhaps from the artistic point of view the most interesting of all the representations on the monument. 228 Ancient and Historical Monuments — Dumfries. Fu;. i57.^The Flight into Fgypt, To far, (t. 228 REPORT ON THE RUTHWELL CROSS. It is shown on fig. i6o (2). This is the boldly designed figure of an eagle or other bird of the falcon tribe that is supporting itself by one claw which grasps firmly a conventionally treated bough that ends with a tuft of foliage. The other claw, the right-hand one, is gathered up under the creature’s body and it is turning its head furnished with a formidable beak towards the dexter side. The treatment is large and plastic, the position and action of the bird natural and effective, and we are retninded by it of the excellent birds represented with such spirit though on a minute scale upon the early Anglo-Saxon coins. On the curved side margins of the panel sundry characters of an inscription can be identified, but as these characters are runic they may be passed over here to be considered later on in connection with the runic in¬ scriptions in general on the two crosses. The upper margin is too much damaged for any characters to be made out upon it. The lower arm of the cross on this face is occupied with the half-figure of an Archer who has drawn his arrow to the head and is about to discharge it upwards and to his right. He has long'hair, and a short cloak hangs from his right shoulder. At the front of his body an object that is obviously a quiver hangs by a broad band. The meaning of the Archer and his relation to the eagle on the top panel are obscure, but it is quite possible that there is no recondite symbolism involved or at any rate present in the mind of the carver, and that the eagle, treated with such spirit and natural feeling, is to the artist nothing but a noble quarry, that the archer will presently transfix with his shaft. On the Bewcastle cross the Falconer, whomever he may represent, is treated in quite a secular spirit, and we are reminded of the varied and interesting designs on the sceat coins in which religious symbolism is often abandoned in favour of some piquant bit of secularity, so that Lelewel remarked that the Anglo-Saxon artist never quite di¬ vested himself of his native paganism. There are some traces of letters on the margins of this panel especially on the sinister side, but nothing can now be made out. Professor Victor thought the characters were runic, and this is to be expected, as upon the correspond¬ ing top arm of the cross the characters are undoubtedly of this kind. The eagle and the Archer are quite unconventional subjects, and do not fit into any of the recognized schemes or cycles, so that they give us no help towards determining what are the likely subjects for the back of the lost transom. These must be left uncertain. Next in order we have a remarkable re¬ presentation of the Visitation, though a portion from the middle part of the figures has been lost. The two female figures, draped in mantles drawn over the head and descending to the feet, probably with tunics under them, stand opposite each other and stretch out their arms to enfold each other in an embrace. Their feet are thrust into unmistakeable ‘ sabots ’ or wooden shoes, and below there comes a plain piece of stone, original and in the same piece with the feet, 4 inches deep, the explanation of which is not easy, but, as noticed above, some miscal¬ culation in the placing of the figures on the panel may account for it. Part of the lower half of the figures has been lost and the defect is made good with modern stones. The inscriptions on the side and top margins of this panel are tantalizing to the antiquary, for enough is left to stimulate the spirit of investigation, while their full elucidation is for ever impossible, since only about half of the inscribed margins is preserved. The characters are certainly runic, and as connected with a figure subject are exceptional on the cross, unless the figure of the Archer on the panel immediately above was flanked by runes. They represent an intrusion of the runic element on the domain of the Latin epigraphist that has some significance. The characters on the dexter margin are (exceptionally) placed so as to be read from a position at the opposite side of the panel. ‘ Martha ’ seems to be indi¬ cated on this dexter margin and on the top a word or words beginning with M, in which ‘ Mariam ’ has been read. This top margin, how¬ ever, ends with a clear R. It is quite possible that through the scriptural association of the names ‘ Martha ’ and ‘ IMary ’ the carver was led into the curious blunder of using the former name instead of Elizabeth. There are runic characters on the sinister margin, and some¬ thing will be said about them on a later page [posted, p. 241). The principal subject on this southern face, in a panel 2 feet 8 inches high, is a figure of 229 HISTORICAL :\IOXUMENTS (SCOTLAND) COMMISSION. Christ with Mary Magdalene at His feet, see fig. 15S. The Christ, a majestic figure, lacks something of the grace and finish of the Christ in Glory on the other side of the cross. His right hand is raised in allocution or in bene¬ diction and the left veiled in the cloak holds a book. The hair, falling in wavy tresses to the shoulders, is better preserved than in the other figure but the features are worn away. The nimbus is cruciferous. The figure wears a tunic and an ample cloak, draped quite differently from the cloak of the Christ on the northern face. The figure of the Magdalen is remarkable alike for boldness in conception and crudity in execution. She is only shown to the waist and at first sight looks as if she were kneeling, but the curved mass in front of the figure is really that of her hair, which with her right arm and hand she is pressing against the feet of Christ. The arm and hand are singularly awkward and the worst piece of figure work on the tw’o crosses. The hand is a monstrosity when we compare it with the claws of the eagles on the upper panels which leave nothing to be desired in the matter of truth and delicacy. The inscription, in Latin characters and in the language of the Vulgate, occupies all the four margins of the panel beginning at the dexter side of the uppermost margin, con¬ tinuing down that on the sinister side, then resuming at the top of the other side just under the beginning on the uppermost margin and descending this dexter margin, finally to finish along the bottom line. The lettering here is not nearly so regular or sharply cut as that on the northern face, and the characters are rather smaller and less well formed. There is no separation between the words either by spacing or, save in one case, by the use of a point. The letters that are legible are A strvm VXgVEXTI&STANSRETROSECUSPEDESEIUSLACRI MISCOEPITRIGAREPEDESEIUS-&CAPILLISCAPITISS uiTERGEBAT,^ and the characters are i|- inches high. The carver had made some miscalcu¬ lation in regard to his spacing and had not sufficient room to finish in the same capitals his last word tergebat. The stone is a ^ ATTULIT ALABASTRUM UNGUENTI ET STANS RETRO SECUS PEDES EJUS LACRIMIS' COEPIT RIGARE PEDES EJUS ET CAPILLIS CAPITIS SUI TERGEBAT. good deal worn here, but the finish is clearly what is commented on postea, p. 243. For the sake no doubt of economy of space there are no divisions indicated between the words, save one point after the word Eius coming before &capillis. Next below follows the subject of Christ healing the blind man treated with three- quarter length figures, with a plain panel underneath that seems intended for an in¬ scription. Both figures have long hair and ample cloaks and Christ wears the cruciferous nimbus.^ A careful inspection of the left cheek of the Christ shows that here too, as in the Christ of the principal face, there was a moustache, but no beard. On the Christ with the Magdalen, on the other hand, there is no trace of either feature. The blind man, who is bearded, stands rigidly upright and in profile, but Christ, seen almost in three- quarter view, leans a little forward in accord¬ ance with the action at the moment selected. The right arm is advanced and over it fall the heavy folds of the cloak. This hand, it has been suggested, held the rod with which the eyes of the blind man were being touched, while the left hand is brought round to the front underneath the other. This hand is dis¬ proportionately large but is boldly carved and the artist has evidently taken pains with it. See fig. 158. In the very numerous ex¬ amples of this subject in Early Christian sculp¬ ture Christ places the fingers of His right hand upon the eyes of the blind man and holds in His left a cross or a scroll. Here at Ruthwell the right arm in its present position would not be long enough for this, and the left hand is not holding any object. Hence the arrangement of the subject here is abnormal if not unique. At the part where these figures are cut off there was a fracture of the stone, and when the cross was put to¬ gether by Dr. Duncan some cement was used for the purpose of making up. By an error the folds of the drapery of the two figures have been prolonged in this material so as to encroach on the space of the plain panel, which seems to have originally ended in a 1 The cross is indicated here by two parallel incised lines, whereas there are three lines in the case of the nimbus of the two other figures of Christ on the monument. The nimbus of the Bewcastle Christ has no mark of the cross. 230 Ancient and Historical Monuments — Dumfries. West side. South face. Fig. 158.—Portions of the western and southern faces of the Ruth well Cross. To face p. 230 . Ain i, !!/ a>!ii Historical Moinaiu nts—Dum fries. Fig. 159.—The Annunciaiion. To face p. 231 REPORT ON THE RUTHWELL CROSS. straight edge under the figures without any border. This panel measures 12 inches in height by an average width of 14 inches, and is sunk | inch below the surface of the margins to right and left and underneath, which are used, the two former for parts of the inscription explanatory of the scene of healing, the last for the inscription referring to the figure subject next below. The sur¬ face of the panel is not uniform for it curves forward a little in the upper portion of it, and it is not dressed smooth but is covered with neat pick-marks. Now those parts of the stone where there are inscriptions, as on the margins round the various panels, were evidently dressed very smooth before the inscriptions were cut, and the panel with which we are concerned has certainly not been so prepared, for the surface is far from having the same smoothness. This fact and the curvature of the surface suggest that there may once have been an inscription here which has been carefully chiselled out and the surface dressed down with the pick. This was certainly not a part of the iconoclastic work of the seventeenth century, but it is quite possible that at the Reformation, though the monument itself w'as spared, a Latin in¬ scription offensive to Protestant eyes may have been judicially effaced. No trace of former lettering can however be discerned at present on the panel. The theory that the surface was prepared for an inscription never actually cut is unlikely in view of the finished character of the whole monument and the abundant and carefully cut lettering on other parts of it. The marginal inscription here is a good deal broken away. It begins at the top of the dexter vertical margin after a Latin cross, with the words et praeteriens • viDi after which the margin is broken. We resume at the top of the sinister margin A NATiBiTATE ET SA and then after another break b infirmitate. The words are not an exact quotation but are suggested by the opening verses of the gth chapter of St. John. The impression of grace and ease which we have already derived from the figure of Christ with His feet on the heads of beasts is in¬ tensified when we turn our attention to the next figure panel, which comes below the plain panel just noticed and is figured on fig. 159. The subject here is the Annuncia¬ tion shown by the two figures of the Angel and Mary. When this was perfect there can have been few representations of the theme in Christian art so poetic in its expressive simplicity. What lightness and elasticity in the celestial intruder, what grace in the slender form of Mary drawn a little back with head shyly dropped but poised so charmingly ! The gesture of her right arm with the hand upon her bosom suggests a slightly deprecating movement, that contrasts most effectively with the impulsive advance of the angel. Mary, who seems to be wearing a cloak as well as a tunic, is drawing a fold of the former round her with the left hand, the gesture carrying out the idea of modest retirement. The long tress of hair seen on the right side of her figure is a conspicuous feature. The contrast in the actions of the feet of the two figures is noteworthy. The left one of the angel is seen in profile and suggests movement, whereas Mary’s feet are apparently repre¬ sented as lacing the spectator in an attitude of rest. The composition has been well thought out and the whole gives us the im¬ pression of maturity. It is most unfortunate that this panel has suffered such injury, for with the Christ and the Flight into Egypt it evidently contained the best figure work on the monument. The inscription began at the dexter side of the upper margin of the panel and the words here faintly legible are ingressus axgel, but as the i is inches from the edge of the margin there is room before it for the et with which the Vulgate account of the scene begins. The side and the bottom margins are broken or worn away, but at the top of that on the dexter side are some fragmentary letters which may be read TECUM • BE. Now the passage in the Vulgate, Luke i. 28, runs et ingressus angelus ad EAM DIXIT : AVE, GRATIA PLENA ; DOMINUS TECUM ; BENEDiCTA TU IN MULIERIBUS. Mea¬ suring by the space occupied by the letters INGRESSUS it is found that the rest of the top and the sinister margin would accommodate the above words down to dominus, while there would be room on the opposite margin for the rest. This would leave the margin be¬ neath the feet of the figures free for an in¬ scription relating to the subject below, which is that of the Crucifixion. If the letters here were on the scale of m.aria on the north side 231 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. there would be room on this margin for the words CRUCIFIXERUXT EUM, words that are specially emphasized in the Hiberno-Saxon manuscripts. It is the lowest panel on this, the present southern, face of the Cross, that offers this subject of the Crucifixion. The stone here widened out on both sides to an extreme breadth of about 2 feet 8 inches, the lower dimension of the shaft proper being about I foot 9 inches, but on the present east side the extra width has been taken off. The height of this base or plinth is about 3 feet 8 inches. Full advantage of the space thus afforded was taken, and the Cross of Christ with the figure upon it is displayed on a comparatively large scale. The upright stem is about 7 inches wide, the cross arms about 4 inches wide, the figure nearly 3 feet high. Here again the damage has been disastrous, and of the head and arms of the figure little remains, though there is some indication that the former was bearded. The lower part is better preserved and enables us to see that the figure was nude save for a small piece of drapery apparently folded round the loins and knotted at the right-hand side below the waist, thus forming a short skirt or kilt that ends above the knees. The legs below are well drawn and are not crossed. What else there was upon the panel is not easy to say. The orb of one of the two chief heavenly bodies is apparent above the sinister arm of the cross, and the other may be assumed above the dexter arm. Below on both sides of the stem of the cross there were representations, and some have seen here the two crosses of the thieves, others the figures of soldiers, etc., but it is not possible to arrive at any decision. The Head of the Cross. This description of the two principal faces would not be complete without a notice of the head of the cross. The form of this is very noteworthy and it bears on the much-controverted question of the date of the monument. The transom, or horizontal arm of the cross, is to be entirely ignored. On the other hand the upper and the lower arms of the cross-head are original, and have preserved the primitive outline as well as sculptural representations of the highest interest. The carver who fitted the modern transom to the original lower arm misunderstood the form, and the junction is bungled, though at the upper side of the tran¬ som it is managed correctly. Here it will be seen a sweeping curve fills in the angle, while there is a second curve between this and the end of the arm. This double curve is well marked and absolutely original, for it is re¬ peated b}^ the inner edges of the margins, where no alteration would have been possible. The bottom of the cross-head, it will be ob¬ served, is a little wider than the top of the shaft, and the junction of the two is thus tact¬ fully emphasized. The projection has how¬ ever in parts been broken or cut away, and an attempt to do this is quite clearly seen in the photograph, fig. 155, on the western face where the stone has been brutally scored with a pick. There is no sign here or anywhere else that the round of a wheel-cross has been cut away, and it cannot be too clearly under¬ stood that for the theory of a wheel-cross-head which has been suggested there is no founda¬ tion at all. The form of head that is clearly shown by the original portions that have survived is that of the well-known cross from Rothbury in Northumberland preserved in the collection of the Society of Antiquaries at Newcastle-on-Tyne, shown No. 8 on fig. 163, and of other examples which will after¬ wards be noticed. It may be noted that the Bewcastle cross must have had a head of a different pattern. The shaft here is in parts continued in a plane surface for 7 or 8 inches above the top of the uppermost figured panels, but there is no sign of a curve or of a projection marking the beginning of the head, as there is on the Ruthwell cross, where moreover the figure work is carried up into the head instead of ceasing some inches before it begins. The Narrower Sides of the Cross ; THE Ornament. The figure subjects are confined to the two broader faces of the shaft, and on the narrower sides, now east and west, the ornamentation is composed of animals and foliage. Here, and on the east face of the Bewcastle cross which is one of the broader faces, we have what is on the whole, both for design and execution, the best examples in early medi- 232 Ancieul and Historical Moniinicnts — Dumfries. Fig. i6o (i). —The Uj'per Arm of the Cross-head. To face ., 252 Ancient and Historical Monnments — Dumfries. Fig. i6o (2).—The Upper Arm of the Cross-head. To face p. 232 . REPORT ON THE RUTHWELL CROSS. aeval sculpture of the foliage scroll with animals pecking at the fruit, a motive popular from the earliest Christian times down to the Gothic epoch. On the Bewcastle cross there is a single panel the whole height of the shaft, but at Ruthwell there are on each side two panels one on each of the two original sections of the shaft, the decoration being the same on both. Upon the subject of both animals and foliage there is something to be said. The latter is designed in the scroll form familiar in later classical art, with an un¬ dulating central stem giving off at intervals on this side and on that tendrils that curve round in a circular sweep and afford lodge¬ ment within their convolutions to the pecking animals. These last are not only the birds of the orthodox classical tradition but quad¬ rupeds and also fantastic creatures whose tails end sometimes in foliage. The squirrel makes his appearance, but there is no instance of the fox, which on the strength of the text ‘ the little foxes that spoil the grapes ’ is sometimes introduced in a design of this kind, as on a Sasanian silver vase in the British Museum. The statistics are as follows. At Bewcastle there are, counting from below, a complete quadruped, two fantastic mammals with only forelegs, two birds, and two squirrels. At Ruthwell the western face gives us, lower panel, from below, a bird with fantastic tail, a quadruped, evidently an otter, two birds, two fantastic mammals ; upper panel, a bird, and, possibly, a squirrel : on the eastern face the lowest remaining scroll no doubt contained an animal but it cannot now be made out ; next comes a quadruped, possibly again an otter, then two birds and lastly two partly fantastic mammals ; a partly preserved quad¬ ruped and a bird survive in the upper panel. The creatures turn their heads alternately to right and left. There is a charming variety in these animals, and their heads with the action of pecking or of biting are natural and spirited. On the west side where the plinth has its full projection, it can be seen that there was conventional foliage on the plinth also below the shoulder where it receded by a set-off for the shaft, but the exact scheme of the design is in this portion not easy to make out. From the level of the first of the animals however the stem and its offshoots are clearly defined. On the upper stone, above the original joint, the work is not so well preserved as on the lower. The same is true of the east side, though here the plinth has been cut away and all indications of work upon it are lost. The questions of the nature of the foliage, that is to say, of the particular plant that has been subjected to the process of conventionalization, and of the animal forms, will later on receive atten¬ tion. See postea, p. 250. The west side of fig. 158 gives a specimen of the work. The Narrower Sides of the Cross : THE Runic Inscriptions. The upright margins of the lower panels of the narrow sides and the horizontal margins bounding them above are covered with inscriptions in the runic characters. These characters are about 2 to 2\ inches high and where the surface of the stone has not received actual damage they are perfectly clear and legible. There were similar inscriptions on the margins of the upper panels but they are very imperfectly preserved. Owing to this the purport of what is written on these upper panels is quite obscure, but on the other hand the content of the inscriptions round the lower panels is clearly made out and is of the highest possible interest. The matter here is a por¬ tion of an Anglo-Saxon poem that occurs elsewhere in a more extended form, and the existence of this fuller version makes it possible to restore with practical certainty parts of the inscription of the cross that have been accidentally damaged. The inscription in its literary and epigraphic aspects receives full attention elsewhere in this report (see postea, p. 268 f.). It is noteworthy that the execution of the runic letters is exactly the same as that of the Roman characters on the other two faces of the Mone. In each case the lines of the letters are finished at each end with a kind of dot, not made with a drill but with a pick, and the cutting as regards depth and sharp¬ ness is similar, so that the same workman or workmen may have incised both the Latin majuscules and the runes. A further point is worthy of notice. In the matter both of the Latin and the runic inscriptions there are differences in each case in the work on opposite faces. The Latin lettering on the principal, 233 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. now the northern, face is larger, bolder, and better spaced than on the southern face. There are as a rule points bettveen the words, and the et is inscribed without contraction, while on the southern face words are run together udthout separation and the et is twice contracted. In the case of the runes there is also a difference. On the present eastern face the runic characters are smaller than on the western, so that in the space from the top of the lateral margins to the place where the stone has been fractured and re¬ paired with cement there are on the eastern face twenty-two lines of runes and on the western only eighteen. See the illustration, p. 269, where the spacing has been copied with as much accuracy as is practicable. It is curious that on the eastern face the lines of characters on the two margins correspond horizontally, but on the western face they are not exactly on the same levels. It should not be omitted that there are alterations in two of the characters on the present eastern and western faces of the cross. One was originally cut as an E, but was altered to a U; the other as a G, altered to OE. This shows that an error of a letter in a runic inscription would be quite possible. At Ruthwell it will be noticed that in the general scheme, the two broader faces with the figure sculpture bear the Latin inscrip¬ tions, while the margins on the narrow faces are devoted to the runes. The latter however overflow, as it were, on to the broader faces and occur on the present south side, formerly the east face or the back, on the top arm of the cross round the panel with the bird, and also on the margins of the panel of the Visita¬ tion below the cross-head, and perhaps on the ‘ Archer ’ panel. All the rest of the writing on these two principal faces is in the Latin language and characters. This overflow is an additional proof of the intimate connection between the two sets of inscriptions. On the other hand on the Bewcastle cross the only inscriptions are in the runic letters, and when we compare the Ruthwell runic inscriptions with those at Bewcastle we find that there are characteristic differences. The differences in the forms of the characters will be noticed later on, but some note may be taken here of distinction in technique of cutting and in arrangement. The incisions at Bewcastle are broader and, owing in great part to the wearing of the face, shallower than at Ruthwell, and the lines do not end with the same pronounced dot or round depression. The whole surface of the stone is, as already explained, far more weathered, and as a consequence the char¬ acters are not so easy to decipher, though it is quite a mistake to assume that they cannot on the whole be made out with a substantial degree of certainty. See postea, p. 241 f. They differ moreover from the Ruthwell letters in the variety they present in their sizes. At Ruthwell the letters are all about the same height, approximately a couple of inches, and most of them are cut in a situation somewhat awkwardly chosen, in successive short lines across the very narrow vertical strips of the side margins. At Bewcastle the arrangement is more studied. The characters in the first place vary in size according to their distance from the eye, some of those on the uppermost parts of the shaft measuring as much as 5 inches in height. The letters giving the title of Christ above His figure on the western face are nearly 3 inches high ; those in the long in¬ scription of nine lines beneath this figure which are nearly on the level of the eye measure about 2| to 3 inches. The spaces apportioned to the inscriptions are regularly schemed out and appear to carry further the idea of a special inscribed panel, which we find on a modest scale on the back of the Ruthv^ell shaft, under the figures of Christ and the blind man. The nine-line inscription at Bewcastle has a border on each side marked off with a line. At Bewcastle the corners of the shaft are neatly finished with roll mouldings, at Ruthwell they were left square. The Runic System of Writing. A fuller note upon the runic inscriptions is necessary in order to place the reader in a proper position to form a judgment upon the date and provenance of the monuments. The runic system of writing is purely Teutonic and is chiefly represented among the more northerly members of the Teutonic stock. It belongs to the Germanic peoples just as the Ogham system belongs to the Britannic Celts, and any runic inscriptions discovered in the Celtic districts of the British Isles are as certainly intrusive as Oghams are when 234 REPORT ON THE RUTHWELL CROSS. found in Anglo-Saxon regions, as was the case at Silchester. The fact that runic inscriptions figure so largely on the two crosses is enough to preclude any suggestion that they are of Irish origin, though it needs hardly to be said that there are interesting points of connection between Irish artistic work and that found in Anglo-Saxon England, of which due account has to be taken. Runes were not however indigenous in the North, but the characters were derived from the Greek or, to a lesser extent, from the Roman alphabet, and their introduction appears to have been due to the Goths of about 200 A.D. when they were living, in touch with classical lands, on the northern borders of the Black Sea. Some of the characters betray at the first glance their origin in letters of the Greek or Roman alphabet but of others the exact derivation is obscure, and the matter is complicated by the fact that it was not always the familiar classical capital letters that were used, but the minuscules such as we know in ordinary print, and also cursive forms of the letters in popular use for every day writing. The modifications which the characters underwent in being turned into runes were largely due to the fact that they were meant to be incised on wood. This material offers itself as a rule in the form of long narrow pieces, derived from the natural shape of the tree trunk and the bough, and the characters would run from left to right or right to left along the strip of material avail¬ able, so that the upright strokes would be cut at right angles to the grain of the wood and the horizontal strokes, like those of an F, would lie along it. Experience evidently showed that these horizontal strokes tended to be confused with the lines of the grain, and for this reason they were made to take an oblique direction. Hence the Latin F becomes the runic char¬ acter shown in column i in fig. i6i, the Greek or Latin T the form seen in column 17. The cross strokes of the H, column 9, also slant. The characters in an inscription are almost always the same height, and were so made at first in order to fill evenly the up¬ right space allowed by the breadth of the primitive wooden stave. Perhaps owing to this tradition, the vertical strokes are in runic writing specially emphasized, possibly because of the crisp pleasant feeling, which every boy with a knife in his hand has experienced, of the clean cut against the grain of the wood, so that the classical A has the dexter upright stroke made vertical and the other stroke and the bar are run obliquely so as to make the runic character seen in its simplest form in column 26. The custom of reversing so that left becomes right, or of turning letters upside down, was freely adopted, so that on the Franks Casket, line e, the ‘ s ’ of column 16 maj'^ face either way. The combination of two letters into one runic character after the manner of ligatures or diphthongs also makes its appearance, and such compounds are called ‘ bind-runes ’. Specimens will be found in the second column of fig. 161. Stops are sometimes used, but in quite arbitrary and casual fashion, and there is no division between words nor any recognition that the end of a line is a good place to terminate a word. Lastly it must be noted that for some unexplained reason the classical (and Semitic) order of the letters. A, B, C, etc., is not followed, but the first letter is F, which with the five following makes up the word ‘ futhorc ’, so that this term is applied to the whole set of letters in the same sense as our word ‘ Alphabet.’ The runic system of writing was introduced, as has been said, in southern Russia, but it was propagated not in the westward direction as would have been expected, but towards the north. The reason of this was the introduc¬ tion in the fourth century by Ulfilas of a Gothic alphabet closely following the classical, in which he wrote his famous vernacular version of the Bible. This superseded for these Dacian regions the runic futhorc and barred the way against its diffusion towards the west. To the north however the way was open and the runic system of writing, with twenty-foiir characters equal in number to those of the Greek alphabet, became established in the Teutonic regions near the Baltic at a date that may be fixed about the third century A.D. From this centre it was diffused into all the countries to which the various Teutonic tribes of the north betook themselves on their migrations. In Gaul in the sixth century we learn from a couple of illuminating lines in one of the poems of Venantius Fortunatus that runic writing on 235 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. •25 '52- X VjO >o -- :=£r X o to to c^ CN o >ecC Cv OO (N > OO Cvj > CN, cv >- <5 <^ >< tN cv Cv w ^ Zz_ ZZ-. ^ ^ -, ''£> CN? Cv < ^ >Z 2if_ 2yL . /< :V lO CM X X X X lxl: L3«L..' cv tO CV 1*1 ^ D - Ch cv U 2Z xz n -r-- 2) ^ : 03 CD ^ _-iN CZl 00 L. 4 _ -> '4“- ■" t=: SN 1 N V — 3- so X- (O CL LX S to Cv ^ 4: cv o z ^ -- -7^ zl: O', X 3: :z: zzn lyR ; -X- — C?1 03 /Cs CO (\ O X X X X X K O — C y - V sO 'O X fX ^r=: Z=1 X yV .C'" li.ir^isr-T eZ lOZ lO 'f- o ^ yi/ » f7~ -// to X ^ -Ql. . /~\ £ n to H Of D ci; cn yzri;: yz: Cv L. ? i S“ H vi 0 g 2 Q* < u s || zi 3 to UJ Z ^ -i o 3 < < 2 S f ^ SI O '<) 'o g '7'“ O 1 S 'f z; .1 1 IJVISlEi D 2 R ^ g 1 ? 2 o „s H -si 3 S .cu SNIOO '0- <=> 6 ^ 1 ° 2 in ' COFFIN OF SAINT CUTHBERT, fc98 •n (N rs V -> ^ o o ^ N (H g s 7 P ^ In 3 H 2 iri u (n g § jq (j d I s H O D p: ft: u N U ^ "w ^ . s ^ S o (Cl [t. MANX 3T0NE5. , C, 1050-1200. B 1 ND RUNES « (§2-^ <|c^ “ijr~ (“x-/■ (s^~^ X to VjJ h-o <3D —' X Fig, i6i. —^Table of Runic Futhorcs. REPORT ON THE RUTHWELL CROSS. ashen tablets was quite well understood and that for this purpose smooth wooden staves might take the place of paper.^ Runic writ¬ ing however was never really in vogue in the southern and more thoroughly romanized lands which had passed under Germanic control and the few inscriptions, none of them lapidary, known in Gaul and southern Ger¬ many, cease about the year 700, after which time no continental inscriptions save a few of the Viking period and character can be found. On the other hand in the north it was entirely at home, so that runic inscrip¬ tions swarm in the Scandinavian kingdoms, while Great Britain comes next in the richness of its equipment. The runic futhorc of twenty- four letters was carried by the Saxons and Angles on their migration to the west and south that brought them first into what is now north Germany and Holland and then into England, and there is a group of in¬ scriptions known as ‘ Anglo-Frisian ’ that represents the mode of writing employed in the fifth, sixth and seventh centuries in Fries¬ land and in Great Britain. The earliest is on a coin found in Friesland and dating about the fifth century, fig. 161, line j8, but a particu¬ larly interesting example is an inscription of about 500 A.D. incised on wood on the blade of a small model sword found in a Frisian terp or artificial mound, reared to keep build¬ ings out of the way of floods. This is shown in line j8 on fig. 161. The earliest English inscription is also on a coin, which is in the British Museum, dating in the sixth century, fig. 161, line y, and there are other datable inscriptions on coins and other objects of the seventh century from which it is possible to make out a set of characters used up to and about the year 700. Such futhorcs are given in fig. 161, which requires a word of explanation. The vertical columns are numbered, and at the top of each is placed a modern letter or combination of two letters, runic characters correspond¬ ing to each finding their places below. The horizontal spaces are marked with Greek letters and each shows the runic characters found on the monument or group of monu- ' Miscellanea, lib. vi, c. xviii, ad Flavum, in Migne, Patrol, curs, compl, Ser. Lat., LXXXVIII, p. 256. ‘ Barbara fraxineis pingatnr runa tabellis, Quodque papyrus agit, virgula plana valet.’ ments named in the column on the left of the fig. The uppermost space, marked a, contains a standard futhorc derived mainly from a list given in a manuscript from Salz¬ burg now in the Vienna library, that is a copy of one ascribed to the hand of the famous Alcuin. The characters were pub¬ lished by J. M. Kemble in Archaeologia, vol. xxviii, Plate XV, fig. 7. The other horizontal lines give only the characters found on the particular inscribed monuments indicated. These monuments, Frisian or English, j8 to t), are all of early date. Those on the two lowest of the horizontal spaces on the other hand, k, A, are much later, and the Bewcastle and Ruthwell Crosses occupy a position between. The Anglo-Frisian and Scandinavian Futhorcs. The history of the Anglo-Frisian futhorc is interesting in comparison with that of the set of characters used in Scandinavian lands. In the latter region characters as time went on were lost till the futhorc came to consist only in sixteen characters, many of which showed marked variations from the forms of the older common Germanic futhorc that remained in use in the Anglo-Frisian province. In this province moreover not only did the original forms for the most part endure but by modifications of the older ones extra characters were gradually added till the whole number was thirty-three. Some of these later characters. Nos. 29 and 30, are used in the inscriptions on the Ruthwell and Bewcastle crosses and this fact has of course an important bearing on the question of the date of these monuments. It is how¬ ever a very difficult matter for the runologist to decide how early a particular form of letter may have been employed, and certain usages might be expected to be of a local character. These two characters on the crosses do not, as will be seen in line a, occur in the Salzburg MS. the original of which is ascribed to Alcuin, who as a Northumbrian born might be ex¬ pected to know all the characters in use in his own time, about the end of the eighth century, in the region of his birth and up¬ bringing. Hence the appearance of these characters, not recognized in this manuscript. 237 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. on the crosses is a somewhat weighty argu¬ ment against a date for them in the seventh century. The argument is however sensibly weakened b}’ the following consideration. There is a set of runic characters inlaid on the blade of a sword of the pattern known as ‘ scramasax ’ that was found in the Thames and is now preserved in the British Museum, and this futhorc, that agrees in the main with that of the Salzburg manuscript, does not show these later characters. The form of this weapon however belongs rather to the Viking or Danish period of our national antiquities than to the seventh or eighth century, and from this and from the place of its discovery, London, it may be ascribed with some confidence to the ninth or tenth century, by which time the extra characters would certainly have been in use. Their non- appearance on the scramasax and in the Salz¬ burg futhorc may simply show that though in existence at the time they were only used locally or for other reasons were not regarded as a necessary part of the futhorc. It is noteworthy too that though the two crosses with which we are now dealing are obviously contemporary yet the runic characters are by no means entirely the same, as will be seen by comparing the two sets of characters in fig. i6i, 6 and i, ^ and this is just a case of that freedom in the use of the letters which makes it impossible to give too much chronological importance to the appearance or absence in special cases of this or that particular char¬ acter. When this is said however it still remains a fact that the appearance on the crosses of these seemingly late additions to the runic futhorc is an argument, though not by any means a conclusive one, against a date in the seventh century. It is the opinion of some well qualified to judge that in the present state of knowledge on the subject runic evidence of this kind should not be allowed completely to override historical or ' The Bewcastle ‘ C ’ has a straight sloping side stroke, the Ruthwell ‘ C ’ a curved stroke. The ‘ S ’ at Bewcastle sometimes has the vertical strokes prolonged to the whole height of the letter. On the former monument the form of ‘ O ’ in column 23 is never used ; on it there are several examples of ' bind-runes ’ while there is only one doubtful example at Ruthwell, on the uppermost limb of the cross at the back, before the letters read as FAUCETHO. archaeological considerations of weight and substance. There is at any rate no doubt whatsoever that the character of the runes on the crosses is distinctively Anglo-Frisian. This futhorc differs from that found in Gallic and other continental inscriptions in the use of a special character for O, a modification of the original A rune, shown in column 4 in fig. 161. This consideration moreover, coupled with the fact that south of the Channel and the Rhine runic writing died out altogether at an early date, renders hopeless from the first any attempt to connect the monuments with the agency of Normans, or of ecclesiastical scholars and artists from any of the later mediaeval centres of art and culture in France or the Rhineland. If the Normans had brought runes over with them to France these would have had the Scandinavian not the Anglo-Frisian characteristics, and by the time the Vikings settled in Normandy the Franks had forgotten all about runic writing. It may be noted that neither in Frisia, nor in France, Germany, or Italy, are runic in¬ scriptions on stone anywhere in evidence save in the case of certain later monuments of Viking date and origin. The lapidary in¬ scription in runes is a speciality of Scandinavia and England. Runic Evidence for the Date of the Crosses. Within what limits of date and place the Ruthwell and Bewcastle monuments, in their runic aspect, can be located is a question not easy to answer. The runes prove that they are certainly English in their origin, not Celtic, and not continental, but within what periods of old English history and in what parts of Great Britain their incision was possible is another matter. Save for the presence of the extra characters the futhorc of the monu¬ ments corresponds to those of the older in¬ scriptions up to about yoo a.d. shown in lines j 8 to 77, and might be of the same early date, but as a fact this Anglian futhorc con¬ tinued in occasional use for a couple of cen¬ turies or more after 700. The very limited knowledge that exists at present as to the chronology of the later stones and of most of the other monuments on which these char- 238 REPORT ON THE RUTHWELL CROSS. acters appear, and the uncertainty as regards the dates and local provenance of English manuscripts where runes are to be found, make dogmatic statements hazardous. There is, however, one class of datable objects that offer specimens of runes, and these are coins. On coins some chronological con¬ clusions may be based, and the evidence of coins is against any late survival of Anglo- Saxon runes. Runic inscriptions on our native coins as a whole are rare, but they are much more common with us than in Gaul where there are only the faintest traces of runic characters on the early triental coins. On our English coins of the seventh century there are names of kings in full runic writing, but in the eighth and ninth centuries such runic characters as appear are mixed with Roman letters that seem to be superseding them, and we hnd a name written sometimes partly in runes and partly in Roman capitals and at other times in the latter alone. The latest example given in the British Museum catalogue of Anglo-Saxon coins ^ is North¬ umbrian of the ninth century, and from beginning to end there are no runes on the coins of Wessex. On the other hand, a runic N is used for the last letter of the name ALHSTAN on the ring in the Victoria and Albert Museum, the supposed property of that prelate who died bishop of Sherborne in 867 A.D. The fact that runes seem to survive on the coins longer in the case of the names of the moneyers than in those of the kings ’ may be taken as showing that this was a popular rather than an official style of writ¬ ing, a character runologists have assigned to it from its first inception among the Goths. It is a very doubtful question therefore (i) how far in the ninth or tenth century, runes of this kind were employed in the south of England, as for instance at Winchester, and (2) how near to the Norman Conquest their use may in any part of the country have ex¬ tended. It is certain however that, as in Europe generally so in Great Britain, runic writing is distinctively a northern speciality, though there are runic monuments in the south, such as the inscribed tombstones at Canterbury and at Dover, of which due account must be taken. ^ London, 1887, p. Ixxxv. ’ Catalogue, ubi supra, p. Ixxxvii. Post-Conquest runic inscriptions exist, but this fact lends no colour to any argument for a corresponding date for the two crosses, because these inscriptions are not of the Anglian type. Owing to the Danish inroads Scandinavian forms of runic writing were introduced into this country, and there are many inscriptions in the Isle of Man that exhibit this character. There is also one specially important runic inscription of this kind in Cumberland, upon a font of the twelfth century, that is of Norman date and work¬ manship, preserved at Bridekirk, while on a carved tympanum at Pennington near Ulver- ston, and on a stone of Carlisle Cathedral, there are inscriptions in Scandinavian runes also of Norman date. The Manx and Cumbrian runes are quite different from the older Anglo- Frisian ones. The distinction between the two sets of characters is made abundantly clear by a comparison of lines k and A in fig. 161 with those above them. Some of the letters, such as those for F, U, R and L remain constant, but in the case of most of the others there are marked changes, e.g., in C, H, S, B, A, in face of which any confusion between the Anglian and the Scandinavian or Manx futhorcs is quite impossible. English ecclesiastical scholars of the twelfth century or those Norman ones that swarmed at the court of David of Scotland, would certainly not have possessed runic knowledge as a matter of tradition or training, and if they had sought for expert assistance in this matter they would undoubtedly have been taught the futhorc in its Scandinavian not its older Anglian form. It is in the very highest degree improbable that they would have found any craftsmen who were masters of the older system, and could compose and carve fairly lengthy inscriptions in good old Northumbrian runes without introducing any Scandinavian characters. The extra letters of columns 29 and 30 are it must be clearly understood not Scandinavian or Alanx, but are modifica¬ tions of previously existing Anglo-Frisian runes. Some remarks on the phonetic value of these extra characters, and on their relation to the other elements of the Anglian futhorc, will be found postea, p. 272 f. It is impossible to emphasize too strongly the fact that on the two crosses there are now legible between 400 and 450 runic characters, and that none 239 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. of these characters displays any tendency towards Manx or Scandinavian forms such as occur on the Bridekirk font or the Norman tympanum at Pennington. The extra charac¬ ters are as Anglian as the rest. The inscriptions on the crosses are in old Anglian runes, and had they been cut in the twelfth century, it would be a case of atavism of a most curious and exceptional kind. That those responsible for the monuments should desire to bring about this archaeological wonder is equally unlikely. M’ould these Latin-trained and Latin-writing scholars have surrendered the whole of the inscriptive spaces of the Bewcastle cross to antiquarian ex¬ periments in a vernacular style of writing that had been out of date for centuries, even incising in runic characters the title of Christ over His efhgy ? M'hat was noticed above, p. 239, about the popular character of runic writing must here be taken into account. To ourselves of course a verse of old Anglo- Saxon poetry or a stave of Anglian runes is of more interest than pages of monkish Latin in the Anglo-Norman style, but the ecclesi¬ astics of that day cared no more for the Anglo-Saxon vernacular or for runes, than the scribes of Ezra’s school for the old Hebrew ballad poetry which they suffered to perish, or the dour Scottish Highland minister of later days for the Gaelic charms and songs and folk-lore the memory of which he would have been glad to bury in oblivion. The Content of the Runic Inscription AT Ruthwell. The foregoing has been occupied with the runes as a system of writing. The contents of the inscriptions thus executed must now have a word. Some amazing interpretations of these have been in times past promulgated, but not by competent scholars or by careful observers. Guesses have been made by older archaeologists that were founded on drawings, squeezes, or casts, from which no really satis¬ factory conclusions can be drawn, and there is a similar danger in reliance on the more modern aid to the investigator, the photo¬ graph. Only inspection of the originals, not only prolonged but repeated under varying conditions of light and aided by a magnifying glass, can avail to distinguish the accidental marks of weathering from the strokes of the tool that give the form of a letter. Scholar¬ ship and diligence have led in the case of both the crosses to a satisfactory result, and it is possible now to ignore the various conjectural renderings based on insufficient knowledge or study, as well as the obiter dicta of those observers who have decided after a hasty glance at the Bewcastle cross that it was impractic¬ able to decipher the runes. A glance at the west face on fig. 154 will show how far this is from being the case. It is now of course perfectly well known that the runes on the lower panels of the two narrow sides of the Ruthwell cross contain some verses from an Anglo-Saxon poem about the Cross of Christ, a more extended version of which is furnished by the manuscript quoted as the ‘ Vercelli Codex.’ The exist¬ ence of this written version makes absolutely certain the general sense of the clearly-cut but damaged inscription on the Ruthwell monument. The point is this. After Dr. Duncan had set up the cross and made it available as a whole for study, but before the version in the Vercelli Codex was known, he submitted drawings of the runes to a Danish scholar who produced a most ludicrous render¬ ing, that was capped by another still more ridiculous attested by a well-known Danish professor.^ So far however were the runes from being incapable of scientific treatment that when a competent Anglo-Saxon scholar, J. M. Kemble, examined them he was able to decipher them with such accuracy that his rendering was afterwards found to be in almost exact agreement with the words of the same poem in the Vercelli Codex, of which at the time Kemble had no knowledge^ Kemble is the father of the modern study of Anglo- Saxon antiquities, and he never gave any better proof of his mastery of this field of study than this brilliant success. In the case of the runes on the other parts of the Ruthwell cross little or nothing can be done in the way of interpretation. The only characters that are quite clear on these other parts are those on the lower part of the sinis¬ ter margin of the upper panel on the east face, the north-east corner of the shaft, and there are the runic equivalents of the letters 1 ArchcBologia Scotica, IV. 313 f., Edinburgh, 1857. 2 ArchcBologia, XXVIII, 327. REPORT ON THE RUTHWELL CROSS. D^GlSG^F which do not occur in the Vercelli version of the poem and have not been interpreted. It must be noted too that these letters are set differently from those across the margins of the lower panels, and agree in their vertical alignment with the lettering on the broad faces. It seems therefore im¬ probable that other parts of the poem were inscribed on the upper panels. On the head of the cross, along the margins of the panel with the bird, runic characters are on the dexter margin easy to be read, but those on the other margin cannot be deciphered with any assurance. Stephens read into these signs the often-quoted words cadmon me fawed, ‘ Cadmon made me ’, and interpreted them to mean that the poet of that name was the author of the Anglo-Saxon poem in¬ scribed on the narrow sides. It needs hardly to be said that the ‘ me ’ in such a case would refer to the whole monument, not to a mere portion of the inscribed matter, and Cadmon is not known as the appellation of a carver. The fact is that neither the name Cadmon nor any other word or words can now be deciphered on the top or the sinister margin of the panel, but on the other hand, as the illustration, fig. i6o (2), shows, the letters on the dexter margin can be read, as Stephens gives them,^ faucetho with a possible M^ as a bind-rune above and before it. Dr. Henry Bradley describes the former word as ‘ mere jargon, not belonging to any known or possible bid English dialect,- and with ‘ d^egisg^f ’ it must be accepted in the meantime as among the runic inscriptions where the letters are clear but the sense wanting, see p. 278. As was men¬ tioned above, there are runic characters on the sinister margin of the panel with the Visita¬ tion. Read from above downwards and from a position on the dexter side of the monument they are quite certainly the runic equivalents of DOMINN followed (probably) by .E, after which comes the upright stroke of another letter. This wovdd give ‘ Domin(n)ae,’ which makes perfect sense, and is of importance as a Latin word, not a proper name, written in runic characters. Nothing else can be even tenta¬ tively interpreted on the runic parts of the cross. ‘ The Riiihwell Cross, etc., Copenhagen, 1866 (separate publication from the Author’s Old Northern Runic Monuments of Scandinavia and England). - Enc. Brit, nth Ed. art. ‘ Caedmon’. The content of the main inscription on the narrow sides of the cross is made the subject of a separate study which will be found postca, p. 268 f. The Runic Inscription at Bewcastle. In the case of the runes at Bewcastle there have been the same conjectures advanced on insufficient evidence, and statements have more than once been published to the effect that interpretation is here hopeless. These guesses and confessions of failure have been quoted as a reason for ignoring the runes altogether, and a distinguished foreign writer on mediaeval art, who ascribes the monument to the twelfth century, has ‘ arrived at the conclusion that the Bewcastle inscription has, hitherto, been interpreted in too many different ways for any certain date to be based upon it This hardly conveys a just impression. It is true that as at Ruthwell so here various interpretations of a fanciful kind were offered by the antiquaries whose attention was hrst called to the monument, but since the Rev. John Maughan, who was Rector of Bewcastle for 37 years, published in 1857 his reading, this has been generally accepted by recent scholars and is quite well accredited, though it cannot of course be established on the exceptional evidence that exists in the case of Ruthwell. It is in truth quite erroneous to suppose that the interpretation of the Bewcastle runes is a hopeless matter. Time and patience are however essential if there is to be any assur¬ ance of success. Most visitors to that not very accessible spot arrive in the afternoon when the light is full upon the western face where is the inscription, and are disappointed that there is so little to be made out. An hour or two earlier, when the sun was just coming round from the south and the light struck across the inscribed face, the markings would have been seen distinct in light and shade ; and at such a time very many of the char¬ acters are quite unmistakeable though in the case of others accidental weather and lichen marks make sure identification difficult. Now Maughan expressly tells his readers that it was only by examining the characters ‘ Commendatore Rivoira, in The Burlington Maga¬ zine, April, 1912, p. 24. 241 16 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. repeatedly in all sorts of lights that he was able to assure himself of their forms, and the readings that he finally published were the outcome of much time and care. Professor Victor,^ as a trained scholar, came to Bewcastle as he tells us ‘ with a prejudice rather against than in favour of Maughan’s reading ’ but confesses that against his expectation he found that Maughan was in large measure right ! - As the result of a repeated examin¬ ation of the stone in every part in different lights with a magnifying glass there is no hesitation in saying that of all the characters Maughan affected to read thereon well nigh half can be read now with reasonable assur¬ ance, and confirm in general the readings of Maughan. There is really no practical doubt about the general interpretation of at any rate the first six of the nine lines of the main inscription, and these run ‘ This sigbecn thun setton Hwaetred, Wothgaer (there follow letters apparently forming a third or a third and a fourth name) aeft Alcfrithum . . kun- ing ’, etc., etc., words that have been ren¬ dered ‘ This slender sign of victory was set up by Hwaetred, Wothgaer * * * * for Alcfrith (* *) king ’, etc., etc.. The familiar pre¬ position ‘ aeft our ‘ after ’, naturally intro¬ duces a proper name, and this shows that the characters which follow it are not a mere fortuitous set of letters belonging to different words that happen to come together in this order, but a personal appellation. The word ‘ kuning ’ at the end of the line is quite clear. There are still plainer runes than these upon the Bewcastle cross. The title of Christ, ges- sus CRiSTTUS, above the figure below which is the main inscription, is absolutely clear, but the most legible word in the whole monument is to be found low down on the northern side, where, as Maughan pointed out, it was shel¬ tered by the contiguous wall of the church and preserved from the action of the weather. This word, with the runes and the translitera- 1 Vietor’s work, Die Northumhrischen Runensteine, Marburg, 1895, must be in the hands of everyone who studies this subject. ^ Die Northumhrischen Runensteine, p. 14. ' Ich habe die Pausen zwar nicht ohne Riicksicht auf die mir wohlbekannte, wesentlich Maughan’sche Lesung bei Stephens, jedoch eher mit einem Vorurteil gegen als fiir sie, gepriift und wenigstens in der Hauptin- schrift iiber Erwarten viel Maugansches Wiederge- unden.’ tion, is shown in No. 3 on fig. 163. Readers of Bede will remember that Oswy, King of Northumbria had a son Alcfrith, whom he seems to have made under-king in Deira, and who as a friend and abettor of Wilfrid was instrumental in bringing about the famous Synod of Whitby, so momentous for the history of the early Anglo-Saxon church. Alcfrith, about whom we hear a good deal from Bede as well as from Eddius, married, the historian tells us, the princess Cuniburga daughter of Penda of Mercia, a lady of such distinction that when later in her life she presided over the convent at Castor near Peterborough she so impressed herself on the locality that down to modern times a path in the parish has been known as ‘ Lady Connyburrow’s wayCuniburga’s name still confronts us on the Bewcastle cross. Maughan professed to read a good many other words and names on the cross that can¬ not now be established, but it needs hardly to be pointed out how important in the light of history is the conjunction on the stone of these two known names of a Northumbrian prince and his royal bride, connected as he was with that notable event, the Synod of Whitby of 664 A.D. The adherents of a seventh-century date for the cross or crosses, for Bewcastle carries with it Ruthwell, natur¬ ally claim this as an argument on their side of almost overpowering weight. The attempts on the other side to explain away ‘ Cyniburyg ’ and make it mean something quite different from its obvious significance need hardly be taken seriously. The name is certainly not a modern forgery, for William Nicolson, Arch¬ deacon and later on Bishop of Carlisle, read it on the cross more than two centuries ago.^ The Palaeography of the Latin Inscriptions. On the palaeography of the Latin Inscrip¬ tions the following may be noted. Most of 1 Bridges, History of Northamptonshire, Oxford, 1791, ii, 499, ‘ In Castre field is a ridge or balk, from Kyneburga, corruptly called " Lady Connyburrow’s way ” The writer visited the spot a year or two ago and found the name still remembered. 2 Professor Albert Cook published in Yale Studies in English, No. L, New York, 1914, an extremely useful series of reprints of notices of the Bewcastle Cross, under the title ‘ Some Accounts of the Bew¬ castle Cross between the years 1607 and 1861 ’. Nicol- son’s notices are printed pp. 3 f., 9 f. 242 REPORT ON THE RUTHWELL CROSS. the letters in these are Roman capitals modified in a fashion represented in the great illumin¬ ated manuscripts of the Hiberno-Saxon school, such as the Books of Durrow and of Kells, the Gospels of Lindisfarne, and the Gospels of St. Chad at Lichfield. In these manuscripts capital and minuscule letters, displayed on the sumptuous pages where a few words only of text are set in an elaborate decorative scheme, take sometimes quite fantastic shapes but admit of a normal alphabet being formed from them. From an alphabet of this general character are drawn most of the letters in the inscriptions on the stone, and those that are capitals are given in line aa in fig. 162, the inscription round the Christ / 3 ^, 8 in fig. 162. H is of minuscule form in both. M takes a remarkable shape alike on the stone and in the manuscripts, see squares 88, 12, 13. On the great quoniam QUIDEM page in Lindisfarne, the verso of fol. 139, it appears as a vertical line crossed by three parallel bars. It also occurs in Lin¬ disfarne in ordinary minuscule form, and the form in squares 12, 13, is really a minuscule, the line joining the three upright strokes being placed half-way up them instead of at the top. The capital M is used however in Durrow. The N appears in the form, of squares 14, 15 also in the manuscripts, and curiously enough it is there used in this its majuscule form in the half-uncial writing of otot LETTERS RUTHWELL CAPS. / 2 3 4 3 0 7 8 9 /o 11 12 131 Zis 16 17 18 19 20 21 22 23 24 25 A AA B B C C D D E E F F C H 1 1 L L M 1 N 0 oc p P Q R R s X T T u U V V X X dd RUTHWELL minuscules T h w w q c (ift yy HIBERNO'SAXOKMS^ ORNAMENTAL CAPS L E F c I L U 0^ F P X T u X yy S-S Do. Do. MINUSCULES. h t) h ffim a a abbreviations of ET (AND) e z HJB G- ERNO (X ■5AX c- ON/^ -CT M 5. s A !&&. cx LAT A ER d EX AW u FLE 5 MOD. &. TlRSldNS a Fig. 162.—Table of Letters. of the northern face, where the words are spaced freely and divided by points, having been taken as providing the standard forms. Others of the letters are minuscules and are shown in line j8j8. The manuscripts agree with the inscription in this mixture of capitals and minuscules, but the particular letters do not always correspond. Among the capitals A, C, D, E, F, I, L, O, R, S, T, U, V, X, are found in practically the same forms in the inscription and on the elaborate pages of the manuscripts, though of course the pen works more freely than the pick or chisel and can give more and sharper detail, with varia¬ tions in the thickness of lines. The B is com¬ monly in the latter of minuscule form, but on the cross is a capital. G, which in the manu¬ scripts is either of normal shape, or, as in the Book of Kells, takes a very curious form, appears on the stone in the shape of the common minuscide shown in the square the text, the desire being to avoid confusion between the minuscule ‘ n ’ and ‘ r O is a capital in both and appears in both in the two forms shown in squares aa, 16, 17. P, which is normal on the stone, has in the manuscripts a flourish below the loop which looks like a second loop and almost turns the letter into a B. The normal form however appears in the Book of Durrow, beginning the second word in St. John’s Gospel. Q is a minuscule both in the inscription and in the manuscripts. T occurs on stone and in manuscripts both in majuscule and minuscule form, and the latter use of it on the stone at Ruthwell is a point of interest. On the margin below the scene of Christ with the Magdalen on the southern face the inscription ends with the word TERGEBAT. There was not room here for the full word in capitals, so the last two ^ Sir E. Maunde Thompson, An Introduction to Greek and Latin PalcBOgraphy, Oxford, 1912, p. 374. 243 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. letters were made smaller and placed one on the top of the other. The A is a capital but the T a minuscule, and it is interesting to note that in the Gospels of Lindisfarne the ver}^ same thing occurs, as will be seen in the word ERAT on the recto of folio 211. This is a small detail but it is b\’ no means without signifi¬ cance as illustrating the rather close resem¬ blance in the palaeographic aspect between the writing on the stone and in the early manuscripts. There is nothing on the former that does not appear in the latter, and points of connection like the M of squares 12, 13, are certainly remarkable. It must not of course be taken that the forms on the stone are only to be found on monuments of early date. The majority can be found paralleled on much later monuments. For example, this form of M is early, but is not confined to the earliest group of manuscripts, for West- wood gives two examples from the ninth century on plates 21 and 28 of his Facsimiles of Manuscripts,'^ while it is prominent also on a coin of Edward the Elder at the beginning of the tenth. To illustrate further this point there may be introduced a brief reference to a datable inscription on stone from the very end of the Anglo-Saxon period. This is the well- known inscription on the sun-dial from the north door of Kirkdale church near Kirkby Moorside, Yorkshire, and it dates within a few years of 1060 a.d. It suffices to say that the ornamental A, the diamond-shaped O, the S with straight lines, occur in it. On the other hand a comparison may be made between the Ruthwell lettering and that found in the famous Benedictional of iEthelwold, bishop of Winchester from 963 to 984. This is a notable work of West Saxon art of the tenth century, and the facsimile publication ^ is worth examining in view of certain con¬ siderations afterwards to be adduced, postea, p. 262. Now the Benedictional is a work of the southern scriptoria that were less open to the influences of Celtic calligraphy than those of Northumbria, and the majuscule characters throughout are of the classical Roman form, without any of those fanciful elements which ' London, 1868. ^ Oxford, for the Roxburghe Club, 1910. The manuscript is in the library of the Duke of Devon¬ shire at Chats worth. occur in the early Hiberno-Saxon manuscripts and on the Ruthwell cross. When capitals are in use in the Benedictional there is no such admixture of minuscules as we have just come to know. A final point that is of some interest con¬ cerns the abbreviations for and or rather et. In early mediaeval manuscripts this takes two forms. One is the old Tironian sign,’- an inheritance from classicaT antiquity, that re¬ sembles the Arabic numeral 7, and the other is a ligature of the two letters that has survived to our own time in the familiar &. The 7 sign was apparently specially favoured by Celtic scribes, and Professor Lindsay believes that the Irish set the fashion of its use to the scribes on the Continent.^ The Boniface Gospels at Fulda, which he accepts as of the first half of the eighth century in date,^ use it, and so does the datable Book of Armagh of the year 807,* but on the Continent M. Prou reckons that it was taken into the minuscule script from about the tenth century, and Canon Reusens is of the same opinion—‘ des le X® siecle, apparait le signe en forme de 7 ’ etc.^ In that century and the following it was certainly very common, and likewise in the twelfth though it alternated then with other indications of the conjunction. It is used on the Bridekirk font of the twelfth century, but it also occurs on the Kirkdale sun-dial, and on one of the Hartlepool stones of about the eighth century. The second form of the abbreviation noticed above is a simple ligature of an uncial E of a rounded form with a minuscule T.® It occurs universally in the sumptuous early Hiberno- Saxon manuscripts above referred to in forms 1 Tironian signs or notes are the elements of an ancient system of tachygraphy equivalent in its inten¬ tion to our shorthand. The name comes from one Tiro, a freedman of Cicero, who was supposed to have invented the system in order to take down his master’s words. See M. Prou, Manuel de Palio- gvaphie, Paris, 1910, p. 118 f. ^ Contractions in Early Latin Minuscule MSS., Oxford, 1908, pp. 12, 34. “ Early Irish Minuscule Script, Oxford, 1910, p. 5, and facsimile, Plate III. * Ibid., p. 26 and Plate IX. ^ Elements de Paleographie, Louvain, 1899, p. 104. ® Canon Reusens says a capital T reversed, x> but there is no question that the letter is a minuscule, see the forms of the two letters in squares ee, i, 2, in fig. 162. 244 REPORT ON THE RUTHWELL CROSS. such as those shown in squares ee, 3 to 5, taken respectively from the Gospels of Lindis- farne, the Book of Durrow, and the Book of Kells. Here the two letters have so far coalesced that the dexter portion of the horizontal head of the T serves as the central horizontal of the E while the sinister por¬ tion is cut off and appears as a separate stroke. In the Kells example the upper curve of the E is made into a loop like that of the cursive e of our handwriting. Square 6-7, shows another Lindisfarne example, square 8 one from the Bodleian ‘ Rushworth ’ Gospels or Gospels of MacRegol of about a.d. 800, and square 9 a specimen from the Canterbury Gospels of the latter part of the eighth century, in the British Museum,^ while there follow in squares ii to 13 the two specimens of the ligature which appear on the Ruthwell Cross. The E here is formed with two closed loops, but the lower curved part of the minuscule T and the sinister portion of the head are still in evidence. The after history of the ligature is interest¬ ing to trace. In the eighth and ninth centuries it is the common contraction for et especially in well-written manuscripts and appears also in the Benedictional of ^Ethelwold of the tenth. The form in square 16 is from a capitular of Charles the Great of the year 825, in 17 from the Benedictional. In these last there is a modification that is almost universal in the manuscripts of the later period in which the ligature occurs. The upper curve or loop of the E is atrophied and reduced to a small flourish like a hook, while what was originally the sinister portion of the head of the T is tacked on to the lower curve of the E of which it forms the termination. Exceptionally this hook or flourish is found earlier, for square 15 gives an example from the fragmentary gospels of about the seventh century in Trinity College, Dublin,- but here the portion of the head of the T is still distinct. The final morphological change seen in examples of the twelfth century turns this hook or flourish into a complete loop which is taken in to the general sweep of the curves of the letter. Squares 19 and 21 contain twelfth-century examples, the former from the Cottonian MS. of the Anglo-Saxon * Sir E. Maunde Thompson’s Introduction, fac¬ simile No. 141. “ National MSS. of Ireland, I, 2. Chronicle, Dom. A, VIII ; the latter from a fragment of the Bible in Edinburgh University Library. It will be seen at once that these are the immediate progenitors of the familiar sign for ‘ and ’ of the modern printer, as seen in square 23. Squares 24 and 25 show some forms of the Tironian sign 7 used as an abbre¬ viation of the conjunction. The question of the relations in point of usage between these two forms for et, and of the chronological significance of the shape of the ligature on the Ruthwell cross may have a word. Both forms are possible at any period, for the Tironian sign and the letters et are alike classical, and the ligature, as we have just seen, has survived to modern times. The ligature is more dignified than the 7, and the latter comes first into use where writing is compressed. The early use of it in the Book of Armagh of 807 is thus explained, for Sir E. Maunde Thompson notes of the manuscript that the principal object of the scribe seems to have been ‘ to pack into the page as much as possible and in a twelfth-century manu¬ script, the Gospels of Maelbrigte,^ the ligature (see square 18) is used at the beginning of a sentence, the 7 within the sentences. The ligature is universal in the finely-written Hiberno-Saxon manuscripts, and would be expected in lapidary inscriptions, though the late Saxon example at Kirkdale of about 1060 (p. 244), has the 7 used several times and no ligature. There is no doubt that the form of the ligature in which the E is completely formed with an upper and an under curve is earlier than the one in which the upper curve is atrophied into a hook or flourish. Though square 15 is a curious exception, this is borne out by the examples in squares 3 to 9, and the form of the ligature on the Ruthwell cross is certainly early as it resembles these far more than it does the later examples, such as that from the Benedictional of .iEthelwold of the tenth century (square 17) which is more than half-way towards the modern &. At the same time, the dated examples which appear in Fig. 162 of these forms of the ligature show that no rigid system of chronology can be derived from them. The point may seem a small one, but such details sometimes furnish * Introduction, etc., p. 380. ^ National MSS. of Ireland, 1 . 40. 245 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. very valuable chronological indications. In this case the minor piece of evidence is quite in favour of the early date pro¬ posed for the cross. The Form of the Cross and of the Cross- Head AND THE Geometrical Ornament. This critique of the inscriptions may be suitably followed by a few words, first, on the significance of the form of the monument as a whole and of the cross-head, next on the linear or geometrical ornament on the two shafts, and, thirdly, on the floral and animal decoration, the figure sculpture being reserved for treatment to the last. It was noticed above that the section of the shafts of both the crosses is oblong and not square, and that this is characteristic of cross¬ shafts in general. As a rule these have greater extension of surface at the front and back than at the sides and exhibit in this respect to some extent the character of slabs. The morpho¬ logist explains this by drawing out a scheme of the evolution of the free-standing cross. First we have unhewn or rudely shaped mono¬ liths on which is incised a small cross alone or in company with an inscription. The stones at Kirkmadrine in Galloway, with many in Wales and other Celtic regions, are of this character. Next, the cross becomes more important and grows into a predominant position on the slab, being often carved in relief and not merely incised. A further ad¬ vance is made when the form of the slab is affected, it may be by piercing it with openings where the arms of the cross meet in the centre, or by extending the ends of the arms a little beyond the main outline of the slab. Later on, so it is assumed, this process is continued and the cross form is gradually extracted from the slab and rendered in all its proportions as a solid thing of three dimensions. A re¬ miniscence of the slab is however still retained in that the cross is wider at front and back than at the sides. Naturally such an evolu¬ tion would take time, and the free-standing cross would be a much later product than the slab with incised or modelled cross on its face. Hence morphology might be considered to furnish an argument against an early date for the two crosses on the assumption that these must have been preceded by cross-slabs from which they were gradually evolved. Underlying this argument however there are two fallacies. It is true that all these inter¬ mediate forms occur, and may be found exemplified, for example, among the incised and sculptured stones of the Isle of Man, but it would be very rash to assume that they follow each other in chronological sequence. Furthermore, the starting point is not the incised cross on a slab, for this is merely a representation, and the real original is the free¬ standing cross of three dimensions that can be seen and handled. When King Oswald before the battle of Heavenfield set up a cross of wood of substantial size,^ he did not start by incising an outline on a board, but reproduced in a rough form yet in all its dimensions the traditional object of Christian adoration. In the same way when we are told that St. Cuth- bert set up a cross by his oratory on Fame Island,^ or read in the Latin Lives of Irish Saints recently edited by Mr. Plummer ® of the crosses, sometimes of stone, which they erected, we must believe that the idea of the cross in such cases materialized at once in a solid shape, with shaft and transverse arm and head, without the interpolation of any process of ‘ evolution ’. Hence a free-standing cross of stone is in itself quite possible at the early date suggested for Bewcastle and Ruthwell, and the shape of the shaft in section may have been motived by quite other considerations than the reminiscence of an earlier slab. The form of the cross shown in the crucifixion scene on the base of the Ruthwell shaft shows great proportionate width, and can never have been imagined by the artist as of a corre¬ sponding thickness. With regard to the form of the cross-head the following may be noted. Ornamental forms of the cross occur early both in drawn and modelled representations of the object, and in reproductions of it in three dimensions, such as the small portable crosses often found in Anglo-Saxon cemeteries, as for example Nos. 4, 5, and 6 on fig. 163. It is noteworthy however that the cross forms used in classical and classically-derived art, that is, among the Mediterranean peoples and in Gaul, are very simple, and the elaborate and fanciful 1 Bede, Hist. Eccl., iii, 2. ^ Bede, Vit. Cudb., c. xxxviii. ® Oxford, 1910. 246 Ancient and Historical Monuments —Hittnfries, hlC, l li;. i6j. — Variims ilhistralivc pieces. 246. 5 j- fiS’! ■c,> ‘T •• . ' *, ’ v ' ;■' •■™li:S ■■I f^-.' ’r.'i'^' R m?i m ,l)aCvi Mj. Wig'll tjf vji; SI <.• •* ; 1 ''•;" ■-j!fS I 1%*' m.’.. ■t- - ■■■m -•v '>v: -_’i* ■• ’ M '.'XQ'-V ..V B ■‘.'• ■<.y -^i ■ i. ^ r' * > '*A REPORT ON THE RUTHWELL CROSS. forms common in our own country appear to be of Teutonic origin, for we find them in early Germanic tomb furniture both in this country and on the Continent, and it was largely from this source that they were taken over for the carved stones and illuminated manuscripts. The normal ornamented cross form in classical lands is the cross with arms broadened out at the ends, or cross ‘ patty ^ at first a mere reproduction of the familiar ‘ serif ’, or broadened-out termination of the strokes of letters. The scheme of it is shown in fig. 164 {a). Later on this broadening out is carried down in an even curve to the centre where the arms intersect, and this curve, the arc of a circle, is continued for the opposite arm, so that the two arcs are struck from the same centre, fig. 164 [b). This form of the cross occurs on Syrian buildings of the fifth and Fig. 164. sixth centuries, and No. 7 on fig. 163 shows it on a piece of Anglo-Saxon tomb furniture. It is an innovation, apparently of Teutonic origin, when an arc of a circle is used to define the outline not of two opposite arms but of two adjacent arms of the cross, so that a form is produced like that of the well-known pectoral cross at Durham, of the seventh century, found in 1827 on the body of St. Cuthbert, the general outline of which is rendered, fig. 164 (c). Fig. 163 (4), a pin from a Jutish grave in Kent, shows the same scheme with the circles carried further round so that the expansion of the arms is greatly increased. If this increase be cut off by the arcs of other circles, as is suggested by the dotted lines on the dexter arm of fig. 164 (c), a form is readily arrived at similar to that of the head of the Ruthwell Cross, and it may easily be seen that this is a natural method of treatment for a # ^ That is, spread out in the form of an extended ‘ patte ’, like a cat’s claws when she is stretching her¬ self (Littre). monumental cross-head of the seventh century. The accepted form at the time for the decora¬ tive cross was something like fig. 163 (4), but this could only be cut in wood on a large scale by using a very wide piece of stuff, whereas the modified form, where the second arc cuts off the projecting parts, would suit the squared log of timber employed to make an early monumental cross of that material. Such a wooden cross, like that of King Oswald, would certainly pre¬ cede, and so give its form to, the monumental cross of stone, and this gives a simple explana¬ tion of the Ruthwell shape. This double-curved form is very popular in the north of England, and was evidently in common use shortly before the Norman Con¬ quest, for it is that of all the four cross-heads of a date in the earlier part of the eleventh cen¬ tury that were found in the foundations of the Chapter House at Durham. ^ It occurs how¬ ever earlier, as on the important and fairly early cross at Rothbury, Northumberland (No. 8 on fig. 163) at Masham, at Lastingham, Yorks, and elsewhere, but mostly in England, though there are Scottish examples from the interesting ecclesiastical centre at Hoddom in Dumfriesshire not far from Ruthwell (see this volume, p. loi), and an obviously late one at Dupplin in Perthshire.^ At Hoddom there occurs on a grave cover recently discovered in the ancient burial-ground a cross in low relief that reproduces the form of the head and shaft of the Ruthwell cross as these must have ap¬ peared in combination. The date of the slab cannot be ascertained, but it certainly must have been cut at a time when the Ruthwell cross was still perfect, and undoubtedly re¬ produces its form though not its proportions. A drawing of it is therefore added, fig. 165. Furthermore — a fact suggesting that the double curve may also have appeared cpiite early—there is no question that the same was the shape of the head of the so-called ‘ Acca ’ cross, the fragments of which were brought from Hexham, and are now in the Cathedral Library, Durham. We have the authority of a Hexham writer of the twelfth century in the Historia Regum published under the ^ F. H. Haverfield and W. Greenwell, Catalogue of the Sculptured and Inscribed Stones in the Cathedral Library, Durham, 1899, p. 79 f. - Early Christian Monuments of Scotland, Edin¬ burgh, 1903, p. 319 f. 247 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. name of Symeon of Durham ^ that when Acca, bishop of Hexham, was buried outside the eastern end of the church at Hexham in the year 740 a.d., ‘ two stone crosses adorned with wonderful carving were set up, one at his head, the other at his feet ’. Near this same spot in 1858 was found a portion of a carved cross-shaft, and this with other portions which came to light in the neighbourhood make up the ‘ Acca ’ cross in question, the top part of w'hich is shown in a photograph, fig. 163 (i). It will there be seen that enough is left of the lower part of the head to indicate a form like that of the Ruthw’ell and the Roth- bury crosses, and if this be really one of the crosses set up in 740 its import¬ ance for the dating of the Ruthwell monument is obvious. The floral ornamenta¬ tion of the ‘ Acca ’ cross¬ shaft exhibits, as will be noticed, a motive similar to that of the foliage at Bew’castle, but the head is enriched largely with row^s of pellets, which are familiar elements in Norman decoration. These may be taken in connection wdth the panel of carved chequers con¬ spicuous on the northern face of the Bewcastle cross w’hich carries with it also a pronounced Norman suggestion, for though the chequer ornament in itself is of all times and countries chequers carved in stone are characteristi¬ cally Norman. A fragment from Hexham in Durham Cathedral Library, ascribed to an early date,® is marked with chequers. Che¬ quers and pellets are moreover common in Irish ornamentation, and occur carved in stone in w'ork where no Norman influence ’ Printed by the Surtees Society in its Publication No. 51, Durham, 1868. See pp. xxiv, Ixxii, 14. ^ Canon Greenwell, Catalogue, p. 58. ® Greenwell, Catalogue, p. 65. is to be traced. Painted chequers are found in the illuminations in the Book of Durrow, w'hile chequers in relief in metal work occur on the cover of the Stowe Missal in the collection of the Royal Irish Academy in Dublin of a date between 1023 and 1052, and in stone on the doorway of the ruined church at Maghera, Co. London¬ derry, wLich may date about the same epoch. The Maghera chequers are curiously like those at Bewxastle but the squares are a trifle smaller. Pellets occur on the famous enriched doorw'ay at Freshford, County Kilkenny, in the beautiful bases of the chancel arch at St Caimin’s, Iniscealtra, and very commonly elsewhere in ‘ Irish Romanesque ’, a style influenced by, but in its origin and early development independent of, Norman art. To these affinities between details of the Acca and Bewcastle crosses on one side, and Irish work on the other, may be added the fact that the subject of the hermits Paul and Anthony breaking bread in the desert, conspicuous at Ruthwell, occurs also several times on the Irish carved crosses of about the eleventh century. In regard to these con¬ nections, of which note should assuredly be taken, Mr. Champneys ^ thinks that the Irish borrow^ed from the Northumbrians, but the relations between Irish and Anglo-Saxon art open up, as has been said, a question that has never received systematic treatment, and the matter can only be referred to here in passing. On the Bewcastle shaft but not at Ruthwell occur panels of interlacing work. Such orna¬ ment is of course excessively common both on the carved crosses and in the manuscripts and metal work, alike in Ireland and Great Britain. The Bewcastle panels are very finely wrought, and it may be worth mentioning that two of them agree in part with inter¬ lacing patterns found in the Gospels of Lindis- farne of the date of about 700 a.d.® A sun- ^ Irish Ecclesiastical Architecture, London, 1910, p. 86. 2 See postea, p. 265. The Bewcastle panel below the chequers on the north face may be compared with fol. 92 r. of the Gospels of Lindisfarne (the beginning of the Gospel of St. Mark), the middle panel on the sinister side of the stem of the second I in ‘ initium in the centre of the page ; and the Bewcastle panel at the bottom of the south side with fol. 2 v. (a whole page of ornament) the top parts of the panels just under the indentations of the arms of the central cross. REPORT ON THE RUTHWELL CROSS. dial occurs in the middle of one of the foliage panels on the southern face at Bewcastle. This has little chronological sig¬ nificance, for though later Saxon and Nor¬ man sundials are common yet a sundial may be seen in situ on the southern face of the very early Saxon church at Escomb, Co. Durham, and a Roman example not unlike that at Bewcastle is preserved in the Museum at Chesters on the Roman wall not many miles away.^ The Foliage and Animal Decoration. Attention may now be paid to the question of the provenance and affinities of the foliage and animal decoration on three sides of the Bewcastle cross and the two narrow faces of that at Ruthwell. Besides the tall panels with animals and birds in foliage scrolls, there are panels of pure foliage ornament at Bew¬ castle, that are for the boldness of their design and their just distribution among the best things of the kind in art. There needs only to be noticed here the peculiarity that on the principal panel, on the southern face, the stems divide and the several sprays inter¬ penetrate. The presence of unmistakeable bunches of grapes shows that the motive is the vine. Now the division of a stem is in foliage ornament generally so rare ^ that special attention has been paid to this anomalous treatment on the Northumbrian monument. The source of it is however evident. It comes from the vine ornament on Roman ‘ Samian ’ bowls, an ornament which in this particular form is so common on some types of these vases as to be almost inseparable from them. Nos. I2, 13 on fig. 163 show two examples from the North in the Tullie House Museum, Carlisle. On the ‘ Sa- main ’ vases this division of the main stem ^ Figured in An Account of the Roman Antiquities preserved in the Aluseum at Chesters, London. 1903, p. 198. ^ Rare, that is, out of the English Anglian region. Here it is very common, and there are numerous carved stones in the North that exhibit it. In Gaul, Italy, the Hellenized East, on the contrary, the motive will be sought for in vain. The use of the divided stems on the Acca cross is a link of connection between that monument and those now under discussion. of the scroll into two or three parallel shoots is almost constant ; but in Roman work the shoots do not interpenetrate. The interlace¬ ment on the Bewcastle cross is in accordance with Teutonic practice in the seventh and following centuries. That the long panels on the two crosses also contain vine orna¬ ment, and indeed that the vine is the main foliage motive on both the monuments, may seem a bold statement, but the fol¬ lowing considerations it is believed will justify it. The motive of floral scrolls with birds or other animals is of every age from the late classical to the advanced mediaeval. The particular kind of foliage at the basis of the conventionalized treatment varies. The acanthus was first employed, and birds are introduced with it in the enrichment of the magnificent Greek silver vase from Nikopolis in the Hermitage, Petrograd. In late Greek times the vine came into use, and vine scrolls are common in Roman decoration. The vine was at once co-opted into Christian art and given a sacred significance, the birds or animals pecking at the fruit signifying,the faithful drawing sustenance from the True Vine. As such the motive makes its appearance among the very earliest catacomb paintings, and this as we shall presently see is the motive on the crosses. After the Carolingian renaissance however the classical acanthus comes again into favour and from about 800 till the intro¬ duction of naturalistic foliage in the early Gothic epoch the acanthus in its Carolingian form rules supreme, so that it is very hard to find any early mediaeval floral ornament in which it does not make its appearance, and this applies specially to the eleventh and twelfth centuries. The fact that there is no trace on the Ruthwell and Bewcastle crosses of the Carolingian acanthus is one of very great moment from the point of view ot chronology. It would be certainly surprising if such an amount of floral ornament in several different forms were designed and executed in post-Carolingian days without one trace any¬ where of the dominant, almost universal and exclusive, foliage motive of the age. This, it may be noticed in passing, is almost as strong an argument in favour of an early date for the crosses as the occurrence of the names Alcfrith and Cuniburga. 249 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. The floral ornament on the crosses with which we are dealing belongs to the vine period, or at any rate is of vine and not of Carolingian acanthus character. It is obvious however at the first glance that there are details in these floral panels that are not in the least like vine foliage. The animals are sometimes pecking at what are clearly bunches of grapes, but in other places a cluster of berries appears en¬ closed in a sheath of leaves, and elsewhere again the berries seem to have been trans¬ formed into the inner petals of a flower. Examples uill easily be found in the illus¬ trations, figs. 152, 154, and 158. On the Bewcastle cross especially there are rather elaborate flowers. A bunch of grapes growing on the same stem as leaves that form for it a sort of sheath is a botanical impossibility, for in the vine the cluster and the leaf have each its separate stem. In Gaul, Italy, the East, wherever the vine is a familiar plant, no such solecism in its treatment will be found ; how then are we to explain the anomaly of the union of berries and foliage on the crosses ? One possible explanation would be that the plant is not the vine at all, but another and perhaps a better one may be found in the following consideration. The floral scrolls in question may have been designed by an artist who had no personal acquaintance with the vine as a plant, but was taking the motive at second hand from specimens of Roman vine foliage in which the exact anatomy of the leaves and fruit and stems was not clearly indicated. There is a Roman vine scroll carved in stone at Carlisle on a sarcophagus or trough, shown No. II on fig. 163, that is the sort of monu¬ ment that might have given the general idea, while the piece of Roman vine carving shown No. 9 on fig. 163 illustrates the manner in which the mistake above referred to may have arisen. As a fact each bunch of grapes and each leaf has here, as in nature, its separate stalk, but the appearance, partly due to the fact that some of the stalks are broken away, is in parts that of leaves covering the clusters and forming a sheath for them. The actual piece in question is at Budapest, but similar work may have existed in the North of England and given the artist of the cross a hint for his detailed treatment. It is not of course suggested that an Anglian artist evolved the whole scheme of the ornament in independence. The general motive of the foliage scroll with animals introduced was a matter of established tradition, and the de¬ signer was familiar with it from examples on portable objects. In working out the general motive he would receive practical help from Roman examples. The split vine stems on the Bewcastle panels are a proof that this was the case. The variety and ex¬ cellent decorative feeling shown in the treatment of these details of the plants is another matter. There we have to do with the artistic qualities of the design and execution, and these will more fittingly re¬ ceive notice in connection with the figure sculpture. There are certain leaves among the foliage forms which are reminiscent of the classical palmette, and are worthy of a moment’s attention, because similar leaves occur on other datable objects of Anglo-Saxon art. For example, a leaf from Bewcastle resembles one on a small bronze object in the Cambridge Museum found in Suffolk and dating perhaps about 500 A.D. The similarity is obvious, and both leaves resemble one that occurs on a piece of early Teutonic metal work at Hanover, that can be dated about the fifth century.^ Again, on the Ruthwell and Bew¬ castle photographs (figs. 152, 154, and 158) there are several fantastic beasts with two fore¬ legs and a tail that goes off into foliage. No. 10 on fig. 163 is a small object in cast bronze found in a Jutish tomb in Kent and dating in the latter part of the seventh century. The creatures are, like the leaves, curiously alike, but too much should not be made of these resemblances in detail, because in the wealth of floral and animal ornament which we meet in other periods of the middle ages other approximate parallels might perhaps with some searching be discovered. At the same time, as was noticed before, such points, though small in themselves, are sometimes of much significance. Critique of the Figure Sculpture. By far the most important element in the decoration of the crosses is the figure ^ Safin, Altgermanische Thierornamentik, p. 176. 250 REPORT ON THE RUTHWELL CROSS. sculpture, to which attention must now be directed. One very effective criterion of date for monu¬ ments of this kind is iconography. Certain subjects belong to particular periods and do not occur outside ascertainable limits. Endeavours have been made to demonstrate the late date of the crosses on the ground that certain subjects upon them are not known in art before the more advanced mediaeval period. Definite information on this point is therefore a desideratum, and it is not difficult to supply it. Taking the figure subjects in order as they have been described, antea, p. 224 f., it may be noted that the Evangelist John with the eagle and Matthew with the angel are familiar Early Christian subjects, occurring for ex¬ ample in mosaics of the eighth century in S. Vitale, Ravenna; ^ John the Baptist, robed and carrying the Agnus Dei, is to be found very conspicuously displayed on the sixth-century ivory chair of Maximian at Ravenna. Christ in Glory is one of the subjects incised in wood on the coffin of St. Cuthbert of 698, ^ but the animals under His feet have been noticed as, apparently, unique. The subject of St. Anthony and St. Paul is of course Egyptian. Its occurrence on some of the Irish crosses has been already referred to, and it has also been recognized on the Scottish carved stone slab at Nigg, Ross-shire. It occurs in twelfth-century work at Vezelay in France.® The Flight into Egypt, rendered in a manner curiously like the representation on the cross, makes its appearance on the golden medallions from Adana in Cilicia now at Constantinople, that date from the sixth or seventh century, and were published by Professor Strzygowski in the appendix to his Etschmiadzin-Evangeliar (Wien, 1891). On the other side of the cross the appear¬ ance of the bird of the falconidae group need cause no surprise. It may be meant for an eagle, and this would explain the appearance of the archer in the panel below. There is no doubt however that the peculiar merit in ' Garrucci, Storia dell' Arte Cristiana, IV, Tav. 261, 263. * Haverfield and Greenwell, Catalogue of Sculptured Stones, etc., p. 141. ® Professor Albert S. Cook, The Date of the Ruthwell and Bewcastle Crosses, Yale University Press, 1912. an artistic sense of the representation should be explained in connection with the love of the Anglo-Saxons for the sport of hawking. It is because the artist knew and loved the falcon of sport that he fastens on this bird form with such keenness. It has been debated how early in the national history the sport was established in this country. The latest writer on the subject^ says that it originated —probably as an importation from the East —among the Germans in the second and third century a.d., and by the sixth century it was in full practice. Professor Montelius notes that in one of the seventh-century tombs at Vendel in Sweden there were found the bones of a sporting falcon, that had evidently been buried with its lord.® The Anglo-Saxons adopted the craft from the continental Ger¬ mans, and the letters of and to Boniface about the middle of the eighth century ® show that at that date English kings were sending over to Germany for falcons of good breed of which there was at the time in Kent a poor supply. Earlier than this however the sport must have already become very popular, for on the early Anglo-Saxon coins, the so-called sceattas, birds are shown repeatedly perched on the wrist of a male figure, and it is difficult to see what can be inferred from this save the vogue of hawking. An archer occurs not only on the Franks casket of about 700, but on a piece of stone carving from Hexham that probably formed part of the sculptured decor¬ ation of the church Wilfrid built there about 675. The Visitation is one of the subjects on the Adana medallions, and filled one of the panels, now lost, on the ivory chair of Maximian at Ravenna."* The Magdalen Washing the Feet of Christ does not seem to be known before the ninth century, but the next subject, Christ Healing the Blind Man, is one of the most familiar on the Early Christian sar¬ cophagi, and in ivory carvings. As was in¬ dicated, antea, p. 230, the theme is rendered * Fritz Roeder, in the article ‘ Falkenbeize ’ in Hoops’s Reallexicon der Germanischen Altertumskunde, Strassburg, 1913. ^ Kulturgeschichte Schwedens, Leipzig, 1906, p. 224. ^ Printed in ]\Iigne, Patrol. Curs. Compl., Ser. Lat., tom. LXXXIX, Boniface to iEthelbald of Mercia in 742, p. 750 ; .Rthelberht II of Kent, 748-762, to Boniface, p. 776. ^ Felix Ravenna, July, 1912, p. 283. 251 HISTORICAL MONUxAIENTS (SCOTLAND) COMMISSION. in an original fashion. The Annunciation is of course a favourite subject in all periods of Christian art save the earliest, and it occurs on the Adana medallions, but there are special features in the representation here that must presenth? be noticed. The Crucifixion, as is well known, is a motive Lorn which the Chris¬ tian artist in the earlier periods stood aloof. It never occurs in the Catacomb paintings nor in the mosaics of the great churches of the fourth and immediately following cen¬ turies. Apart from a few engraved gems the earliest knowm representations are on the carved wooden doors of S. Sabina at Rome and in an ivory carving in the British Museum, both of the fifth century, and in the famous ‘ Rabula ’ codex, a Syriac MS. of the Gospels in the Laurentian Library at Florence dated in the year 586. It w'ould be quite possible in the seventh century, and the dress of the Ruthwell figure w^ould suit an early date. Later on of course the subject becomes exceedingly popular. On the Bew^castle cross the figure of the Falconer has not unnaturally excited sus¬ picion, as it appears to indicate a condition of the sport more advanced than would be possible in the seventh century. The figure, whomever he may represent, stands with the right hand grasping the light stick used for beating the reeds or bushes where birds may be in covert and holds a falcon in orthodox fashion on his gloved left wrist. Below is the creature’s perch. From what has been said above however, it may be judged that the representation would not be impossible at the latter part of the seventh century, though a later date might seem historically more pro¬ bable. To sum up, of the fourteen subjects here enumerated ten are distinctly authen¬ ticated as of early date. Of the four remaining ones Anthony and Paul and the Magdalen at the feet of Christ are not so authenticated, but there is no reason why they should not appear in seventh-century work. The Fal¬ coner and the Fagle or Falcon must be judged on the considerations just adduced, and to some may present difficulties. If the subjects of the figure sculpture on the Ruthw'ell cross indicate on the whole an early date and suggest in some cases an oriental provenance, so too does their style. This style is of course in its origin classical, but it is far more Greek than Roman. It would not be easy to find a Christ-figure in Early Chris¬ tian or mediaeval art more like the Christ of the crosses than the one shown in fig. i6b. This is an outline drawing of a painted figure of the Saviour that existed formerly in the Early Christian cemetery at Alexandria not far from ‘ Pompey’s Pillar ’. It has now perished, but it was published by Neroutsos Bey in his study ‘ L’Ancienne Alexandrie,’ ^ from which it is here reproduced. The resemblance in general character and expres¬ sion is quite unmistakeable. It was natural that the Hellenic tradition should be main¬ tained in greater purity in the eastern parts of the classical world than in Italy, and sculp¬ ture not devoid of the old Attic grace and charm could be produced in Egypt, Syria or Asia Minor dowm to the fifth or sixth century, to the time indeed when the advancing tide of Saracenic invasion overwhelmed these seats of the older classical civilization. Good as some of this late but still Hellenistic sculpture may be, it is not better than some which is found on these Northumbrian crosses. The ^ Paris, Leroux, 1888, p. 49. 252 REPORT ON THE RUTHWELL CROSS. figures here are not Roman in type, but in their grace, elasticity, and slender proportions are Greek, and this applies specially to the Christ, the figures in the Annunciation, the Flight into Egypt, and the Bewcastle ‘ Fal¬ coner Such qualities would be most in evidence in work that stood comparatively near to the classical period and was Greco- oriental in its origin. Some of the subjects have this significance, and it was noticed with surprise by Dr. Stuhlfauth ^ that the attitude of Mary in the Annunciation is the standing one that is characteristically oriental, of what he terms the primitive Syro-Palestinian type. It is another question how sculpture of this type came to be executed in Northumbria. No direct early connection between this region and the Hellenistic East can be proved, but the possibility of such a connection is obvious. The most prominent figure in the Anglo- Saxon Church in the last half of the seventh century was Archbishop Theodore, a native of this very Hellenistic East. As early as the end of the fourth century St. Jerome in com¬ menting on the description in Ezechiel of the ancient commerce of Tyre noticed the com¬ mercial activity of the Syrians of his own day, that carried them into all parts of the Roman empire.^ These Hellenized orientals maintained a traffic in works of art, and at times craftsmen also migrated to the West, for there are inscriptions attesting the presence of Syrian and Alexandrine glass workers in the Rhineland and Gaul.® In an article in the Byzantinische Zeitschrift for 1903^ M. Brehier adduces early mediaeval evidence to show the presence of Syrian merchants at the chief towns of Italy, Gaul, and the Rhine¬ land, who imported into the West the products of eastern Mediterranean lands. Among these, ivory carvings of the late Hellenistic style are quite a possible commodity, and carvers in stone from the same Greco-Oriental centres ^ Die Engel in dev altchristlichen Kunst, Freiburg i. B.. 1897, p. 71. ^ Hieron. in Ezech., cap. xxvii, vers. 15, 16, Migne, XXV, 255. ‘ Usque hoclie autem permanet in Syris ingenitus negotiationis ardor, qui per totum mundum lucri cupiditate discurrunt ’, etc. See also St. Jerome’s Epistle cxxx, in Migne, vol. XXII, p. 1112. ’ J. Pilloy, Etudes sur d’anciens Lieux de Sepultures dans I’Aisne, III, St. Quentin, 1912, p. 296 f. * Les Colonies d’Orientaux en Occident au Com¬ mencement du Moyen Age. may conceivably have found their way to the remoter parts of the West. These are of course mere conjectures, but the monuments and the style of their sculpture are facts, and surprising as these are there must be some explanation of them could it but be found. In any case there is no question that the best of the Ruthwell sculpture finds its affinity in the latest good work of the classical schools in the Hellenistic East. The Style of the Work in its Bearing ON THE Question of Date. In relation therefore to the history of European art as a whole the earliest date to which the monuments could be assigned is the most likely one, but it is different when we view them in connection with the Anglo-Saxon polity and with the story of the Northumbrian Church. It has been argued with some co¬ gency that work on such a scale, in an in¬ tractable material, so full of detail, and of such remarkable excellence, is unlikely at so early a date in the history of Teutonic Britain, and that the particular form of the monuments, the large free-standing cross of stone, is at the time hardly possible in view of the previous history of Christianity in the North. It is pointed out in this last connection that Bede lays stress on the fact that when in the year 634 King Oswald erected an extemporized cross of wood on the spot where he was about to achieve the victory of the Heavenly Field, this was the first symbol of the Christian faith that had been set up in the Bernician realm, and it is asked whether it is reasonable to suppose that within forty years such very elaborate examples of the decorated cross of carved stone on a monumental scale could have been made and erected in this v^ery region. Here then is an undoubted difficulty that has proved decisive in the minds of some of our leading archaeologists, and forced them, while maintaining the Anglian origin of the crosses, to relegate them to a later period than the seventh century. Three questions have here to be dis¬ tinguished. One is the general question of the possibility or likelihood of great achieve¬ ments at comparatively early periods ; a second concerns the history of Anglian Chris¬ tianity with especial reference to the religious 253 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. use of the symbol of the Cross and to the results of the Synod of Mliitby, while the third relates to the political history of North¬ umbria. Common sense postulates a gradual process of development leading up to great achieve¬ ments as the ultimate outcome. As a matter of fact, common sense in this is for once at fault, for in the history of the arts the greatest works in a particular style or phase of art sometimes make their appearance surprisingly early, and everything in the same kind that comes after represents a decline rather than an advance. There is a contagious enthusiasm that may inspire the efforts of men in whom the spirit of a new movement is becoming incorporate, and this may carry them at a bound to heights that orchnary plodding steps would take a long' time to climb. Anyone unaware of the history of the Great Pyramid would certainly judge that it came at the end of a prolonged period in which this type of monument had been gradually improved and enlarged through the centuries. In truth the colossal tomb of King Chufu, though much larger and better executed than any of its numerous fellows, is one of the very first in its kind, indeed the very earliest of all that were planned of the fully developed pyramid shape from the outset. Only two prior examples, and those imperfect pyramids, are known. Again, in the long series of ancient Baby¬ lonian seals the earliest are by common con¬ sent the finest. In Greece the great Ionian temples of about 600 b.c. were in magnitude and perfect workmanship exemplars for all the after periods of the art. The Pantheon at Rome of the time of Hadrian is the earliest and at the same time the grandest of all domes in monumental masonry. What is there to lead up to the western fagade of Chartres—in some respects the most inspired work of the whole Gothic period, but apparently a sudden creation with nothing of the same kind at its back ? In Celtic art of the Christian period there is an illustration of the same phenomenon. Sir E. Maude Thompson writes of ‘ the sudden appearance of highly decorated manuscripts at a certain period without earlier specimens to show anything like a gradual develop¬ ment Where are the earlier works that ^English Illuminated Manuscripts, London, 1895, P- 3- ‘ lead up to ’ the Gospels of Lindisfarne ? In this country there is certainly nothing. In Ireland the Book of Durrow, a masterpiece of the first rank, may be a little earlier, but is itself, like Melchisedec, without a pedigree. In ancient Northumbria Wilfrid’s minster at Hexham, though not the first important stone church in that kingdom, was of the first genera¬ tion, and this building kept up its reputation throughout the Anglo-Saxon period as a land¬ mark in architectual progress, and one of the most notable early churches on this side of the Alps. On the porch of the church at Monk- wearmouth, almost contemporary with Hex¬ ham, there was displayed a life-sized statue in relief in stone,^ an ambitious work to which the whole later Anglo-Saxon period offers hardly a parallel. What was there to lead up to this ? If it be granted then that the carved cross on a monumental scale was a natural outcome of the artistic productivity of the older Nor¬ thumbria, then there is no reason in the nature of things why the most ambitious achievements of this kind should not have been the first. This at any rate is the manner in which in many ages and many modes of production the genius of art has made itself manifest. That the memorial cross was a natural form of artistic expression in Anglo-Saxon England is proved by the very many examples that have come down to us either more or less whole or in fragments. It would occupy too much space to enter upon any discussion of the different uses of these monuments and the varying circumstances in which they were set up. One early record is too significant to be passed over. It occurs in a Life of St Willi¬ bald, the Anglo-Saxon fellow-helper and bio¬ grapher of St. Boniface, written towards the end of the eighth century,^ and is to the effect that when Willibald, who was born in 699, was three years old, he was solemnly conse¬ crated by his parents to the service of God at the foot of the cross which stood near their dwelling, in or about what is now Hampshire. For, the writer explains, ‘ it is customary 1 It was not carved in a single stone, but in five separate stones superimposed. The work has been destroyed but enough remains to attest the former presence of the relief. ^ Printed Acta Sanctorum, Jul. ii, p. 502. 254 REPORT ON THE RUTHWELL CROSS. among Saxon people, on the estates of nobles or gentry, to have for the use of those who make a point of attending daily prayers, not a church, but the sign of the Holy Cross, set up aloft and consecrated to the Lord Such a memorial, marking out a spot con¬ secrated to the new religion but not yet for¬ mally occupied by a church, might well be signalized by the expression used in the inscription on the Bewcastle cross ‘ This Sigbecn ’, this beacon or token of victory, and be regarded as the standard of the advancing forces of the Church, a reference to the Cross as the ‘ sign ’ of Constantine’s triumph being no doubt understood. If the memorial nature of the Bewcastle cross be thus attested in the inscription, the cross at Ruthwell is stamped with the same character by the poem of the Cross incised upon its narrow sides. This poem, like many other appearances of the cross in Anglo-Saxon art and literature, was ulti¬ mately due to the impression produced on Christendom in the year 628 by the announce¬ ment that on the conclusion of a peace with Persia the Emperor Heraclius had recovered the wood of the true Cross. An artistic memento of this event is in the British Museum, in the form of a jewelled pendant of gold found at Wilton in Norfolk, cruciform in shape, and enclosing in the centre an aureus of this Em¬ peror, so set as to bring into special prominence the cross which is figured on the reverse of the coin. The Bewcastle and Ruthwell monuments both glorify the Cross and are inspired with this same idea, but we cannot suppose either of them set up to proclaim the victory of the new religion in a yet heathen locality. In both cases the locality must have been already Christian at the earliest date when the monu¬ ments can be considered possible. At the same time the victory of the romanizing party headed by Wilfrid at the Synod of Whitby in 664 might easily have been celebrated as a new religious triumph and been commemor¬ ated by a cruciform monument. Such a monu¬ ment need not have been set up immediately, though its erection would fall within the period of religious advance in the North to which the energy of Wilfrid largely contributed. It is not without significance that the Northum- bi'ian King Ecgfrith, who was, as we shall presently see, the most likely of all Bernician sovereigns to have had to do with the crosses, is represented in the Northumbrian coinage by a piece exhibiting a cross on a tall stem from which rays of light appear to be stream¬ ing. When we remember that Alcfrith (see antea, p. 242) was one of the chief agents in bringing about the Synod and securing thereby the success of his own personal friend Wilfrid, the appearance of his name on the Bewcastle cross receives fresh importance. Bewcastle is inside the Bernician borders, and within a walk of Hexham, which Wilfrid had made a centre of Romanizing propaganda. It was a Roman station and was accessible by a road that run northwards from the important station Amboglanna, now Birdoswald, on the Roman Wall. Ruthwell on the other hand is not in Northumbria proper but in the British territory of Strathclyde, though this as will presently be seen was at the time of the Synod of Whitby and for twenty years after under the Bernician control. The Christianity of this region, the apostle of which was Kenti- gern, was necessarily of a Celtic type, and some expressions used by Bede when he wrote the closing paragraph of his History seem to show that the people of the district clung somewhat doggedly to their ancient Celtic usages. Hence the erection there of an Anglian monument may have been only a further extension of the proselytizing zeal of the Bernician ecclesiastics. The Question of Date in Relation to the History of Northumbria. This reference to localities carries us forward to the historical question about the extension at different epochs of Anglian or Anglo-Saxon power over regions beyond the normal Teu¬ tonic area. The question here is ;—At what periods in the history of the Anglo-Saxon polity are works of this ambitious character likely to have been undertaken and carried out in these particular places ? If we waive for the moment the considera¬ tion of the crosses from the points of view of religion and of art, and assume them possible at any reasonably admissible historical period, there can be no question at all that the most likely epoch is that of the reigns of King Oswy and King Ecgfrith of Northumbria in the last half of the seventh century. At that time 255 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. and at that time only the effective power of the Northumbrian Kingdom extended to the north and west far and wide beyond the normal Bernician limits, while the kingdom was at the height of its power and, as Hexham showed, of its artistic productivity. Antiquaries may be led b}^ extraneous considerations of one kind or another to prefer a date for the Ruth- well cross in the ninth, tenth, eleventh, or twelfth century, but no one of these, if he be fair-minded, can refuse his assent to what has just been said. It is true that there was a later period when Bernician power was for a time again preponderant in Nithsdale, but the kingdom was then notoriously in a condi¬ tion of decline ; still later the writs of Anglo- Saxon kings ran over the region, but these were not Northumbrian but West Saxon potentates ; and at a still more advanced date, after the Norman Conquest, the district was effectively ruled by a prince famous for his monumental achievements, but he was Anglo-Saxon in nothing but part of his blood and was entirely under the influence of Nor¬ man ecclesiastics. If ever the region about Ruthwell was in any effective sense really Anglian it was in the reigns of the two Nor¬ thumbrian kings just mentioned. A few paragraphs of a purely historical kind are here necessary to explain the relations among the various peoples inhabiting these parts of Great Britain. A convenient starting point in time is about the year 600. Ruthwell is situated in the valley of the Nith that runs into the sea below Dumfries, and belonged politically to the British kingdom of Alcluyd, called later on Strathclyde and Cumbria, which stretched from the Clyde, where was its capital Alcluyd, now Dunbarton, to the Derwent in Cumberland, the old south¬ ern limit of the diocese of Carlisle. Farther to the north and west lay the kingdom of the Scots of Dalriada, to the north and east that of the Piets, while to the south-west what is now Galloway was inhabited by a people whose ethnology is somewhat obscure. On the east was the formidable and increasing power of the Angles of Northumbria. The first great Northumbrian conqueror .Ethelfrith, at the end of the sixth century, is stated by Bede to have ‘ conquered from the Britons more territory than any other chief or king, either subduing the inhabitants and making them tributary, or driving them out and planting the English in their room A counter-offensive was engineered against him from the north in 603 by Aidan the powerful king of the Dalriad Scots, who in concurrence no doubt with the Britons of the kingdom of Alcluyd marched with immense forces against iEthelfrith, but suffered at his hands a crushing defeat at a place called by Bede - ‘ Degsastan ’, identified with some plausibility with Dawston in Liddesdale. Since that time, the historian adds, no king of the Scots had dared to come out to battle against the English. Professor Oman believes that ‘ the victory of Dawston must have confirmed ^Ethelfrith in the possession of the upper waters of the Tweed and its tributaries, as far as the water¬ shed of the Clyde, and, no doubt, of the land round Carlisle and the west end of the Roman Wall also ’. ® The sceptre of Northumbria was next wielded by a hand more powerful still than iEthelfrith’s, and Bede says that King Edwin reduced under his dominion all parts of the country, English and British alike.^ Edwin’s power however went down before a coalition between the heathen king Penda of Mercia and the Christian Cadwallon, the British lord of North Wales. The whole of the North was ravaged, but Northumbrian power in its northern seat was soon after re¬ established by the memorable victory of King Oswald at Heavenfield, a little beyond the Roman Wall four or five miles north of Hex¬ ham. This was in 636 a.d. Of Oswald Bede goes so far as to say that he ‘ brought under his rule all the nations and provinces of Britain, divided as they are into the four tongues, Bri¬ tons, Piets, Scots, and English This implies a great extension of Northumbrian power to the north, for southwards it was always con¬ fined by that of Penda of Mercia, who ulti¬ mately brought Oswald to defeat and death. Once again the Mercian king held Northum¬ bria in his grasp and harried it up to the very walls of Bamborough, but Oswy, who was the next effective Northumbrian monarch, seems to have established some modus vivendi with ' Hist. EccL, i, 34. ^ Hist. Eccl., i, 34. ^ England before the Norman Conquest, London, 1910, p. 251. ^ Hist. Eccl., ii, 9. Hist. Eccl., iii, 6. 256 REPORT ON THE RUTHWELL CROSS Fig. 167. 257. 17 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. the formidable champion of heathenism, for he gave his daughter to Peada Penda’s eldest son, whose coins have his name in runes, and married his own son Alcfrith to that Cuni- burga, daughter of the Mercian king, whose name is inscribed on the Bewcastle cross. These matrimonial arrangements did not how¬ ever eliminate strife, and this time when a decisive action was brought about Penda fell, and for a while Osw}' remained in a position of practical supremacy in the country. After about 660 however the Mercian Wulfhere, a son of Penda, cut short Oswy’s domination in central England but left him free to extend it unchecked towards the north, where, Bede tells us, ‘ Pictorum quoque et Scottorum gentes, quae septentrionales Brittaniae fines tenent, maxima ex parte perdomuit, ac tribu- tarias fecit ‘To Osviu ’, writes Professor Hume Brown,‘ Bede ascribes even greater power than to ^Ethelfrith, or Edwin, or Oswald. In his later years, he * * * was virtual master of Dalriada, Strathclyde, and parts of the land of the Piets * * * Over all three kingdoms * * * Northumbria for nearly thirty years must have exercised an effective suzerainty ’. Of Wilfrid Bede tells us ® that he was ‘ bishop of all the Northumbrians, and likewise of the Piets, so far as King Oswy was able to extend his dominions ’, and Eddius in his Life of M'ilfrid ^ writes of him at a later date as exercising ecclesiastical authority over Britons, Scots, and Piets. This was in the time of Oswy’s son and successor Ecgfrith, who early in his reign suppressed by force of arms a rising of the Piets against his rule,® and subsequently proved how complete was his command of the western side of the country by dispatching in 684 A.D. a naval force against Ireland. The next year however witnessed the disaster which began the ruin of the fortunes of Northumbria, when Ecgfrith and his whole army which he had led into the north were cut off by the Piets on the fatal field of Nectansmere. The result, according to Bede, ® was that from ^ Hist. Reel., ii, 5. * History of Seotland, i, 17. ® Hist. Reel., iv, 3. ^ Historians of the Chureh of York, Rolls Series 71/1. P- 31- “ Eddius, Vita Wilfridi, ubi supra, c. xix. ® Hist. Reel., iv, 26. that moment the hopes and the strength of Northumbria began to fail, the Piets resumed possession of the territories the Angles had taken from them, while the Dalriad Scots, and also ‘ certain of the Britons ’, ‘ Brettonum quo¬ que pars nonnulla ’, recovered their liberty. This last clause obviously refers to parts of Strathclyde, which Professor Oman thinks may have actually been annexed by Oswy to Ber- nicia,’^ and which was certainly during this latter half of the seventh century, up to 685, completely subject to Northumbrian influ¬ ence. How entirely Bernician at the time was Carlisle is shown by the fact that Ecgfrith had evidently made it his base of operations against the Piets of the North, and had left his Queen there in a newly founded Anglian monastery to await his return. Cuthbert was present at the place and received there a mystic intimation of the disaster at the moment he was looking down into a ruined Roman well.^ Carlisle was in his diocese and we are told that he came there to consecrate priests.® The Northumbria of Oswy and of Ecgfrith, if we include in it the territories which Bede describes as ‘ tributary to ’, ‘in the power of ’, or ‘ appertaining to Bernicia, covers the vast extent of country indicated in the map, fig. 167. This is based on the map prepared by Mr. W. J. Corbett for the Cam- bridge Mediaeval History^ but embraces on the direct evidence of Bede portions of the Pictish lands. How far north of the Forth there was any effective Anglian suzerainty cannot be said, but the revolt of the northern Piets against Ecgfrith early in his reign shows that suzerainty was claimed. The kingdom of Alcluyd as a whole was certainly not in¬ cluded and Clydesdale no doubt still obeyed the lords of the old fort of the Britons, while the more southerly portion of the Strathclyde kingdom, Annandale and Nithsdale, was Northumbrian, as was also the whole of the western coast of England between the Solway 1 Rngland Before the Norman Conquest, p. 295. ^ Beda, Vita S. Cudbereti, c. xxvii. ^ ibid., c. xxviii. * Hist. Reel., ii, 5 ; iii, 6 ; iii, 4. = Cambridge University Press, 1913, Map No. 17. The map is headed ‘ England circa a.d. 700 ’ but this is, of course, fifteen years too late, as after Nectans¬ mere the recovery of liberty by the Scots and certain of the Britons meant a curtailment of limits towards Strathclyde. REPORT ON THE RUTHWELL CROSS. and the Ribble. Part of the region of Gallo¬ way beyond fhe Nith we are expressly told by Bede appertained to Bernicia.^ This part was that about Candida Casa or Whithorn which had been the centre of the evangelizing activity of Ninian fhree hundred years before, and Chrisfianity in a Brifish form may have survived there the conquest of Bernicia by fhe pagan Angles. We find it at any rate a little later than the time at which we have arrived the seat of an Anglian bishopric, and Bede menfions among the latest facts in his Ecclesi¬ astical History the establishment there of Bishop Pecfhelm who was consecrated in 730 a.d .2 This proves of course that even after Nectansmere Carlisle must have re¬ mained in Northumbrian possession, and from Carlisle, fhrough fhe sea-power of which Ecgfrifh’s earlier Irish expedifion furnishes evidence,® a hold can have been kepf on fhe maritime peninsulas of Galloway. This Anglian see remained in existence at Whithorn till the end of the eighth century, after which the growing weakness of Norfhumbria caused its withdrawal. As Bede informs us (i) fhat after the defeat and death of Ecgfrifh ‘ some of the Britons regained their liberty and (2) that when he closed his history the Britons were in part their own masters but in part under subjection to the Angles ® it may be a matter for quesfion how far we can assume any survival of Anglian aufhorify in Sfrath- clyde norfh of the Solway. The subject Britons may be those about Carlisle, and southwards. Bede lived through the reigns of several of fhe successors of Ecgfrith and closes his history with some rather gloomy forebodings as fo fhe fufure of his counfry. So long as he was keeping record, he gives us to understand that there was no recovery of fhe losses which had followed Nectansmere,® but there was some revival under King Eadberct, 737-758, who in conjunction with Angus MacFergus the powerful king of the Piets took Alcluyd and recovered in a measure the overlordship of '■ Hist. Eccl., iii, 4. - ibid. ’ antea, p. 258. * antea, p. 258. ‘ Hist. Eccl., V, 23. ® ibid., iv. 26. Bede died in 735 but he ended his history in 731, forty-six years after the disaster. Strathclyde, which Professor Oman fhinks may have remained in vassalage to North¬ umbria till the latter part of fhe eighfh cen- fury.^ The succession of the native kings of Alcluyd sfill however confinued, even after the conquest of fhe Piets in 844 by the Dalriad Scots under Kenneth MacAlpin. In the time of Kennefh’s vigorous successor Constanfin III, 900-942, the native line of Cumbrian princes failed, and kings of MacAlpin’s family bore rule over Strafhelyde, until its quasi-independent sovereignty was merged in that of fhe practical lord of all fhat is now Scotland, Malcolm II, 1005-1034. Malcolm’s great achievement was the crush¬ ing defeat he inflicted on the Northumbrians in 1018 at Carham near Kelso, the result of which was the final cession of fhe Lofhians from fhe Tweed northwards, which became then as it has ever since remained a part of Scofland. This was the completion of fhe long process of Bernician recession soufhwards which illusfrates the gradual decay of North¬ umbrian power. After Eegfrith’s disaster in 685 the kingdom showed none of fhe recupera- five power exhibited more than once in the seventh century, and its history from Bede’s fime onwards is one of internal decay and outward weakening. Disputed successions, and the murders or forcible encloistermenfs fhat are their sequels and that remind us of earlier Merovingian crimes and disorders, broughf the realm to a miserable pass, and it is noticed by bishop Stubbs that of eighf Norfhumbrian kings between 737 and 796 not one died a natural death while in posses¬ sion of fhe sovereignty. The practical ex¬ tinction of Norfhumbrian power and culture, was brought about however in the succeeding century at the hands of fhe Danes, and fhere musf be noted here the introduction in the later history of fhe region of two new factors, one the appearance and operations of fhe Danes, and the other the recrudescence of Anglo-Saxon power in the North in the tenth century, though not the power of fhe Angles but that of the now predominant rulers of Wessex. It was not till 867 that the full force of fhe Viking onslaught was felt in England, and the blow was directed against the North. ‘ The 259 * En^and, etc., p. 334. HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. doomed realm was Northumbria once the suzerain state of all England, but long a byword for its insane and never-ending civil strife In 867 the ‘ Great Army as it was called, of the heathen invaders captured York and destro^'ed a Northumbrian levy led by two rival kings, making itself master of the southern Northumbrian province of Deira. North of the Tyne there appear to have sur- \'ived native princes who were however at the mercy of the conquerors. A few years later part of the ‘ Great Army ’ set to work in earnest to ravage the northern province. Its leader was Halfdene, and the Anglo-Saxon Chronicle, ad. ann. 875, tells us that he ‘ went with a part of his army into Northumbria, and took winter-quarters by the river Tyne ; and the army subdued the land, and often harried on the Piets and on the Strathclyde Welsh ’, while Symeon of Durham adds that a cruel desolation of the Northumbrian province followed, while the army of the Danes was everywhere raging with savage fury. On all sides monasteries and churches were given to the flames, and to cut the story short, from the eastern sea across to the western there was continuous slaughter and fire ‘ All mon¬ asteries were destroyed is said in another place.^ After harrying the province Halfdene pro¬ ceeded to settle it. The Chronicle, ad ann. 8 y 6 , says that he divided the Northumbrian lands and arranged for their tillage, no doubt by the native peasants under Danes to whom the estates had been granted. The whole country became in a sense Danish, though in the northern province of Bernicia English life survived to a greater extent than in Deira. The invaders were heathen, and Christian buildings and m^onuments were burnt or shattered. Symeon of Durham draws a picture of the bare walls of Jarrow monastery standing without a roof,^ and those of the church at IMonkwearmouth half in ruin with trees and shrubs filling the whole interior ®—a memorial of the destruction of this time that lasted till the Norman Conquest. Wilfrid’s monastery 1 Oman, p. 435. ^ Hist. Dunelm. Eccl., ii, c. 6, in Symeon of Durham, Rolls Series, 75/1, p. 58. * Hist. Regum., ibid., 75/2, p. no. * ibid., Hist. Dunelm. Eccl., iii, 21. ® ibid., c. 22. at Hexham was burnt.^ The monks of Lin- disfarne abandoned their cloister in despair and bore away the body of St. Cuthbert on their famous wanderings. The question of the likelihood of the sur¬ vival of destructible Christian monuments after this pagan inroad is one on which opinions may differ. Heathenism did not remain ram¬ pant, for we are told of Halfdene’s successor that he was a Christian,^ but the raid of 875 was no doubt fatal to many ecclesiastical monuments of value. As every student of monuments however knows, the word ‘ de¬ struction ’ is used by mediaeval writers with considerable looseness, and it was not every Christian object that came under the notice of the Vikings, nor in the case of every monu¬ ment actually attacked was the destruction complete.^ The case of Monkwearmouth is typical. Here we know the ruin was great, but the well-known western porch is certainlj^ a relic of pre-Danish days and has not been destroyed. The life-sized figure relief in stone below the gable, that has already been men¬ tioned,^ was perhaps hacked to pieces and was afterwards for neatness’ sake hewn away as it is now flat to the ground, but on the lower part of the porch there is delicate decorative work that shows no sign of having been wantonlv injured. There is no mark of Danish hands on the Bewcastle and Ruth- well crosses, unless it were they that knocked off the head of the former, but as we are told that the Strathclyde Welsh were harried the Ruthwell monument may have lain in the road of the ravaging bands. It is probable that the question. Could the crosses have sur- ^ ^Ired of Rievaux, in Canon Raine’s Priory of Hexham, Surtees Society, No. 44, p. 190, says of this destruction ‘ quidquid de lignis fuerat, ignis absump- sit ’. The church is described by him (p. 191) as appearing subsequently to the Conquest in much the same condition as those of Jarrow and Monkwear¬ mouth. 2 Hist. Dunelm. Eccl., ii, 13. ^ The Chronicle of Lanercost states distinctly that Wilfrid’s church at Hexham remained till the Scottish incursion of 1296, a.nd it is described in the twelfth century by Prior Richard of Hexham. We are ex¬ pressly told by Symeon of Durham of a stone cross made by Bishop ^thelwold of Lindisfarne early in the eighth century, that when the Danes ravaged Holy Island they broke off the head of the cross, which was, however, afterwards joined on again by being run with lead. ^ antea, p. 254. 260 REPORT ON THE RUTHWELL CROSS. vived if they had been in existence at the time ? will be answered by each one with uncon¬ scious reference to the opinion he has pre¬ viously formed as to the date of the monu¬ ments. It comes to a balance of chances. The monuments may have been in existence and by good fortune have escaped, while it is doubtless more likely that if they were standing in 875 one or other of them would have attracted hostile notice. Independently of this Danish argument, the facts about Northumbrian history now detailed afford some solid ground for conjecture as to possible or likely dates. In the time of Oswy and Ecgfrith Northumbria was great, and the site of the Ruthwell cross was entirely within its control. After that epoch for a good part of the eighth century that part of Strathclyde nearest to the Solway would be within the Bernician sphere of influence, but the age was not one for great achievements in any depart¬ ment of activity. This does not mean that artistic production in the eighth and ninth centuries was at a standstill. The number of carved stones that exist to this day in different parts of the ancient Northumbria is so great, that we must postulate considerable and long- continued industry in this kind, and it is interesting to note that Scandinavian motives appear on not a few of them, testifying to the permanent influence of the Danish immigrants. It needs hardly to be said that there are very few indeed of these numerous carved stones that show any approach to the artistic qualities of the work on the Ruthwell and Bewcastle crosses. What are the relations between these two monuments and all the others is a ques¬ tion that has never yet been systematically treated. From the historical point of view the eighth century would be possible for the crosses but hardly the ninth, for by the end of the former all the outlying possessions of Northumbria had been lost,^ and in the middle of the latter the line of Anglian kings of North¬ umbria comes to an end, and they are succeeded by a Danish dynasty. When we pass to the tenth century the second of the two new factors referred to above comes into play, the great and aggressive power of Wessex. Symeon of Durham says that after the death of Guthred, the Christian successor of Half- dene, King Alfred of Wessex assumed the government of Northumbria.^ It was how¬ ever Alfred’s son and successor, Edward the Elder, who exercised in person his power in the North, though exactly what was the form of his dealings with it is a matter of doubt. With the advent of his son, the mighty .®thel- stan, the supremacy of Wessex in this region was asserted by force of arms, even to the extent of a formal annexation of the Danish kingdom of Northumbria. Of his youthful brother Edmund who succeeded iEthelstan the Chronicle records that after harrying Cumberland he gave it over under conditions to Malcolm I of Scotland. This was in 945. The district in question was not the ancient kingdom of Cumbria or Strath¬ clyde as a whole, but rather the Scandinavian parts of the modern Cumberland, where the Norsemen have left such clear traces of them¬ selves in the local nomenclature. The next king of Wessex, Eadred, confirmed the power of the now consolidated English kingdom over the partly Danish Northumbria, and it re¬ mained an integral part of the united realm under the rule of Eadgar. By this time however the whole of the ancient Strathclyde was in the hands of the Scots, and Carlisle with the country to the north of it remained under Scottish control through the rest of the Anglo-Saxon period and down to the reigns of William Rufus and Henry I. The former of these possessed him¬ self of the place by force of arms and fortified it with a mound-and-bailey castle on the site of the present later Norman stronghold, while the latter founded there a bishopric. In the troubles of the time of Stephen however Carlisle passed again into the power of the Scots where it remained all through the memorable reign of David I, 1124-1153, who before his accession had administered this part of Scotland as earl of Cumbria. Under David th^ limits of the northern kingdom were extended at the expense of England and ‘ throughout the entire reign of Stephen, the Eden and the Tees ’, rather than the Solway and the Tweed, ‘ were the boundaries of the two countries After David’s death at 1 Oman, p. 334. 261 ^ Hist. Dunelm. Eccl., ii, 14. * Hume Brown, History of Scotland, I, 66. HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. Carlisle in 1153 Henry II of England restored the older limits. Both the West Saxon kings and David of Scotland controlled the sites of the crosses and possessed the power and the natural inclination which would have made the setting up of ‘ signs of victory ’ or of religious monu¬ ments of anj' kind a congenial task. David of Scotland’s participation in the work is however necessarily excluded. As has been seen alread}’ and will be made more clear in the sequel, the runic inscriptions on the crosses, alike in their form and in their content, supply convincing evidence that they are of pre-Conquest date, and it is only by ignoring or misinterpreting these that any plausible plea for a twelfth-century date can be formu¬ lated. The tenth century undoubtedly offers more possibilities. It was a great artistic period in which were produced some of the masterpieces of Anglo-Saxon craftsmanship, and was an age of high aims and correspond¬ ing achievement. If it were not for the Anglian character of the monuments that have so many more parallels in the North than in the South, and for the Northumbrian lan¬ guage and lettering of the inscriptions, the theory of a tenth-century date would have a good deal in its favour, though there would still be other difficulties in the way of it. There is one possibility here that in a Report like the present cannot be ignored. An argument not without some plausibility might be started on the basis of the Anglo- Saxon translation of Bede ascribed to King Alfred the Great. ^ This implies a keen interest in the heroic age of the North felt at the time in West Saxon circles, and its promulgation would undoubtedly extend and intensify that interest among the leaders of thought and action. The northern writers claimed that Cuthbert appeared in a vision to Alfred in the marshes of Athelney and promised him victory, and that Alfred took over Northumbria after the death of Guthred the Christian successor of Halfdene.^ iEthel- stan showed his reverence for the northern saint in conspicuous fashion by his magnificent gifts to the shrine of St. Cuthbert at Durham, amongst which are the world-famous em¬ broidered stole and maniple.^ It might be argued that HDthelstan, or a like-minded prince of his house, gave expression to the reverential feeling of the day for the age of the North¬ umbrian saints and heroes by erecting monu¬ ments in a semi-antiquarian spirit in honour of some of these departed worthies, on which their names might be inscribed. The frag¬ mentary inscribed stones at Collingham and Hackness in Yorkshire might be worked in to support such a theory. Antique diction and lettering might be consciously used in accordance with the memorializing idea. This theory of a date in the tenth century, though it might be seriously urged, would have still against it the arguments (i) that the figure sculpture would be hardly possible at a date so long after the Early Christian period, and (2) that the absence of acanthus foliage in a post-Carolingian monument would be extraordinary. The foliage in the Bene- dictional of .®thelwold and on the embroidered stole at Durham, both works of the tenth century, is acanthus. The Question of Foreign Workmen. From the historical point of view it is clear that the two crosses would find a more natural location in the seventh century than at any succeeding period, and the question must now be asked whether on artistic grounds such a date be reasonably possible. It must be repeated that the work in question is in some ways so excellent that its appearance at any time in Anglo-Saxon England is some¬ thing like an artistic miracle, and we have really to seek for the epoch when it would be, not most likely, but, least surprising. The co-operation of foreign artificers is possible in an outward sense at any conceiv¬ able period for the crosses, but the presence of such is only to be looked for at periods when the country was safe and prosperous. This was the case in the Northumbria of the seventh and early eighth century, and also in the tenth and first half of the eleventh, when England had the reputation of being a wealthy country. Prior to the Conquest we are told by William of Poitiers that Germans ^ Plummer’s BesdcB Opera Hisiorica, I, cxxviii. ^ Symeon of Durham, Rolls Series, 75/1, p. 211. - Symeon of Durham, Hist. Dunelm. ecclesiae., ii, The date of the gift was about 934 on the occasion of 10, 14, Rolls Series, 75/1. one of .Rthelstan’s military expeditions to the North. 262 REPORT ON THE RUTHWELL CROSS. versed in all the arts were accustomed to settle among the English, who, he says in the same sentence, were themselves, men and women alike, expert in the various arts.^ The hypothesis however that continental craftsmen were responsible for the fabrication of the crosses, though many have favoured it, becomes more difficult of acceptance the more critically it is examined. There is, it is true, the direct evidence of Bede that when Benedict Biscop was building his monastery at Monkwearmouth in 674 he sent over to Gaul for masons to erect a stone church after the Roman manner, and also for workers in glass, a material that no one at that time in Britain could manipulate. The contemporary life of Wilfrid by his companion Eddius is good authority for the fact that this energetic prelate in his journeys on ecclesiastical busi¬ ness took with him stone masons and men skilled in almost every art,® but nothing is said in the passage about the presence among these of foreigners. In the case of the song- master iEdde or Eddius (the writer of the Life) mentioned in the passage in note 3, Bede tells us * that Wilfrid invited him from Kent, which would suggest that the other craftsmen were also of Kentish origin. There may of course have been Franks among them, as Wilfrid was in close touch with some of the Gallic clergy, but there is no necessity for assuming this.® As was noticed above, the presence on the crosses of inscriptions in Anglo-Frisian runes is almost destructive of the hypothesis under consideration, unless it can be saved by assum¬ ing that foreign—Gallic, Italian, or Greek— ’ Gill. Piet., apud Duchesne, Hist. Norm. Script. Ant., p. 211, B. ‘ Anglicse nationis femina3 multum acu et auri textura egregie viri in omni valent artificio. Ad hoc incolere apud eos Germani solebant totium artium scientissimi “ Historia Abbatum, c. 5. ’ Vita Wilfridi Episcopi, in Historians of the Church of York, Rolls Series, 71/1, p. 22, ‘ episcopalia officia per plura spatia agens, cum cantoribus .Edde et Eonan, et caementariis, omnisque paene artis insti- oribus * * * ’. * Hist. Eccl., iv, 2. ^ Writers of the twelfth century speak of Wilfrid importing craftsmen from overseas, even from Rome, but there is nothing about this in the contemporary authorities. In the twelfth century it was the fashion of the times to regard Rome as the universal provider of all things ecclesiastical, even of the Orders of a purely Celtic missionary saint like Kentigern. carvers executed the figure sculpture and called in local workmen to cut the runic inscriptions. But why, it may be asked, were there runes at all, when the classically trained artists could have expressed all that was required in Latin characters ? The inter¬ penetration on the Ruthwell cioss of the runic and Latin modes of writing is against any idea of such a separation between the two sets of inscriptions. It must be noticed too that the Latin characters are not of Italian or Gallic type but Hiberno-Saxon. This is another difficulty in the way of the ‘ foreign workman ’. The motive of the vine scroll with animals is common to Christendom at large, but the pure foliage panels on the Bew- castle cross have no parallel in continental lands, and if they were devised by foreign workmen represent a ipiite original effort in design not on the lines of any traditional continental style. The Question of Anglo-Saxon Authorship. It is always of course a reasonable assump¬ tion that any immovable monument of local material is also of local execution, that is to say has been wrought by workmen of the district or at any rate the kingdom or political area wherein it is found. In a few concluding paragraphs this assumption may be tested in the present case, and the question discussed whether it is within reason to postulate for the crosses not only Anglian design but also Anglian workmanship. The popular prejudice against crediting our Anglo-Saxon forefathers with any artistic gifts need not stand in the way. Their racial character is supposed to be of a rather stolid and heavy order and most people would doubt if they had it in them at any time to achieve success in the arts. Happily however this is not a matter for a priori argument, for works which are undoubtedly the product of Anglo-Saxon craftsmanship exist in sufficient numbers to prove the popular view erroneous. Anglo-Saxon tombs have given to the light innumerable objects of personal adornment that arc so different from objects of a similar kind found on the Continent that their native origin is fully established. The works in question, brooches, buckles, pendants and the 263 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. like, date from all periods of the sixth and from the first half of the seventh century, and in their design and character they re¬ present a tradition of Germanic art which left its impress on all the lands of the West. As regards design, from the purely artistic point of view they are of no outstanding merit, and they decline in aesthetic value as the sixth century passes on to the seventh. Their strong point is their technique, and in this the Anglo-Saxon specimens are of special excellence. In the casting and chasing of small objects in bi'onze, in fine gold work, and especially in the exact cutting and set¬ ting of semi-precious stones, the Anglo-Saxon craftsman rivalled the best of his continental contemporaries, and is practically the equal of the average practitioner in the greater ages of the industrial arts. This kind of work represented in the pagan Anglo-Saxon ceme¬ teries declines, as has just been said, in aesthetic value as the seventh century ad¬ vances, but before it becomes artistically extinct a new form of decorative art has already made its appearance. In this form of art precision of technique is of slight account, but the qualities of design exhibited in it are of remarkable excellence. The re¬ ference is to the early set of Anglo-Saxon coins known as sceattas, that may begin in the last part of the sixth century and that flourish through the seventh and a considerable part of the eighth. It is impossible to discuss the Anglo-Saxon capacity for art without taking due account of these early coins. They are on a minute scale, averaging about three - quarters the diameter of a threepenny piece, but in the matter of design the coin artists seem simply to overflow with ideas, and they give rein to their fancy in copious and novel devices, often of a most quaint and spirited kind. In the British Museum Catalogue of the Anglo-Saxon series in the national collection it is said of them that the sceattas are ‘ rich, as few coinages of the world are rich, in the variety of the designs by which they are adorned ’. These characteristics appear in full relief when we compare the sceattas with the con¬ temporary, or rather earlier, Frankish coins of about the same size known as ‘ trientes ’. These are neat in execution and more scholarly than the Anglo-Saxon pieces in their rendering of Latin lettering, but in design, with very few exceptions, they are tame and mono¬ tonous to the last degree. These Anglo-Saxon coins must of course be taken for what they are—sketchy studies on a minute scale with no pretentions to exactness of detail or to orthodox figure drawing. Their importance resides in the vivacity and varied character of their designs, and the boldness with which new motives such as foliage sprays are introduced and pressed into the service of the composition. It should be said that the coins, though Anglo-Saxon, are not in their origin Anglian, but belong to the southern and midland districts. The corresponding early coins of Northumbria, from Ecgfrith onwards, the so-called stycas, have not nearly so much merit in design as the sceattas of the more southerly parts of the country, and this may reduce the significance of the coins as evidence bearing on the Northumbrian origin of the work on the crosses. We have here at any rate two forms of art prior to the eighth century that are un¬ doubtedly Anglo-Saxon, and taken together they show that at this early period produc¬ tions of high technical quality and of a vivacity and inventiveness in design that are quite extraordinary could be turned out of Anglo- Saxon workshops. It is noteworthy that the two phases of art are quite independent. The ornamental motives used in each have hardly anything in common, and no one examining a collection of sceattas in conjunction with a set of artistic objects from Anglo-Saxon tombs of the early seventh century would think for a moment that they correspond in date and in locality. Anglo-Saxon craftsmen seem as it were to have followed separate paths each keeping to the tradition of his own particular art, and not all drawing together from a common stock of motives. This consideration may explain the extraordinary phenomenon that we meet with at Lindisfarne at the close of the seventh century. Two important woi'ks of art were at that time and place in progress, executed by members of the same community, for the same purpose of honouring the deceased hero-saint Cuthbert, but in different materials and on distinct artistic traditions. One work was the enriched wooden coffin that was to hold the body of the saint, the other the Book of Gospels that was written and illuminated 264 REPORT ON THE RUTHWELL CROSS. for his glorification. On the first there is incised figure work almost childish in its homely crudity, in the latter linear and conventional ornament drawn and painted with a skill and finish in design and execution that have never been surpassed. Were not these two works both specifically dated and fixed to a known locality it would never have occurred to any¬ one to imagine any connection between them, and even in face of the known facts of the situation it is hard to realize that they were practically contemporary and were carried out within the same walls. This apparent paradox is of course explained by some on the supposition that the wooden coffin is Saxon work, but the illuminated Gos¬ pels are the production of Celtic calligraph- ists, and on this point a word must be said. The illuminated manuscript known as the Gospels of Lindisfarne or of St. Cuthbert, or as the Durham Book, is the finest example in existence, save one, of the so-called Irish style in calligraphy, and is surpassed by the Book of Kells alone. The style of writing and of ornamentation is characteristically Celtic, though there is one peculiarity, the abundant use in the enrichment of the bird motive, that is rather Anglo-Saxon than Irish. There exists however at the end of it a note by a scribe of the tenth century to the effect that it was written by Eadfrith, bishop of Lindisfarne, who is known to have filled the see from 6q8 to 721. The writer of the note was one Aldred a priest, and is very possibly to be identified with an ecclesiastic of the same name who was consecrated to that bishopric, then located at Chester-le-Street, about the middle of the tenth century. He wrote the interlinear Anglo-Saxon translation of the Latin text of the Gospels which is inscribed in the manuscript. The most important item of information imparted by Aldred is contained in the words ‘ Eadfrith bishop of the church of Lin¬ disfarne he wrote this book ’} the word ‘ Professor Lindsay, Early Irish Minuscule Script, Oxford, iQio, p. 2, note, thinks that Eadfrith pro¬ bably executed the work before the year 698 when he became bishop. The fact should not be passed over that in the recently published elaborate work on pre - Carolingian manu.scripts [Vorkarolingische Miniatitren, herausgegeben von E. Heinrich Zimmer- mann, Berlin, 1916), the author on the basis of his comparative studies accepts as established the early date of the Lindisfarne book, and believes also in a seventh centurj’ date for the two crosses. ‘ wrote ’ being the normal past tense of the familiar verb ‘ writan ’ to write. Now a quite plausible comment on this would be that Eadfrith may have indited the actual words of the text, but that its artistic embel¬ lishment is quite another matter and was probably the work of other hands. To this the answer is easy. No one can turn over the pages of the manuscript, which is one of the choicest treasures of the British Museum, without assuring himself that text and orna¬ ment interpenetrate so intimately that they must necessarily be the work of the same hand or hands. In later mediaeval manuscripts of another class theie was sometimes a division of labour, and a worker might indite a page of text and then hand it over to an ornamentalist to put in the enriched capitals and borders. Nothing of the kind is possible in the early manuscripts of the Celtic school. Writer and ornamentalist were one, and in Irish and Anglo-Saxon literature the one word ‘ scribe ’ is used as the title of the executant. Of a calligraphist named Ultan we are informed that he was ‘ a blessed priest of the Scotie nation, who could adorn little books with elegant designs. ... In this art no modern scribe could rival him.’ Ultan was a member of this very community of Lindisfarne,and flourished about 740.- Whether the single executant of the Durham Book was Eadfrith himself, or someone like Ultan employed by him, is another question that might naturally be raised. It would be rash to maintain that a word like ‘ fecit ’, ‘ made ’, implies always in mediaeval literature the personal agency of the person named, but other distinctive forms of expression were certainly in use when the person named only ordered and paid for the work. Oueen iElflaed ‘ fieri precepit ’ in the case of the Durham stole and maniple of the early tenth century, and king Alfred does not claim to have ‘ made ’ his jewel, but only ‘ ordered it to be made ’. In the case before us there * The words are quoted from an Anglo-Saxon poem of the ninth century published in Mabillon, Acta SS. Ord Bened., IV, par. ii, 317-335. In a private communication to the writer Sir Edward Sullivan kindly stated that he had no doubt that in connection with Irish MSS., scribe and illuminator were one. Colgan, Acta Sanctorum Hiberniee, Louvain, 1O45, i, 109. 265 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. is the strongest internal evidence that when Aldred in the tenth century used the word ‘ wrote ’ he meant that Eadfrith was the actual executant. He goes on in his note to say that ^Ethelwold, the next bishop, bound the book, and adds the important words ‘ as he weU could clearly implying that he put his own episcopal hand to the task, and this undoubtedly reflects back a presumption of personal knowledge and skill on the part of *Ethelwold‘s predecessor. Furthermore, Al¬ dred repeats at the end of the note the names of those who, as he says, ‘ fabricated and adorned ’ the book, Eadfrith, iEthelwold, Bil- frith the Anchorite who made a jewelled cover for it, and lastly he himself who added the interlinear Anglo-Saxon gloss. No other name is even hinted at. If this evidence, which is undoubtedly very strong indeed, be credited, then in this Anglo- Saxon Bishop of the end of the seventh century we have an artist who in design and execution is fully equal to the most accomplished masters in manuscript illumination of whom there is record. The style of the work is Celtic rather than Anglo-Saxon, and so far it may be called in Northumbria an exotic style, though we must remember on the other hand that it had been taught and practised at Lindisfarne for more than half a century, so that in that particular district it may almost be regarded as vernacular. In any case, whoever may have been ultimately responsible for the work, the fact remains that there is most cogent evidence that it was executed at Lindisfarne in honour of St. Cuthbert a few years after his death in 687, and this is quite enough to show that this epoch in Northumbria was one of great artistic achievements, when exceptional tasks could be undertaken and carried through. There exists also evidence to show that a combination of fine classical and of barbaric elements in a single work was elsewhere possible at the time. The reference is to the remarkable work of art in the Museum at York known as the Ormside bowl. It is made up of several pieces differing markedly * The words are ‘ githryde & gibelde sva he vel cuthjE ’. Of these the first is the past tense of the verb ‘ gethrythan ’, which means to ‘ press ’, ‘ make firm and hence ‘ bind ’, while the second v/ord ‘ gibelde ’ is obscure, but it certainly does not mean ‘ embellished ’. in their ornamental enrichment, but a close examination of the piece as regards both de¬ sign and construction shows that it is a real artistic unity worked out on a single scheme. This scheme is not classical but Germanic, and some of the component parts are ornamented in a typically Germanic fashion, while at the same time there are other parts, and these the most important, on which the work is in a characteristic late classical style that would be at home in Alexandria or in Syria. It has been suggested ^ that the piece is of Anglian origin and Anglian design, but that the Eng¬ lish artist responsible for the latter employed a foreign craftsman, perhaps from Syria or Asia Minor and a companion of the Greek archbishop Theodore, to execute in the classical style the main part of the ornamenta¬ tion. The Germanic portions of the composite piece are however in style not Anglian but Frankish, and it was most probably executed in the Rhineland or in Northern Gaul. If this be accepted as its place of origin it re¬ mains almost equally instructive for the present purpose, for it shows how Germanic and classical elements may be combined in a work of the period, for the style of the Ger¬ manic work shows that it is of the seventh century a.d. The classical ornamentation bears some resemblance to the foliage and animal work iipon the two crosses, and, like it, is distinctly non-barbaric. The consideration of these two seventh- century monuments makes less startling the hypothesis of Anglian authorship for the Ruthwell and Bewcastle crosses. This Anglian authorship might mean, as in the case just dis¬ cussed, design and artistic supervision of the whole of a work, and the actual execution of parts of it, with the relegation of other portions to foreign employees. Or it might mean An¬ glian execution as well as Anglian design for the whole. It may be asked. If the Saxon Eadfrith achieved such a marked success in a style that was not Teutonic, might not Northumbrian carvers, inspired with the large and daring spirit of the time, have used foreign models in relief sculpture to something of the same effect ? It may be freely admitted that the hypothesis is unlikely, but then the monu- ’ By Mr. W. G. Collingwood, in the Cumberland and Westmorland Society’s Tvansactions, vol. XV, p. 381. 266 REPORT ON THE RUTH WELL CROSS ments in themselves are unlikely, so unlikely indeed, that were their mere existence and character known and nothing else about them, it would be declared beyond the bounds of possibility that they could have come into being in the seventh or in any other Saxon century in northern Britain. The monuments however are there, though any hypothesis offered for their explanation must be a forced one, for they are certainly not what would naturally be expected in that region and in the Anglo-Saxon period. In favour of an Anglian origin are some con¬ siderations of weight. The runes and the Anglo-Saxon poem would be at once explained, and the fact must not be lost sight of that the nine-line runic inscription on the principal face of the Bewcastle cross occupies the most prominent position on the face, level with the eye of the spectator, which it holds more than is the case with the figure sculpture. In this inscription the names of certain obviously Anglian personages are given as having set up the cross, and this is a piece of direct evidence not lightly to be put aside. The curious treatment of the vine, so unlikely in a con¬ tinental designer, and the emphasis on the bird form point in the same direction, while the bird itself, as seen for example on the top piece of the Ruthwell cross, fig. i6o (2), is treated in a very original style and is quite unlike a Roman eagle. The eagle with the evangelist John on the Ruthwell cross is a remarkable creation, and gives colour to the interpretation of the Bewcastle ‘ Falconer ’ that makes him a John the evangelist treated in an unpre¬ cedented and daringly secular fashion. Such boldness can be understood when account is taken of the designs on the sceat coins. The figure common on these that holds two crosses, one in each hand, sometimes changes one cross for a bird held falcon-fashion, and in one example he has dropped them both and ex¬ changed them for two flowering stems. This is not treating the cross with traditional reverence. The sudden appearance on the coins of freshly treated foliage motives in which no classical element is discernible has its bearing on the remarkable and wholly un¬ conventional design of the foliage panels on the north and south faces of the Bewcastle cross. The representation of the Christ figure with a moustache but no beard, if the fact be accepted as stated here (see pp. 227, 230), is almost decisive in favour of the theory of a Saxon carver, for the moustache is distinct¬ ively Germanic and non-classical. Moustached heads occur on the sceat coins and are not uncommon on the carved crosses of Ireland, but the Christ figures on the latter are not so treated. A bearded Christ, or one with the face smooth, is of course quite normal, but the Ruthwell Christs have no sign of a beard while the moustache seems hardly to admit of a doubt. On the hypothesis of Anglian workmanship could be best explained the curious lapses into crudity of treatment which occur here and there in the sculpture. The Magdalen’s figure is the crucial instance (see fig. 158). The general motive is very boldly devised, the striking mass of the votary’s hair being very large in its treatment and worthy of the thought of a great artist, whereas the arm and hand are childishly bad, and it is hard to believe that any classically-trained sculptor could have been responsible for them. The awkwardness of the figure of St. John the Evangelist on the head of the cross has already been explained on the supposition that the artist was more interested in the bird than in the man, and treated the latter in a sum¬ mary fashion. See antea, p. 225. If an Anglian origin for the crosses be accepted, the most likely period would be the reign of Ecgfrith rather than that of his father Oswy. Oswy had at one time wielded more power than any English sovereign had up to his time possessed, and Ecgfrith entered upon an inheritance, diminished indeed in the south through the prowess of Wulfhere of Mercia, but offering splendid potentialities in the directions of the north and west. He was not only a warrior but evinced a taste for the monumental, and his reign till it was cut short by the disaster of Nechtansmere bade fair to be as fruitful in the arts of peace as in those of war. Eddius calls him ‘ Rex Christianissimus ’.^ He was the founder of the monasteries of Jarrow and ]\Ionkwearmouth, was present at the dedication of Wilfrid’s church at Ripon,'^ and through his wife, the ^ Vita Wilfridi, c. xvi. - ibid., l.c. 267 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. saintly Ethelthryth, made donation of the land for the building of Hexham abbey.^ The spirit of his times might very naturally embody itself in monuments of exceptional scope and beauty like the Ruthwell and Bewcastle crosses, and they would take their place in the artistic movement in Bernicia of which the recognized landmarks are Hexham, Jarrow, and Monkwearmouth. THE RUNIC INSCRIPTION ON THE RUTHWELL CROSS. 1 . The deciphering of the runes . . 268 IT The text of the runes . . . 270 III. The transliteration of the runes . 272 IV. The transliterated text . . . 274 V. The inscription and the poem . 274 (1) The correspondence of the two texts .... 274 (2) The content of the poem . 276 {3) The authorship of the poem . 277 (a) Caedmon . . . 277 (b) Cynewulf . . . 278 (4) The metrical evidence . . 279 (5) The poem and the O.E. riddles 280 (6) Summary of the literary evidence .... 281 VI. The linguistic evidence . . . 283 The reading of the runic characters carved on the north and south borders of the east and west faces of the Ruthwell Cross provides us with four short groups of verse, which can be said at once to be Northumbrian in dialect and probably an early rather than a late example of that dialect. Hopes of an exacter impression of time and place depend in the first instance on the trustworthiness of the text established, and therefore on the accuracy with which the runes can be deciphered and transliterated. As the editorial history of the text shows, both these processes allow in the result considerable difference of opinion. The following text is the result of a new and independent and repeated examination of the runes on the monument. And in the vexed circumstances in which busy modern debate has left this fragment it may be well to state the principles on which they have been read and treated. ^ Vzia Wil/ridi, c. xxii. 1 . The Deciphering of the Runes. The canons observed in the preparation of the runic text are these :— (1) No letter has been admitted to the text that cannot be certified to-day on the stone, and no appeal has been made to casts or photographs or to anything but the stone itself. (2) There are letters of which only parts are decipherable—one only of two upright shafts, for instance, or part of an upright, or the upright but not the lateral strokes, or only one of two laterals, or again a lateral but not the turn-up stroke. In some of the cases, the letter cannot be inferred without prejudice to the transliteration. The last rune on the N.E. border for instance might, epigraphically, have been either of the two forms of the g-rune ; and the third rune in .t.css at the foot of the N.W. border might similarly have been either of the two runes used to distinguish the front and back rounds of c. Such letters, therefore, have not been admitted to the text. (3) In other cases, where part of the letter is lost, the part that is decipherable may provide a form which can belong to one rune only. Such letters are held to be established epigraphically by the characteristic portions of the forms which belong to them alone, and they appear therefore in the text. The examples are:—the d-rune of .o{d)ig, the /-rune of h{l)afard, the A-rune of {h)celda, the rf-rune of .ismce.rce{d)u, the e-rune of .ist{e)mi, the ce-runQ oi f us{cB), the g-rune of h.a{g), the /-rune of stre{l)um, and the A-rune of {h)ince. (4) Where readings remain uncertain or are absent, letters once read in whole or in part by former editors are recorded in the notes. Victor in his latest version admits such readings to the text on the ground that the latter should rest on the whole historical material.^ They are certainly part of the whole linguistic evidence; and they are admitted, on merit and with discretion, to 1 ' Ich habe . . . mir . . . bei der festellung des lese- buchtextes gesagt, dass nicht sowohl die jetzige lesung, so wie sie erscheint, als vielmehr ein auf dem ganzen, auch historischen material beruhender text zu geben ist ’. Letter from Vietor to Schipper, 23 Oct. 1909. Cf. Zupitza-Schipper, AU-u-Mittelenglisches Ubungsbuch, Elfte Auflage, 1915. 268 REPORT ON THE RUTHWELL CROSS. 1 llhfiacR :IM|iicn iFIR'A-KY IhtfflCHEA iFHiiruN !7ftL)AF jFRHlARD jHFIk'DAlC lilHFjNlDO iF^TFIrst/e. vl^Flj 15M !FliFHA.RA4i) In-kv/I lUUN |KET •flflijMEN IFFriBAA.T iTHi AD ,H I |HI IIFIna. ^F'igo IF’pa l^'lECH niiTTi X^b-TH fnIah flP'iEW FFIal iHFI'iDe 'W'lON 'XF'iga '.rX'LG jflX'iUG ilhT'lST 'iIXP'llGA 1 Phi o|:>) ilXfilGF 1 li I : ITHB jtPHFlLOD/E. i \Si\ 15T 'finilMEN !|li * I I I I I I ' I I : p1 » N(S I BI ‘ » I I I ' I i 1 I < I 1 t I I I I ‘ 1 i ! ^ EAST Fig. 168.—Runes on the sides of the Cross. 269 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. the later discussion. In the present editing, however, the}' ha\-e been relegated textually to the notes rather than inserted in the text itself which conforms, and is confined, to what is decipherable. (5) The plates or transcripts which may be used in collation,^ subject to the conditions stated above, are : — 1705. Hickes, Thesaurus, Gram. Island., pi. 4 . . . = H. 1726. Gordon, Itinerarium Septen- trionale, pi. 57. . . . = G. 1789. Cardonnel, Monumenta Veiusta, II, pi. 54 . . . . = C. 1832. Duncan, Archaeol. Scot., IV, pi. 14.= D. 1838. Kemble, Archaeologia, XXVIII, PP- 353-5 • • • . = K. 1866. Stephens, Ruthwell Cross . . = St. 1867. Stuart, Sculp. Stones of Scot., II, pis. 19, 20 . . . = Stu. IT The Text of the Runes. (i) On the north or right-hand border of the east face :— ' / /i/ORTH £/l5T ..XMRMHF NTe KPH WmlTTIX- Tf Nti prrm Ftj XFfxii xi'itTxf .Fflix G.'.. ...rmf .a. 3-6. Occupy 5 inches on right hand of transom, and are preceded by 4 inches of worn and cemented surface. Nothing legible. The stone had probably worn before it was cemented over. There is room for three signs. The first may have been a cross, followed by two runes =ow or un. Cf. Lindis. Mt., 27, 31, ongeradon, and Rush., tmgeredun. 20. Sweet {O.E.T., p. 125) says is impossible, and suggests an h rune with some strokes worn away. There is no trace of such wearing. is clearly and sharply cut. 48. Illegible. H., ^ Complete collation will be found in Zupitza- Schipper, op. cit., pp. 3-6, and (omitting Stuart) in Grein-Wiilker, Bihliothek der aqs. Poesie, II, 1894, pp. 111-4. G., D., K., St., and Stu., m. 50. Left up¬ right, traces of a lateral, and top and bottom of a right shaft can be read. All plates read d. 54. Three or more inches of cement cover one line, and all but the whole of another. For the first H. gives fare ; G., C., St., and Stu., fore ; D., only r. Of the second the bottom of two shafts emerge from the cement. H., G., D., and K. omit ; C. allows for it, but cannot read it; Stu. shows ends of four shafts below the cement. 64. Illegible. Stu. gives an imper¬ fect 5 ; D., a single upright. 65. Practically the whole outlines of u visible. 66. Top half of one of the two runes for g legible. Stone quite worn away below. 66. Below this line the border extended 2 feet 9 inches. Runes were probably carved to the foot, which would give some thirteen lines or more, and more than thirty runes. All traces now effaced. (2) On the south or left-hand border of the east face ;— £. SOUTH £AST .... IF Km mm NFFfPRH NPThF Ih Tl HFRWF .hrtRFHfi nm fif» If ft.fh.. IF ... .It effRf .ihfmi. 9ir Ef I. Under the left of the transom the first line of the border is in part worn, and in part cemented over. H. gives g and two uprights ; G., the fragments of letters ; C., i and e or m. 14. The rune is the u in outline, with centre part much worn, but with distinct traces of they stroke there. H. and D. give r ; G. and C., u; K. and St., y. 27. Top lateral only shows. H., G., and C. give 1 . 33. Left upright and parts of two laterals are all that remain clear, h is clear in plates from H. to D. 43. H., D., K., St., and Stu. give a ; G. and C. give 0. But the turn-up stroke of the second lateral is clear, though the lateral itself is not. 48. Part of the lower loop of a probable b is all that is visible; but b is distinct in H., G., and C. 55. Left upright, and faintest traces of curved obliques, are legible ; but not the right REPORT ON THE RUTHWELL CROSS. upright. H. gives a clear d ; the others, less clear or doubtful d’s. 57. Begun as m or e, and corrected to u. 59. The two points and upper half of middle upright are clear. The lower part is less clear. The whole letter, though faint and worn, is not to be doubted. 69. The top right slope of a letter which may have been either of the two runes, for g is visible. Earlier readings do not help. 72-3. Illegible. G., H., St. give c for 73, but Stu., d. 76. 3|- inches of cement cover a line. No early readings. 79. Right upright of letter clear, but nothing else. 86. The first lateral is clear, and the second just visible. 87. Illegible. D. and Stu. give b. 91. Only right upright and part of right oblique visible. 94. Illegible. 96. From this line to the foot of the border is 2 feet. The sculptor would have nine or ten lines at disposal, holding about forty runes, since at this point he was carving on an average four runes to the line. (3) On the south or right-hand border of the west face :— 3. 30 aTH IA/ 55 T. ARhf nh Pi mi mm [^fiR m mm }m fir m ih pr EL... h... Ih p.'i PiT. mm xiM.M N.PX I. There is room for a sign before the first letter. H., G., K., St. give a cross. In¬ distinct in C. and G. 27. Upper lateral of letter gone ; stone worn away. I 3 ut the align¬ ment clearly implies the full cc-rune. 39. The upright is clear, and faint traces of lower lateral visible. H., G., Stu. give ce. 61. Letter much worn. Left upright only visible. H. gives h ; G., C., D., and St., an upright only. 62-4. 2|- inches of cement. No readings. 65. s very plain. 66. Upright and one lateral traceable. D. and Stu. give ce. 67-8. Only bottom part of one letter visible. No readings. 72. Bottom of an upright only. St. and Stu. give (Z. 79. Bottom of an upright remains, perhaps the shaft of r. 85. A left upright and one lateral can be read. 88-9. Nearly 2 inches of the stone is worn away after this on the right edge of the border. No readings. 91. Whole outline of h quite clear. 92. Lost, where edge of stone is worn. 94. The first half of the rune for the back stop g is clear. The whole rune ^ (gar) is thus given. From here to the foot of the border is i foot 5 inches, carrying perhaps six lines and about eighteen runes. Illegible now. (4) On the north or left-hand border of the west face :— A'OPT/Y mrirnti ximfHW rmXHM NIf Nllf^ 4S SS ao 6^’ 70 XWFHHfl. Nin.I.fi ...ir.A. .T. m I. Two bottom shafts clear, and the possible beginnings of a middle oblique. Stone broken off on left ; G. gives clear m. 8. Top of up¬ right, and beginning of top lateral visible. H., G., and C. give 1 . 25. Upright and lateral of n clear, but bottom of shaft rather worn away. 29. Bottom of left shaft, whole of right shaft, and two laterals legible. This gives a certain h. Early plates all give h. 37. Cut first as X {gifu), and joined afterwards at top to form the ce-rune. 41. Whole outline of n plain. 51. Top of upright only visible. H., G., and C. give n. 52. Fracture of stone across the second oblique of h, but whole letter clearly given. 55-9. 3 inches and more of cement cover one line and half of another. No early readings. 60-1. The bottom of an upright and the whole of another upright can be read on the right of the border. 62. A right-hand curve at the foot of a letter is all that is visible. 65. From this point on only the letters printed in the text, and read under each other on the right hand of the border, are visible. One line is covered altogether, and 2\ inches of stone are worn away on the left edge. The text must be given up as irrecoverable. 76. A right upright and the suggestion of an oblique can be traced. 79. An upright is traceable. 271 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. 82. The top of the shaft is worn away, but the letter is clear. 83. Below this is i foot 5 inches of border. This would hold six lines and about twenty runes, but nothing remains. III. The Transliteration of the Runes. The following notes deal with certain orthographic distinctions obsen^ed in the present transliteration;— (i) O.E. c=Gmc. +k was the symbol of two sounds, a back voiceless stop, and prob¬ ably a front voiceless stop. Before original back vowels, a, 0, u, and their umlauts cb, e, y (=Gmc. a, 0, u, with f-umlaut), and before consonants I, n, r, etc., it was a back stop consonant. Before original front vowels, i, e, before Gmc. +;', and, when final, after front vowels, it was fronted and pro¬ nounced probably on a front stop.^ In MSS. both sounds are as a rule written c. The glossaries hardly distinguish between the two at all; and though k appears in some West-Saxon MSS., it is nowhere used consist¬ ently as a discriminating sign. In the earliest runic inscriptions the form for the original sound is (; and the oldest English runic inscriptions, such as that on the gold solidus imitated from a coin of Honorius,^ and that on the Chessel Down scabbard mount,® show an intermediate form /\. After the fronting of the sound before front vowels, the Anglian runic alphabet, and, so far as we judge from the few specimens in the South, the Anglian only, retained the original sign in a modified form for the new sound, and invented a new sign for the back stop. The Anglian futhorc therefore has separate symbols for the two sounds, {cen) for the front, and {calc) for the back. Ruthwell, further, has two forms for the latter sound, /f\ for the normal back stop, and (a modification of X gar, or perhaps of the /-rune, ger) for a front variety of it. In the present reading the distinction between back, front, and front-back sounds is kept by printing c for the back stop, c for * Sweet, H.E.S., pp. 142-44, and A.S. Reader, 8th ed., §§ 110-30; Wyld, Trans. Phil. Soc., 1899, pp. 134-41 ; Sievers, Gram, of O.E. (Cook), § 206 ; Biilbring, Altenglisches Elementarbuch, § 493 and note. 2 Baldwin Brown, Arts in Early England, III, pi. 3, and pp. 68-69. ^ Stephens, Handbook of O.N. Runic Mon., p. 246. the front, and k for the front-back. The examples are :— /|\ = c in crist, cwomu. 1^—c in ic, riicnse, kyninc. ^ =k in unket, kyninc. (2) O.E. g was also the symbol of more than one sound, and its phonetic value is not always clear. Generally speaking it was a voiced consonant, back or front, and stop or open, according to the nature of its associated sounds, and under the same condition as those differentiating e. O.E. g = Gmc. before original back vowels and their umlauts, before O.E. a^Gmc. a -f nasals, and before con¬ sonants I, n, r, etc., also medially* between back vowels, and finally after back vowels, was a back open-voiced consonant. At the end of the O.E. period, as a late and rare development, it became a back stop voiced ; but in early O.E. its use as a back open voiced is clear, and its history presents few difficulties.1 In MSS. it was written g, and the original rune for it was X {gifu). But before front vowels, and before all diph¬ thongs, and their umlauts, and after front vowels, it was fronted, and became a front open-voiced consonant. MSS. do not dis¬ tinguish with any consistency. But later West Saxon often indicates the front sound by writing e after g. From this usage, from the nature of the sounds associated with it, and from the subsequent history of the word, its front value is generally clear. The Anglian runic alphabet expresses this value by means of the old rune X {gifu), and provides a special symbol ^ {gur) for the back. There is a third g in O.E., = Gmc. +/, which was either a front open-voiced conson¬ ant, or a vowel {i) with consonantal value. In either case its value was assimilated with that of the front g before the first texts in the Latin alphabet, and MSS. therefore observe no distinction in representing the sounds. The original j rune was ^ ; but though it appears in the O.E. Runic Poem as 4 {ger),'^ and in various abcdaria, there is no evidence ‘ Sweet, H.E.S., pp. 146-49, and A.S. Reader, §§ 110-30; Wyld, op. cit., pp. 147-50; Sievers, op. cit., § 206 ; Biilbring, op. cit., § 492 and note. * V. 32. Cf. V. 87. Bruce Dickins, Runic and Heroic Poems, pp. 16 and 22. REPORT ON THE RUTHWELL CROSS. that it was in use in England. The letter {iav), which may be a survival or modihcation of ger, is found in two inscriptions—Dover, gislheard and Thornhill, gilsuiih ’—with the value of front g, and in two doubtful uses on the Brunswick casket ^; but there seems to be no clear example in inscriptions of the value Gmc. +/. After its assimilation with front g it would be expressed by X [gif^i] and it is to be noted that in the Bewcastle gessus, gifu represents the Latin consonantal i. Ruthwell has only the back and front sounds of g = Gmc. +g. In this text the distinction is kept by printing g for the back consonant, and g for the front. The examples are :— =g in god, galgu, gistiga, so . gum, h . a(g). X=g in geredae, almehttig, gistiga, .o(d)ig, gi. roe . . d, giwundaed, alegdun, limwoe- rignse, gistoddu. It should be noted also that in the matter of these two sounds of c and g, the only evidence for their separate values, other than that of the inscriptions, is the indirect philological evidence of the original Gmc. sounds, of the nature of the associated sounds in O.E., of the later diphthongisation of vowels after front c and writing of c after front g, of alternative spellings in early M.E., and of the subsequent history of the word. The Ruthwell text ^ therefore, with its remarkably developed phonetic sense, is, along with the inscriptions from the North of England, of prime importance in determining the pronunci¬ ation of these sounds in O.E., and generally in dating O.E. sound changes.® (3) In general in runic inscriptions the h of the combination ht (xi) is represented by the rune [55]. The fifth rune however in the word almehtiig is the peculiar thirteenth letter 1 Stephens, Handbook, p. 141. ^ Ibid., p. 149. ’ Ibid., p. 119. * Sweet’s text in Oldest Eng. Texts, 1885, p. 125, Vietor’s in Die Northumbrischen Runensteine, 1895, p. 6, and in Zupitza-Schipper, op. cit., pp. 6-7, and Kluge, Angelsdchsisches Lesebuch, 4th ed., 1915, p. 114, observe the distinction in transliterating. * It is to be noted, however, that the distribution ol these symbols, and particularly of the c and c runes, on Bewcastle is by no means so clear or con¬ sistent as on Ruthwell, and involves certain difficulties as to the phonetic employment of the differing signs on the former monument which are not raised in the present case. of the runic poem. In early inscriptions this sign is rare. It is not found in the Scandinavian, and in German its value is uncertain. There are at least hve other examples of its epigraphic use in England. On the Dover stone it appears in g^slheard ^; on a Thornhill gravestone in Eate^nne ^ ; on the Brunswick Casket in Sigh^r, and in Mung- paelyj' ®; and on the Cross Shaft at Urswick Church in Toro^tredae.''^ In Anglian futhorcs and in MS. runic alphabets of the 9th, loth and nth centuries it appears with the names eoh and ih, and the sound values of eo, i, h, ih, k.^ Some of these latter were no doubt given to it from the form of the name, from the West Saxon eoh, and the Anglian ih ; and as the phonetic values which it bears in inscriptions may have been a similar inference, it is not easy to determine the original sound of which it was the symbol, or the exact sounds for which it stands in the six Anglian cases. It was probably at no time an un¬ ambiguous symbol. But it appears, in un¬ certain usage, to have represented the two sounds, i, x a-nd it is the second of these for which it clearly does duty in the Ruthwell and Urswick examples. The original repre¬ sentation of the -xt sound appears to have been -ci. This is the usual spelling in the glossaries, where -cht also occurs, and -ht only occasionally and doubtfully.® In Caedmon’s Hymn -ci occurs in niaecti, dryctin, and allmectig. The connection of the combination with its normal W.S. spelling -ht has been kept in the present text ; but the letter has been distinguished from ordinary h by a diacritic. (4) 3 ^. the twenty-second letter of the runic poem occurs twice in the inscription, in kynihc and uiiket. The origin of this rune appears to have been the cursive Greek double gamma, and it is possible that its earliest value was that of «+g. It represents here, however, the front or back nasal occurring in O.E. only before g or c (k). In MSS. this 1 Stephens, Handbook, p. 141 ; value i. ^ Ibid., p. 149 ; value i. ^ Ibid., p. 119. * Collingwood, Trans. Cumberland and Westmore¬ land Antiq. and Archceol. Soc., N.S. 9, 1911. ^ A list of these names and values is given by IMiss Anna C. Paves, Mod. Lang. Review, VI, 1911, PP- 450-51- ® Sievers, op. cit., § 221, n. i ; Biilbring, op. cit., § 481 ; and Chadwick, Studies in O.E., 1899, p. 241. 18 273 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. sound was not distinguished from the ordinary dental nasal -n. The runic distinction between the two sounds has been kept in transliteration by emplo^’ing for ^ the ordinary phonetic symbol With these explanations of phonetic dis¬ tinctions cori'esponding to formal differences in the runes themselves, the general prin¬ ciples obser\ ed in transliterating the runic text may now be summarised. They are as follows :— (1) The transliterated text should corre¬ spond throughout to the runic ; and, to secure this, a single runic sign should be represented wherever possible by another single symbol. On this principle, then, p> [wyn) =w, ^ [thorn) — ]y [cesc] —CB, and ^ [epel) =ce. (2) There is one rune to which this principle has not been applied. [ear) is the rune for the O.E. diphthong -ea ; cf. Beagnoth on the Thames scramasax,^ and Gislheard on the Dover stone.^ It is transliterated here in heafun CBS, fear ran, and in the single instance -ed, in the last line, by the bind form -ed, to mark the fact that the diphthong repre¬ sents a single rune.^ (3) The transliteration should preserve all the runic evidence intact, and should therefore observe the distinctions discussed above in the -c and -g series, and in the -n and -n, and in the -h and -h runes. (4) Letters inferred from characteristic parts should be distinguished from letters deciphered in whole. This has been done by printing the former within brackets. IV. The Transliterated Text. /K=c X=g \i=cB K-=c X=g ^=n !>=}? ^=CB 1 Stephens, Handbook, p. iii. 2 Ibid., p. 141. ® It should be remembered that in Northumbrian, and especially in early Northumbrian, this rune might represent also -eo. Confusion between these two diphthongs is a frequent, though not a distinctive, feature of this dialect. Cf. Biilbring, op. cit., § 108 ; and Luick, Hisiorische Grammatik der Eng. Sprache, Leipzic, 1914, § 119 and § 133, Anm. i. Further, the Salzburg alphabet gives the name eorand the value eo for the rune. . . geredcC hinge god almehttig Jja he walde on galgu gistiga . o(d)ig f. men . u . 5. ic riicnas Kyninc heafunaes h(l)afard (h)gelda ic ni dorstae . ismgerae(d)u unket men ba get . ad . . ic ... . ij? blodae . ist(e)mi . 10. bi Crist waes on rodi hwejrrae ];er fus(ae) fearran cwomu . ]7j?ilge til anum 15. ic ]:>get al bi .... s . . . ic w . s mi. so . gum gi. roe . . d h . a(g) mi]? stre(l)um giwundsd alegdun hiae (h)inae limwoerignae 20. gistoddu . him .i. aes . . f. . m . . . ea . . u . . i. ]?e V. The Inscription and the Poem. (i) The Correspondence of the Two Texts. The poem containing the lines to which these inscribed verses correspond, the Vision of the Cross, is preserved in the Codex Vercel- lensis (Cod. CXVII), a miscellany of O.E. prose and verse in the Cathedral library of Vercelli.^ The handwriting of this MS., which appears to be the same throughout its 135 written folios, is of the late loth century ; ^ and the 156 lines of our poem (11. i04’^-io6®) are in the usual dialect of such collections; 1 Wulker, Codex Vercellensis, Leipsic, 1893—a facsimile of the poetical parts ; Forster, II Codice Vercellese, Rome, 1913—a facsimile of the whole MS. ; and Forster, Der Vercelli Codex, CXVII, Halle, 1913— selections from the prose homilies. 2 960-980, Keller, Angelsdchsische Palaeographie, Berlin, 1906, p. 46 ; 970-980, Keller, Angelsdchsische Schrift, in Hoop’s Reallexikon der germanischen Alter- tumskunde, I, 1911, p. 102; 'later decades of loth century,’ Forster, II Codice Vercellese, pp. 11-14 ; and Brandi, Geschichte der ags. Lit., Strassburg, 1908, p. no ; ‘ 2nd half of loth century,’ Holthausen, Elene, Heidelberg, 1905, p. ix. Wulker’s ‘ offenbar aus dem anfang des 11 Jahrhs.’ (Codex Vercellensis, p. vii) has not been endorsed since Keller. 274 REPORT ON THE RUTHWELL CROSS. that is, in normal mixed late West Saxon. It is probable, and has hitherto been more or less assumed, that the basis of the majority of such texts was originally Anglian, and that it subsequently passed through various hands and finally into those of a West Saxon redactor. But it is by no means clear this West Saxon character is due merely to later, or to the latest, scribes. And it is at least possible that the linguistic poems in Vercelli are those rather of a mixed literary dialect. Fbrster has pointed out that it resembles the language written at Worcester at that period.^ The MS. may have been produced at Worcester, or the literary language of Worcester may have extended beyond the boundaries of the diocese in the loth century. Exacter inter¬ pretation of the language of the MS. will probably follow on the lines suggested by Forster.'^ The following is the text of 11 . 39 to 49® and 11 . 56'^-65® of the poem in The Vision of the Cross from the Vercelli MS ^;— Ongyrede hine ]:)a geong haeleS, ];aet waes god aelmihtig, 40. Strang and stidmod ; gestah he on gealgan heanne modig on manigra gesyhSe, J^a he wolde mancyn lysan. Bifode ic, J?a me se beorn ymbclypte : ne dorste ic hwaeSre bugan to eorSan, feallan to foldan sceatum, ac ic sceolde faeste standan. Rod wses ic araered : ahof ic ricne cyning, 45. heofona hlaford ; hyldan me ne dorste. jmrhdrifan hi me mid deorcan naeglum ; on me syndon }:>a dolg gesiene, opene inwidhlemmas ; ne dorste ic hira aenigum sceddan. Bysmeredon hie unc butu aetgaedere ; call ic waes mid blode bestemed, begoten of ]?aes guman sidan. 56.Crist waes on rode. 1 Forster, It Codice Vercellese, pp. 19-20 ; and Der Vercelli Codex, pp. 33-35. Cf. Cook, The O.E Elene, Phcenix and Physiologus, Yale, 1919. pp. vii, viii. ^ Cf. Brotanck, Anglia, Beiblatt, XXVI, 191s. pp. 225-38. “ Grein-Wtilker, Bihl. der ags. Poesie, II, 1894, pp. 116-125 Cook, Dream of the Road, Oxford, 1905 ; Sweet, A.S. Reader,^ 1908, pp. 154-158 ( 11 . 1-89) ; Kluge, Ags. Lesebnch,* Halle, 1915, pp. 110-114; Wyatt, A.S. Reader, Cambridge, 1919, pp. 130-132 (11. 28-89). Hwaedere Jiaer fuse feorran cwoman to ]iam aedelinge ; ic jiaet eall beheold. Sare ic waes mid {sorgum) gedrefed ; hnag ic hwaedre J^am secgum to handa, 60. eadmod elne mycle. Genamon hie }?aer aelmihtigne god ahof on hine of dam hefian wife ; forleton me ]?a hilderincas standan steame bedrifenne ; eall ic waes mid straelum forwundod. Aledon hie daer limwoerigne, gestodon him aet his lices heafdum ; beheoldon hie daer heofenes dryhten ; and he hine daer hwile reste, 65. mede aefter dam miclan gewinne. A translation of these lines is; ‘ Then the young man, that was God Almighty, stripped himself, strong and steadfast. Bold in the sight of many he mounted the high cross when he would redeem mankind. I trembled when he clasped me, yet I durst not bow to the ground or fall to the lap of earth, for I must needs stand fast. I was raised a Cross : I lifted up the great King, the lord of heaven : I durst not bend. They pierced me with dark nails, and the wounds are visible on me, the open wounds of malice. I durst not in¬ jure any of them. They mocked us both together. I was all wet with blood, and streamed on from that man’s side. . . . Christ was on the Cross. But eager ones came from afar to that noble one. 1 saw it all. I was sorely troubled with (sorrow) ; yet humbly and with zeal I bowed to the hands of these men. Then they took the Almighty God, and lifted him from that heavy torment. Those warriors left me standing, covered with blood. I was all wounded with arrows. There they laid down the limb-weary one ; they stood at his body’s head, and gazed there on the lord of heaven. And for a time he rested there, weary after the great struggle.’ Now, if we compare these two te.xts, other than philologically for the present, we may note some differences in those lines or parts of lines occurring in both, (i) The subject of . . geredcB in R. i. is god almehttig ; in V. 39 it is geong hcelefr with \^cEt wees god celmihtig as an expanding phrase. {2) R. 2 has ]>a he walde on galgu gistiga ; V. 40 has gestah on gealgan heanne. But it is wrong here to ' MS. mid. gedrefed. Cf. Ruthwell, 16, so. gum. 275 HISTORICAL MONU]\IENTS (SCOTLAND) COMMISSION. suggest that the ‘ author ’ of the poem would not have been guilty of the indecorum of attributing to Christ the desire to mount the Cross in the sight of all men. The word walde does bear this meaning, and the difference is a material one of motive, and not merely of expression. (3) R. 3, /. men, is represented by V. 41, on manigra gesyhde. (4) R. 7, . [h)celda it ni dorstce, is V. 45, hyldan me lie dorste. R., that is, omits the object which hyldan usually takes, and repeats the subject. (5) R 8, . ism(srcB{d)n . unket men ha cet .ad. . is V. 48, Bysmeredon hie line hutu cetgcEdere. (6) R. 9, it .. . is V. 48, eall ic wees. (7) R. 14, . '^])ilce til anum is V. 58, to ]>am ce^elhxge. But it will not do to infer from the occurrence of the latter phrase in Genesis and in Daniel that the Vision poet is in the line of poetic tradition, and that the Ruthwell writer is not.’- (8) R. 18, giwundeed, is V. 62, forwundod. (9) R. 19, {h)ince, is V. 63, ticer. It will, no doubt, be agreed that in spite of these variants, in some of which the sense takes a slightly different turn, the line for line correspondence of the two versions is too close to be classed as a case of mere resemblance or similarity. We may safely speak of the identity of the two passages. At the same time, the few verbal discrepancies are not in themselves of a nature to support any theorizing as to either the superiority or priority of either text. It is to be noted that the Ruthwell lines are obviously incom¬ plete in sense, and imperfect in measure and alliteration. But it is important again not to build too hasty an argument upon this. The general proposition advanced that lack of alliteration, lack of metre, and imperfect sense in an inscription, and their opposites in a poem in manuscript, establish the poem as the original of the two is very doubtful, and in any case cannot usefully be applied here. For, while there can be no objection to explaining the Ruthwell incompleteness by supposing that the sculptor of the runes found himself with a narrow space at his disposal and chose from a poem such verses as he 1 R. 12-14—eager noble ones came from afar to that lonely one; V. 57-58—eager ones came from afar to that noble one. But who are the fuse, the eager ones ? Cook, op. cit., p. 32, says Joseph and Nico- demus. Wyatt, op. cit., pp. 232-233, suggests angels. thought suitable, and carved them where he had room for them, we must be careful not to pass from this perhaps harmless assump¬ tion to the further and different statement that the engraver of the inscription took his verses from the poem, if by the poem we mean, as we then may, our Vercelli text of the Vision of the Cross. For this would be to go well beyond the warrant of the facts if in no other point than that it suppresses the necessary alternative that an earlier text than either of the two present ones existed,—a Northumbrian text, of course, or at least an Anglian, from which the sculptor selected parts for his purpose, and which a West Saxon compiler transposed into the Vercelli form. From the existence and circumstances of the two versions we are not entitled to say initially that either of them used the other, still less which of the two preceded the other. What we can say at this stage is : (i) that the two can hardly be other than versions of the same essential poem, or part of a poem ; (2) that one is, in its present epigraphic condition, a fragment and imperfect,’ but that the other occurs in what is probably a version of the complete poem; (3) that the fragment is in the Northumbrian dialect, of a date to be determined, if possible, by study, but the text of the longer poem is in late West Saxon. (2) The Content of the Poem. It is of some importance to notice, next, that the lines to which the inscription corre¬ sponds occur in the second of three fairly well- marked parts of the poem.^ Of these the first, 11. 1-27, is in the nature of a prologue describing a vision of the Cross at midnight. It appeared as a tree moving in the air. But it was not ‘ the gallows of any wicked man ’. It glittered with gold and jewels, and was gazed on by angels, saints, and men. Yet it changed in colour, red on the right side, and through the gold could be seen to be wet with blood. At the close of this description the I But note that we do not possess the full in¬ scription as it once was on the Cross. See notes above on the Text of the Runes. We cannot infer anything as to the fragmentariness and imperfection of the original complete inscribed text. ^ Prose translation in Kennedy, Poem of Cynewulf, 1910. 276 REPORT ON THE RUTHWELL CROSS. Cross itself begins to speak. Obviously this first part is an echo of the vision of Constan¬ tine, and similar in its terms to other echoes of the vision in O.E. literature—to those in the Elene} in the Christ,'^ in an nth-century homily,® and in ^Elfric’s sermon on the Invention of the Cross.^ They derive prob¬ ably indirectly from Lactantius or Rufinus, rather than directly from Eusebius.® It has been suggested that the wcBdum geweorttod (‘ adorned with its vestments ’) of I. 15 may be a reference to the actual veiling of the Cross on Good Friday. But in general it is the symbolical Cross, the sign of victory, that is described in terms that are familiar in O.E. literature, and draw on the one hand from the well-knowm vision of Constantine, and, on the other, from the presence in churches of the Altar Cross or the Processional Cross [Crux stationalis) . In the third part, again, 11. 122-end, which serves as epilogue, the poet is also speaking in person. He finds himself praying at the foot of the Cross, ‘ alone with a small band ’, and resolves to worship it more and better. In 11 . I3i‘’-i36® he remembers friends who died before him, and in 11 . i36*’-i46^ he asks when the Cross will come to lead him home from earth to heaven. These lines are similar in turn of thought and expression to those in which Cynewulf closes the poems that carry his signature. But the student of Cynewulf’s epilogues is slow to draw any biographical or personal deductions from this manner of writing. 11. 146*’-! 56 seem almost super¬ fluous, being different in tone, and of poorer quality. They have been held to be an addition by another hand, or a later accretion of some kind. But there is no evidence for this ; and even to hold these lines an artistic mistake, as some do, is a merely modern judgment of no value. The general point ' 11 . 69-104, ed. Holthausen, Strassburg, 1905, and Cook, Yale, 1919. * 11 . 1081-1102, ed. Cook, Boston, 1900. “ Morris, Legends of the Holy Rood, E.E.T.S., 1871, PP- 3 - 17 -. ^ iElfric, Homilies, ed. Thorpe, 1844-46, II., pp. 303-7- ® Eusebius, Life of Constantine, I, 28-31 ; Migne, Pat. Gr., XX, 944-45 and 948 ; Lactantius, Of Manner in wh. the Persecutors Died, XLIV', Migne, Pat. Lat., VII, 260-62, Rufinus’ version of Eusebius’ Eccl. Hist., IX, 9. Cf. Stevens, The Cross in the Life and Lit. of the .d.S. (Yale Studies, 23), 1904. for criticism to notice is that neither prologue nor epilogue of the Vision is at all remarkable, among companion O.E. pieces, for any feature of poetry. With them the Ruthwell lines have nothing to do. The case is quite otherwise with the second or main part of the poem in which the Cross speaks, and tells its history. Of this part again, 11 . 75’^-i2i are of lesser interest. They describe—the Cross still speaking itself—the invention, the decoration, the exaltation above other trees, and the bestowal of ‘virtue’, or power to heal. In its last words it bids the poem proclaim his vision, and promises salvation to the faithful. The motive here is plainly liturgical, and the passage is matched by many Latin hymns in adoration of the Cross—by the Eccc lignum Crucis, the Crucem tuarn adoramus, the Dum Fabricator Mundi, the Crux benedicia nitet, the Vexilla Regis, and most of all by the Range lingua of Fortunatus, which was admired and imitated by Alcuin. It is therefore in 11 . 28-74^* that the final interest of this poem resides. In them the Cross speaks, and tells how it was felled long ago on the edge of a wood, and stripped from its trunk. Strong men dragged it to a hill, and set it up. It saw the Lord hastening to mount it boldly. It could have laid its foes low, but it had to stand fast. Then follow the lines quoted and translated above. In II. it is cut down, and buried in a deep trench. It is enough for the present to say of this part that it is unique in thought and expression among all kindred poems ; and that this distinction lies not in the fact that the Cross is endowed with per¬ sonality, or in the emotional quality of its consciousness, and even in the strange beauty of its language, but simply and historically in this, that it is the boldest adaptation of Christian matter to a northern and barbarian setting that can be found in O.E. literature. (3) Authorship of the Poem. Two theories of authorship have in the past found advocates, and are still occasionally repeated. A few' words will state briefly how the case stands with both. (a) The theory of Caedmonian authorship. —As regards the inscription, this theory was made to rest on the fact that various people. 277 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. following Stephens, claimed to have read on the two borders of the eagle panel on the North face some such words as Cadnion mce faucetho. The epigraphic facts are these; On the right border of this panel the runes are illegible, and not a letter can be traced. On the left border there is a doubtful but possible bind-rune mce followed by the fairly clear letters, faii{ory)cetho. These letters mean nothing, and the inscription here must be abandoned. As regards the poem, no external evidence in the MS. or in the history of the poem suggests their authorship ; and there is no internal evidence to argue upon, since we possess practically nothing written by Caed¬ mon with which to compare it. What emerges from Bede’s account ^ is that Caedmon came to the Abbey of Streanaeshalch some little time after its foundation, when it was already famous and flourishing,- but before the death of Abbess Hild; that he outlived her for some time; and that after composing a number of scriptural poems, he died of old age.^ The Abbey was built in 657-59 ; but its flourishing period was not till after the MTiitby Synod, 664 ; and Hild died in 680. It is a fair inference, therefore, that Caedmon’s composition ought to be dated 670-690. The one ascertained fragment of this com¬ position, the nine lines known as his Hymn, survives in its original Northumbrian dialect in a MS. written rather later than 737, and there are two later continental versions of the same Northumbrian text.^ It is not ab¬ solutely certain that these nine lines represent his original words. The terms in which Bede introduces his Latin version of them at least allow an interpretation of them as a re-trans¬ lation from the Latin prose. Even so, they cannot differ materially from what the original must have been. But, beyond this, nothing else can be proved to represent his authentic composition ; and the collection of religious and liturgical poems of various dates and schools and places, differing in source, motif, ^ Bede, Hist. EccL, IV, 24. 2 Multis doctioribus viris prcssentibus. ® Corporea infirmitate. * The MSS. are Camb. KK., 5, 16 ; Dijon Bibl. Mnnicip., 574 ; and Paris, Bibl. Nat. Cod. Lat., 5237. Cf. Bede, Opera Historica, ed. Plummer, 1896, II, p. 251 If.; Wuest, Zfd.A., XLVIII, p. 205 ff.; Zupitza- Schipper, Vbungsbuch, p. 2 ; and Forster, Ags. Lese- buch, 1913, pp. 2-4. thought, style, and language—the Exodus, Daniel, Genesis A, Genesis B, and the rest of the Junian MS.—are not now generally associated with Caedmon himself.That there was a school of religious poets in North¬ umbria alter Caedmon is clear from Bede’s statement.2 That there was a considerable output of religious lyrical poetry at this date, dependent on ecclesiastical and Latin literature, but retaining many marks of the barbarian way of poetry, is almost certain. The Vision of the Cross might have come out of this school, or have been related to it, or might have occurred at any point of time in it after 680. But we cannot speak in any strict sense of Caedmonian authorship, for the reason that nothing exists objectively, to serve for comparison, as a basis of Caedmonian composition. [b) The theory of Cynewulf’s authorship. —Our knowledge of Cynewulf is derived from four poems which bear his signature in runic characters— Juliana,^ Christ II., Elene, and The Fates of the Apostles. We have there¬ fore a canon, and an established basis of com¬ position, for comparison and argument. But our literary application of this knowledge is limited by the fact that the criteria provided by these four poems are by no means uniform, unambiguous, or even consistent. They do not suffice to determine fully either the period or the place of his authorship, still less to limit the range of his activity, or define the terms of his school. Among a number of alternatives offered, therefore, criticism cannot chose with anything in the nature of proof. The Vision poem, for instance, shares some features with the four genuine poems of Cynewulf, in particular some forms of phrase, with Elene and Christ II. But less weight will be attached now to such similarities of style in deciding questions of authorship than formerly ; and, in point of fact, they are less numerous and less striking than might be supposed. It may be noted, too, for what it is worth, that the types of 1 Sarrazin has recently returned to the view, for which there is much to be said, that Genesis A is probably of original Caedmonian authorship. Cf. Von K'ddmon bis Kynewulf, Berlin, 1913. 2 Alii post ilium in gente Anglorum religiosa poemata facere tentabant. ^ Ed. Strunk, Boston, 1904. ^ Ed. {Andreas and Fates of Apostles), Krapp, Boston, 1906. 278 REPORT ON THE RUTHWELL CROSS. phrase in question are more traceable in the prologue and epilogue than in the middle part of the poem. They do not justify an assumption of common origin for the Vision and the undoubted poems of Cynewulf, still less do they establish anything like proof of it. It is doubtful if they are strong enough to provide an inference of knowledge of the one text, or set of texts, on the part of the writer of the other. Joint share in the same general tradition meets the case.^ But, if this is not enough, there is nothing in the w'ay of supposing that the author of Elene knew' the Vision poem.^ It is best not to deal in such curiosities of conjecture ; and w'e need not assume know'ledge or indebtedness one way or the other. If, however, the assump¬ tion must be made, those who make it must keep in mind the fact that the resemblances on which it has to be based leave it quite open for the Vision to be the earlier poem of the two. Any further arguments, biographi¬ cal or otherwise, based on the alleged ‘ con¬ fessions ’ or ‘ conversions ’ of the Cynewulfian epilogues, are worthless.^ If, therefore, the Vision belongs to what is loosely called ‘ Cynewulfian ’ poetry, w'e have no means of placing it at any particular point in the sequence of that poetry. We can summarise all that the facts entitle us to by saying: (i) that authorship by Cynewulf himself is hardly possible on literary grounds (and is made practically impossible by the linguistic evi¬ dence of the inscription) ; (2) that membership in the group of poems (other than the four signed) wEich sometimes bears his name is just tenable ; but if so, there is evidence again, both literary and linguistic, that the Vision, is early, if not indeed very early, compared > The whole stress of criticism on the personal element or authorship of poems of this period is un- historical, and derives from the modern idea of literary property. Similarity between pieces at this date means ‘ tradition ’ ; and does not mean identity of authorship, or direct copying. 2 Sarrazin, op. cit. . holds that the poet of Elene does once betray his identity with the author of the Vision. This argument turns on the fact that in Elene, 11 . 88- 90, the Cross is made to appear to Constantine not as in the original, in the Acta Sanctorum, but as it appears after the Invention, decked with gold and jewels as in the Vision. Cf. the argument on this anachronism, op. cit., p. 121. ® Cf. Brown, Englische Studien, XXXVIII, 1907, pp. 196-233. with the main body of this poetry. On this view the Vision would not be later than the middle of the 8th century, and is more prob¬ ably as early as the beginning of that century. But neither date nor authorship is susceptible to proof by merely literary argument. And we must leave it at that. (4) The Metrical Evidence. It is a pity that in the present state of our know'ledge little, if any, light can be thrown upon the authorship of an O.E. poem by a consideration of the metrical evidence. The main features of the O.E. line are admitted and agreed upon by all scholars. So prob¬ ably are the five fundamental types to w'hich Sievers reduced the varieties of half verses, provided always it is remembered that these represent a convenience of classification for us rather than fine-drawn conditions of artistry for the poets themselves. Further refinements and modifications upon these types, apt as they are to vary with the w'it or the pedantry of the classifier, are negligible for critical purposes ; and arguments for authorship are not securely based on the mechanical tests evoked by these metrical studies. In the case of the Vision, the argument is less likely to turn on such points as the number of un¬ stressed syllables before or after the main stresses, as on the presence in the poem of a number of expanded lines, and on the relation of these to the normal line. Now', these lengthened lines occur more or less in all O.E. poems, either isolatedly or more commonly in groups. In some poems, such as Juliana and Maldon, they are not found at all. In Genesis, Daniel, Christ, Andreas, and Giithlac they are present both singly and in batches.^ In Judith, a poem of 355 lines, there are 63 long lines and 5 long half lines. In the Vision, out of 156 lines 34 are lengthened. These two latter poems resemble each other in em¬ ploying few variations from the ordinary type of lengthened line. Further, they are alike not only in the frequency and the nature of their lengthened lines, but in their use of w'hat may be called long-lengthened lines, 1 Cf. Sievers, Der ags. Schwellvers, in Paul und Braume, Beitrdge zur Geschichte der deutschen Sprache und Lit., XII, 1890 ; Kaluza, Die Schwellvers in der ae. Dichtnng, Eng. Studien, XXI, 1895. 279 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. in which, in addition to the extra stress in the half line, there are as many as three or four extra unstressed syllables. These crowded and longest forms do not occur in Beowulf, in Genesis, or in Exodus, and there are only a few examples of them in Christ. And, speak¬ ing generall}- of the lengthened lines in O.E., it must be admitted that their scarcity, amount¬ ing almost to absence in Beowulf, and their absence again at the other end of the scale in such a native poem as Maldon, makes it diffi¬ cult to assert that the lengthened line is a primitive feature of O.E. poetry. It looks as though it was a tendency of this poetry to develop the long lines, and a tendency of the long lines to become longer. But this cannot be accepted as a full or final statement. It depends, for one thing, upon assigning a late date to fudith ; and, though this is generally done, it is by no means beyond debate or criticism. And it rests, in addition, on the dangerous habit of generalising from the evi¬ dence of a single text, and that text, as it happens, an unusual type of poem. And, in view of the facts that the lengthened lines do occur sporadical!};' throughout all the poetry, that their origin is uncertain, and that it is difficult, indeed impossible, for us to under¬ stand or explain their employment, it is un¬ safe, to say the least, to cite this technique of O.E. verse as a criterion of date or authorship. In the Vision 7 lengthened lines occur in the prologue (8-10, 20-23) f orie in the epilogue (133) ; and one {75) in the liturgical lines describing the burial and discovery and adoration of the Cross. The remaining 25 (30-34. 39-43. 46-49. 59-69) are in the cen¬ tral portion of the poem in which the Cross itself is speaking. We can hardly dismiss this as an insignificant distribution when we remember that it is practically only in the prologue and epilogue and the liturgical lines that we find the echoes of Cynewulfian phrase and usage, and that the lengthened line is not on the whole a distinctive, and certainly not the distinctive, feature of the technique of the genuine poem of Cynewulf. On the other hand, in the Ruthwell text only 5 of the lines (5, 6-7, II, 12-13, 14-15) correspond to short lines in Vercelli (44, 45, 56, 57, 58). The other nine appear to be lengthened lines cor¬ responding to lengthened lines in Vercelli. Other early fragments of Northumbrian verse, Credmon’s Hymn, Bede’s Death-Song, and the Leiden Riddle are in the normal verse. But the Charms, probably the oldest verse of all, imply its use. If we cannot see, therefore, in the metre and technique of 11. 28-75^ of the Vision, a sure sign of age, we cannot ignore a certain characteristic, which s^iggests, if it does not actually constitute, a difference from the rest of the poem. And there are other reasons for believing that this difference tells appreciably in the directness of earliness and age. (5) The Poem and the O.E. Riddles. One of these reasons concerns the relation which this part of the poem bears to the Exeter Riddles.^ A number of these ^ are in the form Ic seah (7 saw .... a won¬ drous sight), with which we may compare the Vision, 11 . 4-23 ; while in another group ® the subject of the enigma is quickened into life, made to speak in the first person, and recounts its life history. It has been suggested by Sarrazin that the Vision is essentially a poem De Cruce composed of these two riddle types.^ And it would seem, further, that it is not only related generally, in point of poetic personification of an inanimate object, to this form of poetry, but has a certain affinity in structure and spirit to particular riddles written in this style. Riddle 30, for instance, is solved by ' The Cross,’ or rather by ‘ The Tree or Cross.’ ^ Lines i to 4 de¬ scribe the life of a tree in the forest, and lines 5-9 are most reasonably interpreted as refer¬ ring to the subsequent life of the Tree as a Cross. In the same spirit ‘ The Battering Ram ’ in Riddle 53 recounts in a few lines its life story.® It remembers its happy life in the forest on the hillside before it suffered the strokes of the axe. Then in its new form it glories in its power 1 Cf. Tapper, Riddles of Exeter Book, Boston, 1910 ; Wyatt, O.E. Riddles, 1912; Wood, A.S. Riddles, Aberystwyth Studies, 1912 ; Trautmann, Die ae. Rdtsel, Heidelberg, 1915. 2 Cf. 29, 31, 34, 36, 37, 38, 42, 51, 52, 53, 55. 56, 59, 68, 86—numbering as in Wyatt. ® Cf. 2, 3, 5, 6, 9, 12, 14, 26, 40, 63, 71, 72, 76, 82. ^ Cf. Von Kddmon bis Kynewulf, 1913, P- 116. ® Tupper, cross ; Wyatt, no solution ; Trautmann, Baum-Kreuz ; Wood, Beam. Cf. Blackburn, Journal of Eng. and Germ. Phil., Ill, 1900. ' Tu., Wy., Tr., Wo., battering ram. REPORT ON THE RUTHWELL CROSS. in battle, and tells how it prepares a way for warriors. The opening lines of this riddle nray be compared with lines 28-30 of the Vision. Again, in Riddle 72, ‘ The Spear ’ speaks. This is a riddle of thirty lines, of which fifteen are very imperfect and impossible to reconstruct. But the opening lines are plain, and in them ‘ The Spear ’ or ‘ The Spear- shaft ’ tells in the same style of a transposed life.^ It flourished as a tree (the ash), then fell into cruel hands. It was smoothed and polished, and made into a weapon, and lives to boast of its power of speedy slaughter. To take one more example, the solution of Riddle 55 is almost certainly ‘ The Cross,’ ^ and the lines describe the four woods of which the Cross was made, and its life among men in the hall. Now, there are several points suggested by these riddles which a further study of the whole O.E. and Anglo-Latin Riddle litera¬ ture, of which they are examples, will enforce. The first is the fact that in riddles generally a familiar and characteristic device of the poet is that of endowing the object with life, and making it speak, and sometimes speak with passion, as the Cross is made to speak in the Vision. The second is that in a certain number of riddles, which poetically are among the most vivid in the whole collection, the object so endowed and speaking is a tree or is something made from the wood of a tree. The third point is that, as in the Vision again, the tree in speaking recalls its former life, its freedom and freshness in the forest, its humiliation at the hands of its enemy—man, the new shape given to it, and the new use to which it was put by men, and its pride in its new power and function. It was once a living thing, it fell before a ‘ murderer’s will’, it was raised again to honour and usefulness : this may be called the common formula of the tree-speaking riddles. The fourth signi¬ ficant thing is that the riddles that employ this formula are either those the tone of which suggests that they are among the older pieces of O.E., or are those that show the queer, sharp blend of the elements of Heathendom and Christendom that is characteristic of the ‘ Tu., Wy., spear or lance) Tr., der Mauerbrecher) Wo., spear-shaft. ^ Tu., cross ; Wy., scabbard ; Tr., die Harfe ; Wo., scabbard. first O.E. Christian poetry. The great period of riddle writing in England, whether in Latin or the vernacular, was 650-800 ; and the English collection obviously belongs to the 8th centufy, and preferably to the first half of the century. It may be taken for granted that it is not by Cynewulf himself, and not in any sense, however loose, Cynewulfian in origin or quality.^ Theoretically each riddle should be dated in and for itself; but it may be taken that they were written in the main in the first thirty or forty years of the 8th century. We need not suppose that they were all written at one time or by one poet. Probably in the middle or at the end of the loth century, the collector of the Exeter MS. was drawing upon more than one source, and from several smaller and earlier collections of riddles, and the Exeter corpus may be looked on as miscellaneous in origin, and, within certain limits, of date. But that the Vision should combine in its structure two of the main riddle types, and that it should offer as well definite affinities with the Tree group of riddles, are at any rate literary factors to be remembered later, when it is found that linguistically the closest parallel to the Ruth- well text is the Northumbrian version of the Leiden Riddle - — one of the few I'iddles that can be almost definitely dated. (6) Summary of the Literary Evidence. It does not appear that there are any other criteria of usage or style that throw light on the question of the authorship or origin of the poem. The application of minute tests of grammar or syntax, such as that of the presence or absence of the definite article before a weak adj.-|-subst., can hardly be conclusive, given the nature of the trans¬ mission of O.E. texts ; and it is doubtful if they can be applied at all, satisfactorily, to an admittedly exceptional te.xt such as the Vision.^ Nor is there any other literary or textual parallel that can be called into use 1 This is the whole trend of both Riddle and Cynewulf scholarship. An exception is Tupper, Mod. Lang. Notes, Dec. 1910. 2 See later, p. 285. ® Cf. Barnouw, Textcriiische Untersuchungen, Lei¬ den, 1902; Sarrazin, Eng. Stud., XXXV’IIl, 1907. p. 145 ; Richter, Chronologische Studien zur ags. lit., Halle, 1910, p. 93. HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. The O.E. lines on the Reliquary in Brussels Cathedral/ said to contain fragments of the True Cross, are hardly to the point. They read :— Rod is min nama : geo ic ricne cyning basr byfigjmde, blode bestemed. ‘ Rood is my name ; once I bore the great King, trembling and wet with blood.’ The inference is that it was once regarded as a piece of the True Cross. The lines may, or may not, have been taken from the O.E. poem ; and they may, or may not, confirm the presumption that the Ruthwell inscription was also taken from the poem. But even if they were, they come from a part of the poem which might well have existed earlier than the first complete Anglian text of the whole poem. The literary evidence, therefore, may be summed up in the general terms of a broad and plain case. By the end of the 7th century or beginning of the 8th century the various facts in the history of Cross worship, and most of the forms of that worship and of Cross literature were known in England. A church¬ man of that date was familiar with the Vision of Constantine, the Discovery of the Cross by St. Helena, the Dedication of the Church of the Holy Sepulchre, the Restoration of the Cross to Jerusalem by Heraclius, the spread of the Adoration of the Cross from Constan¬ tinople to Rome, and from Rome to the West¬ ern Church, the Latin Hymns of the Church, the observation of both the festivals of the Cross, the Exaltation (May 3) and the Inven¬ tion (Sept. 14), and finally with the exposition by Pope Sergius X. of a Fragment of the Cross for veneration in Rome.^ Throughout the 1 Cf. Logeman, L’Inscription Anglo-Saxonne du Reliquaire de la Vraie Croix, Ghent and Leipzig, 1891; and Cook, Date of O.E. Inscription on Brussels Cross, Mod. Lang. Rev., 1915, pp. 157-61. 2 Brandi holds that the discovery of this fragment in the Pontificate of Sergius was the actual occasion of the Vision poem, which was designed, in this view, to interpret the new religious festival, and draw attention to the fact that, as Bede records, Abbot Ceolfrith of Wearmouth was in Rome at the time. Cf. Sitz. Ber. der kon. preuss. Akad. d. Wiss., Berlin, July 1905, and translation of the paper in Scottish Hist. Review, 1912, p. 139. The two accounts of the establishment of the festival will be found in Liher Pontificalis, Pars Prior, ed. Mommsen, Mon.Germ. Hist., Berhn, 1898, p. 213, or ed. Duchesne, Paris, 1886-92, 1 , 374 ; and in Bede, De Tempornm Ratione, Sexta cBtas [a.d. 701], Opera, ed. Giles, 1843, VI, p. 328. 8th centuries and 9th the adoration of the Cross was at its height in Anglo-Saxon England; but the various forms of Cross worship, and of its liturgy and literature, were known at any rate from the end of the 7th century. In the Vision of the Cross there is not a conception with which we need suppose a churchman at the end of that century to have been unfamiliar ; and every factor in the liturgical setting of the poem allows this early date. On the other hand, there are two sets of facts which make it unlikely that the poem was written much later than this. The first is that it knows practically nothing of any of the later Cross legends of which the liter¬ ature of the early Middle Ages proper is full. It does not know of the growth of the tree planted in Jerusalem of which the Cross was made ; it does not know of the four woods of the Cross, or think of it as made of more than one kind ; and it naturally does not know of the crucifix, but only of the Cross. It is safe to say that at any time from the end of the gth century a Cross poem in which one or other of these legends did not enter would be remarkable. And generally speaking the ‘ earliness ’ of the Vision is sufficiently established by the fact that it knows nothing of the theological mysticism of later Cross literature, and in point of fact is'neither a theological nor a mystical poem, but a ‘ heroic ’ poem on a Christian subject. The second group of facts is that the evidences of bar¬ barian thought and expression, which are striking throughout the second part of the poem, are such as imply, and indeed allow, no great break, of time or tradition, from the heroic poetry of the Germanic North. Treat¬ ment of Christian themes in the spirit and manner of secular heroic poetry is the familiar fact of all O.E. religious verse. It is at its height as habit in Andreas and Elenc ; but at least in Andreas it is more a case of specific knowledge of Beowulf than of unconscious memory of the traditions of the race. The occurrence of poetic archaisms in the Vision is not surprising ; but their presence in that kind of poem, otherwise so closely in touch with both the Church and Latinity is, and still more sur¬ prising is their early and primitive nature. Christ is not a sacrifice or a sufferer in this poem, but a young hero performing a feat of power and endurance. It is the Cross or tree 282 REPORT ON THE RUTHWELL CROSS. whose history and sorrows count. Strong enemies {strange feondas) seized it : there were many of them {feondas genoge) : it was pre¬ vented from fighting them; it was made to carry their outlaws {wergas) : it trembled when the young hero clasped it {ymbclypte) : ^ warriors {hilderincas) took their chief from the Cross, and gathered round the body : in the evening they sang the death song {soyhleof)^ over it. This last reference to the death wail is one of the oldest things in barbarian poetry, so old that, as Brandi says, ‘ nowhere save in Beowulf is the custom mentioned. Cyne¬ wulf and his contemporaries have long for¬ gotten it ’. And of the whole tone of their references we may say that they certainly accord best, if not only, with an assumption of early date. It is possible that, since they occur mainly in the second part of the poem, this part should be referred to an earlier date than the other two. The others are palpably more in the manner of the Cynwulfian poetry ; the second retains appreciably more of the formulae of the heroic verse. If this cannot be proved, it is so likely that it must at any rate be allowed for. In the balance of probabilities, then, the lines corresponding with the Ruthwell in¬ scription occur in a poem preserved in late West Saxon in a MS. of the late loth century which may in its original Northumbrian or Anglian form have been at least as old as 700, and need not on any count have been later than 750. When, therefore, the argument of ‘ selection from a poem for purposes of in¬ scription ’ is applied to the Ruthwell text, it may result in establishing a claim for the poem as the earlier of the two ; but it cannot, by establishing that, prove or even suggest lateness for the inscription. The latter might be taken from the poem and still be as early as 700 ; and to prove the Ruthwell inscription later than the Vision of the Cross is, so far as regards dating Ruthwell, to prove nothing at all. From which it follows that while all the literary evidence concerned allows an early date, it does nothing to determine it ; and ^ Possibly, as Sweet suggested, a confusion between crucifixion and hanging. - It is scarcely possible that this is merely a poetic term for ‘ lamentation,’ and not an archaic feature of the original A.S. burial rite, and therefore an evidence of great antiquity. Cf. Schiicking, Das ags. Toten- klagelied, Eng. Stud., XXXIX, 1908. the student of the inscription must conduct his linguistic case independently of any opinion as to the date or authorship of the poem. VI. The Linguistic Evidence. When we turn to the language of the inscrip¬ tion we must remember that three clear categories of words can be used in evidence : (x\) Words of which every letter has been deciphered and certified on the stone; (B) Words of which a letter or more may be missing, and yet the whole word may be given by the context (epigraphic and literary), or by the Vercelli text, in such a way that the letters supplied to complete it have no effect on the phonological evidence of the rest of the word ; and (C) Complete syllables that give their grammatical or phonological evidence inde¬ pendently, and irrespective of what the exact form of the rest of the word may have been. The examples are :— A. geredcB, liince, god, almehttig, ]>a, he, walde, on, galgu, gistiga men, ic, riicnce, Kynihc, hedfunces, ni, dorstce, imket, ha, hlodcB, Crist, wees, rodi, hwelpm, per, fedrran, cwoniu, til, anum, pcBt, al, mip, giwundeed, aligdnn, hire, linmoerigncB, him. B. and C. {m)o{d)ig, h{l)afard, {h)(Blda, {b)ismcb rce{d)u, {b)ist{e)mi{d), bi , fus{cB), . ppilcB, gi{d)rce{fi)d, so{r) gum, stre{l)um, gistoddu{n), .i.ezs, {h)incB. (i) Now in these words the first feature to be noted, as supplying the first criterion to be applied, is the presence of ce in unaccented syllables where normal O.E. texts almost always write e. In Ruthwell this ae appears in fifteen words; geredcB, hincB, riience, heafuncBS, dorstce, blodce, hwepree, giwundeed, hire, limwoerignee, bismcerce{d)u, fus{cB), . ppilce, ■ i . CBS, {h)ince. In each of these cases where the corresponding word or form appears in Vercelli the unstressed e is written.^ On the other hand, of this unstressed e there is only one certain example in Ruthwell, tvalde. It is agreed that in the whole range of O.E. this CB is archaic, and the change to the normal e is generally referred to the early 8th century, ' Except in the case of giwundeed, for which Vercelli has forwundod. HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. though it can be found in a few cases as early as the end of the 7th century. But it is in the nature of such changes that they do not operate uniquely at a given point of time, and variety of usage is to be expected, and will be found, on either side of the dividing line. No one contends that e cannot be found in an early text, or that <®’s do not linger on in later texts. In a charter of 806 or 810 we have this archaic cb ending in two words, geuuor^iae and gulliae} Just as in the Moore IMS. of Caedmon’s Hymn (737) we have lirofe, and in the St. Gall MS. of Bede’s Death Song (qth century continental, but an accurate copy of the Northumbrian original) we have fore, there, and daege. Spora¬ dic occurrences of e early and of cb later do not interfere with the broad fact that ce is archaic and e is normal in O.E. To suggest lateness, therefore, in Ruthwell on the ground either of the occurrence of walde in the in¬ scription, or of isolated examples of cb in even qth century or loth century texts, is to con¬ fuse the argument. The point is not that e cannot occur in genuine early text, but that wherever j^ou get a uniform or predomin¬ ant retention of cb’s, you have a strong indica¬ tion of age. Both occur in Ruthwell, as both occur in Caedmon’s Hymn. But in both these texts cb’s predominate — in Ruthwell in the proportion of 15 : i in a text of 44 words. This is positive evidence of age not easily to be upset. (2) A second and similar feature of the inscription is its use of i for the normal e in unaccented syllables. This i occurs in 10 words — in gistiga, ni, {b)ismcBrcB{d)u, {b)ist{e)mi{d), bi .... , rodi, bi .... , gi{d)roe[fi)d, giwundcBd, gistoddu[n). Of e there are only two possible examples, neither of them unexceptional, . . geredce and %ihket. This change from i to e is also a weaken¬ ing from an archaic to a normal form, and was dated by Sievers about 750. It can be traced, however, a good deal earlier, and must have begun at any rate by the end of the 7th century. Some care, too, must be taken not to generalize too hastily 1 Cf. Harmer, Bng. Hist. Documents, 1912, pp. i, 70, and 129. The Charter is, however, Kentish, which considerably modifies its evidence. * St. Gall MS., thS ; of which the expansion may be them. Cf. Forster, Ae. Lesebuch, 1913, p. 8, n. i. or rigidly on rather unequal evidence. E for i is certainly as old as 692, for it appears twice unstressed in an East-Saxon charter of that date. In Mercian charters of 736 and 742 it occurs, and in a Kentish charter of 740. On the other hand, a Mercian grant of 769 has the unstressed, and in the Northumbrian genealogies of 811-814 both e and i are written in unstressed syllables.^ In the Leiden Riddle (continental MS. of the qth century, but presuming older Northumbrian) there is a preponderance of i’s with a few e forms in the proportion of 8 : 3, and there are 2 examples of i in Bede’s Song and 9 in Caedmon’s Hymn. The entries on pages 88 and 208 of the Lindis- farne Gospels contain gi’s for ge’s, and the gi form appears frequently in the last chapter of St. John. The entries and the gloss to the Gospel belong to the latter half of the loth century, but there is reason for believ¬ ing the glossater was working from an older copy.- Obviously, as in the case of cb and e so with i and e it is not a question of sporadic occurrence of one or the other, late or early, but of the predominance of the archaic or the normal in texts which we know otherwise to be early or late. E is found as early as 692 : i undoubtedly persists till late, and, as it happens, was apparently written later in Northumbrian, or in Anglian, than in the other dialects, for in the Liber Vitae of the beginning of the qth century it still pre¬ dominates. What is clear is that on the whole unstressed i prevails until about 740, after which date unstressed e predominates. Now, Ruthwell shows a significant predominance of archaic i’s, 10 : 2 in a text of 44 words, a feature that it shares, therefore, with only the earliest, or copies of the earliest, texts. And this evidence is not disposed of by assuming that ‘ when a man was penning an inscription, monumental or otherwise, he was likely to archaize or to manifest peculiarities due to a high-strung condition ’. (3) A third feature of the text of the inscription is its use of /, medially, where some have expected to find the archaic and original b. Medial / occurs in hedfuncBS, h[l)afard, gi{d)rce[fi)d. There is no example of final /. It is not easy to apply the 1 Cf. Sweet, O.E. Texts, pp. 167, 426, 428. 2 Cf. Anglia, Beihlatt, XII, p. 142, and Anglia, XXIV, p. I. 284 REPORT ON THE RUTHWELL CROSS. evidence for this usage in the form of a criterion for date. In normal spelling final h is only irregularly found after the middle of the gth century. But it is twice in the Liber Vitae (gth century) in Ciiobualch and Leobhelni, and a Kentish charter of c. 831 has ob ^eni lande, and one of 832 has ob niineni erfelande. This latter charter has also medial b in hiabentiie and Luba ; ^ and the medial b may have persisted later than the final. On the other hand, / is found quite early. Caed¬ mon’s Hymn has both hefaen and hcben, and in the same MS. of Bede / occurs in the majority of cases, and it can be found in a Kentish charter as early as byg. Also there is only one example of either medial or final b in the Leiden Riddle {ob repeated twice) as against many /’s. If it were possible to generalize from such peculiarities, and from so uneven a usage, and to apply the generaliza¬ tion with any conviction of its worth as a test, we could say that the regular use of medial / for b in Ruthwell suggests a date not earlier than the beginning of the 8th, or possibly the end of the 7th century. But it rather seems to be one of those points in O.E. phono¬ logy, on which date cannot fairly be made to hinge at all. (4) A fourth feature is the frequent loss of final 11. This n is lost in the infinitive in Ruthwell in gistiga and . {h)celda. In our early Northumbrian texts there are only a very few instances of the infinitive, such as the hergan of Caedmon’s Hymn,'^ and the cnyssa and haatan of the Leiden Riddle. In late North¬ umbrian, that is after c. g5o, the loss of n is frequent, but we obviously cannot make a comparison. In the Preterite Plurals of verbs Ruthwell preserves the n twice in alegdun and gistoddti{n), and drops it twice in cwomu and {b)ismcsra{d)u. Loss of n after u occurs in cer¬ tain positions in early Northumbrian ; whereas in late Northumbrian n is almost regularly found in the Preterites of verbs, and by far the greater number of such Preterites cer¬ tainly end in -un or -on. There is one other n-less form in Ruthwell, galgu. W’hatever its explanation, it is paralleled in the foldu of Caedmon’s Hymn, by the eorthu of the Leiden Riddle, and by the flodu and thesu/w of ‘ Cf. Harmer, Eng. Hist. Doc., pp. 7, 8, 9, 129. ^ In the Moore MS., but herga in the other two. the Franks Casket,^ and eorthu appears often enough in the Lindisfarne Gospels. It is to be noted that this dropping of n is not distinctly Northumbrian, or by itself a distinctive proof of age. For there are numerous ex¬ amples of it in the Hatton MS. of the Cura Pastoralis. But it is certainly not, as has been contended, an indication of lateness. All that can be said is, that Ruthwell drops it five times, and retains it thrice,—which is inconclusive. (5) A note may be added on the following forms in the inscription :—(a) dorstcB. This is the Preterite of a verb in the Third Ablaut Class of Preterite Presents, dear (Gothic ga-daurs), Pret. dorste (Gothic ga-daursta). As the Gothic forms, and the Old High German gitorsta, and the Old Saxon gitorsta, and the Frisian thorste, all show, the original vowel in the stems is 0. The Lindisfarne form, gedarste (Mark, 12, 34), gidarste (John 21, 12) is by analogy from the present, and presumably is later Northumbrian. If dorstce, therefore, is to be argued from, we are entitled to assume that the text in which it occurs is at any rate older than that of Lindisfarne, by the time of which the a form was used. There is no diffi¬ culty in taking dorstcB therefore, in its stem as in its unstressed vowel, as early, (b) kynihc. In certain cases, where g became palatal, or as a result of vowel syncope, ng became nc. But the combination appears variously in texts as ncg and ngc.- Of the latter there are two examples, Duningcland in a Mercian charter of 788, and Theodningc in a Mercian charter of 77g. It is to be remembered, however, that in the inscription a single rune represents the nasal stop consonant, often written ng ; and no argument for date turns on it. (c) hddfuncBS. It has been suggested that this un is a later form than the hefaen and the heben of Caedmon’s Hymn ; but while cen and en may represent the older forms, the confusion in such endings goes back to the 7th century ; and in view of the occurrence of heofones and heofone, as well as of heofenas, heofenum, it is possible that hedfunces re¬ presents the original alternative which the forms in Caedmon’s Hymn seem to indicate. * Cf. Napier, The Franks Casket in Furnivall, “ Eng¬ lish Miscellany,” 1901, p. 379 and note, and p. 380. “ Cf. Sievers, O.E. Grammar (Cook), § 215 ; and Biilbring, op. cit., § 566. 285 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. (d) . The history of this word has been summarized in a recent article/ in which it is made clear that its earliest recorded form was ce]yil, that from about 740 onwards this i rapidly changed to e, that about the same time the of the first syllable changed in Northumbrian to e, and that the joint later development of these two changes is the Northumbrian e\>el. . therefore, as far as the word can be read, and without prejudice to its first letter, is strong evidence ‘ in favour of a date not later than about 750 for the inscription on the Ruth well Cross.’ (e) unket. Several arguments have been advanced in favour of the lateness and even against the genuineness of this word.- (i) It has been read and transliterated as ungget, and then described as unfamiliar, with the suggestion that the carver probably blundered. But the second rune in the word is that for the nasal before g or c {k), and is the same as that employed in Kyninc ; and it is followed by the rune, transliterated here, and also in Kyninc b}’ K, as it should be ; that is to say, by the mid-form which signifies some degree of front¬ ing, but not full palatalisation. Epigraphically, therefore, it is correct and normal in regard to both the runes in question. (2) It has been said that the only other occurrence of the word is in a text with late spellings. This is the form uncet in Cockayne’s ‘ Shrine,’ 42, 27. But the occurrence of incit twice in Genesis A, 2732 and 2880, makes it probable that there were two parallel forms, uncit and incit. (3) It has been advanced that the et ' Cf. M. D. Forbes and Bruce Dickins, Mod. Lang. Rev., Jan. 1915. 2 Cf. Cook, The Date of the Ruthwell and Bewcastle Crosses, Yale, 1912. is late and due to lack of stress. But e and i vary so much in such cases that, as has been shown already, an isolated case of one or other proves nothing. And on all three counts there is not only no proof that unket is a late form, but none that it is not even an earlier form than the incit of Genesis A.^ Finally, in summarising the linguistic evi¬ dence, and in drawing conclusions from it, two considerations should be kept in mind. On the one hand, in dealing with such a text as the Ruthwell inscription, the question to ask is not, ‘ Can some of its archaic forms be found sporadically in texts of the ninth and tenth centuries ? ’—for this proves hardly anything about Ruthwell itself, but, ‘ Can such a uniform employment of the oldest forms of the language be matched in any but the oldest documents, and is it likely that such a number of old forms could occur in the narrow compass of fifteen lines, and within them, in unbroken sequences of words, and that text not be itself among the oldest ? ’ On the other hand, among such documents those that offer the best parallel, as being not only in the same dialect but also literary and poeti¬ cal texts, are the Northumbrian versions of Caedmon’s Hymnf of Bede’s Death Song,^ and of the Leiden Riddle.'^ If the upper limit of date is not determined, or even determin¬ able, by the use of such evidence, perhaps the lower limit is. 1 A full analysis of this form by Bjorkman will be found in Eng. Studien, LI, 1917, pp. 76-78 ; cf. also Victor in Anglia, Beiblatt, Jan. 1915, pp. 4-5. ^ Cf. above, p. 278. ® Cf. Brotanek, Texte iind Untersuchungen zur ae. Lit. und Kirchengeschichte, Halle, 1913, pp. 150-94. ^ Cf. Schlutter, Anglia, XXXII, 1909, pp. 384-88, and XXXIII, 1910, pp. 457-66. 286 GLOSSARY. Affronle. —Full-faced (Heraldry). Alb or Albe .—A long linen ecclesiastical vestment with close sleeves. Amice. —An ecclesiastical vestment in the form of an oblong square of linen covering the neck and shoulders, the apparel or embroidered edge of which stands up like a collar. Arcade .—A series of arches supported on columns or piers. Architrave .—(i) The lowest division of an entablature. (2) The moulded border returning round the jambs and lintel of a door or window. Argent .—Silver (Heraldry). H ms.—The sharp edge formed by two surfaces meeting at an angle. Ashlar .—A square hewn stone for building purposes. Aumbry {Ambry). —A cupboard formed by a recess in the wall. Aureus .—Roman gold coin. Bailey. —(i) The external wall enclosing the outer court and forming the first line of defence of a feudal castle. (2) In later writers—the outer court or base-court of a feudal castle. Banquette. —A raised way, running along the inside of a rampart or parapet or bottom of a trench on which soldiers stand to fire at the enemy. Barbican. —A tower or advanced work defending the entrance to a ca.stle. Barmkin. —An enclosing wall — Cf. p. Ixi. Barrel-vault. —A form of vaulted roof resembling the interior of a half barrel standing on its edges. Base-court. —The lower or outer court of a castle or mansion. Basement. —The lowest or fundamental portion of a structure ; the lowest storey (not a cellar) of a building, especially when sunk below the general ground level. Bead-and-hollow moulding .—See Moulding. Beakhead .—An ornament shaped like a bird’s beak, used in Norman mouldings. Bend. —A band of the same width as the Fess running diagonally across the shield from the dexter top to the opposite base (Heraldry); when reversed in direction it is known as a Bend-sinister. Berm .—A platform on the slope of a rampart. Blazon (a coat of arms).—To blazon a coat of arms is to describe it in proper heraldic terms. Bor dure .—A border round the shield (Heraldry). Bressumer .—A wooden girder or beam let into and supported at either end by walls, which carries the other floor members. Bretasche .—A projecting covered platform on a castle wall, giving cover while commanding the wall face below. Cable-moulding. —See Moulding. Calvary or Calvary-mound .—The stepped base of a cross. Canon {of a bell ).—The part by which a bell is hung. Cap or Capital .—The uppermost member of a column. Cap-house .—The small erection covering the stair leading up to the parapet walk of a building. Cartouch .—An ornate panel, oval, round, or angular, which usually is placed in a pediment. Cast—To cast a ford. —To obstruct it by trenching the river bed or by damming the river. Cavetto .—See Moulding. Chamfered .—The arris or edge of a stone, etc., originally right-angled, cut aslope or bevel. Stop-chamfered. —Where the chamfer or splay is not carried to the extremity of the arris or edge but terminates in a wrought stop. Chase .—A groove. 287 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. Chasuble .—The outer vestment worn at Mass, in form an almost circular cape with an opening for the head. Cheeky, cheque, chequy. —Divided into squares [chequers) of alternate tinctures (Heraldry). Chevron. —A charge formed by the lower portions of the bend and the bend-sinister (Heraldry). Cinquefoil. —(i) See Foil. (2) A flower of five petals (Heraldry). Clavicula. —Lat. ‘ Little key ’ : name given to a curving prolongation of a rampart to cover an entrance. Cob. —Clay with an admixture of chopped straw. Co«so/es.—Projecting stones or corbels. Cope. —In masonrj-, the stones laid on the top of a wall by way of finish. Coped-stone. —A stone with the top or upper surface sloping down on each side. Corbel. —A projecting stone, usually moulded, to support a superincumbent weight. Corbel-table .— A row of corbels supporting a parapet or cornice. Couchant. —Applied to a beast hang down but with head uplifted (Heraldry). Counterscarp. —The counter or opposite slope to the scarp or inner face of a ditch. Couped. —Cut clean off by a straight line, as distinct from erased, in which the line is ragged (Heraldry). Crenellated. —Battlemented ; indented parapet of alternate solids and openings. Crocket. —An ornament representing leaves of conventional design, employed to decorate spires, canopies, etc. Cross (Heraldry) : (1) Cross-crosslet. —A cross with each arm ending as a trefoil or treble bud, but later having each arm crossed at the end. M’hen the lowest arm is not crossed, but pointed, it is described as a cross-crosslet'fitchy. (2) Cross-paty [croiz patee). —Strictly a cross with its arms terminating directly in a form resembling a fleur-de-lys; usually a cross with expanding arms cut square at the end, which is more strictly described as a cross-formy. (3) Cross-potent. —A cross having a bar across the end of each arm, making it T-shaped. Crow-stepped. —“ Corbie-steps ” : projections in the form of steps on the sloping sides of a gable. Curtain or Curtain-wall .— An enclosing wall. Cushion-Capitals. —A capital consisting of a cube rounded off at its lower angles. Cusps. —Each of the projecting points between the small arcs or “ foils ” in Gothic tracery, arches, etc. Discontinuous Impost.—See. Impost. Dog-tooth ornament. —A typical mediaeval carved ornament, consisting of a series of pyramidal flowers of four petals. Dormer, Dormer window. —A window pierced vertically in the slope of a roof. Drum-tower. —A projecting round or hemispherical tower. Embrasure. —(i) Opening in a wall or parapet for cannon ; (2) the inward slanting enlargement of a door or window. Enceinte. —Enclosing wall : often applied to space enclosed. Engaged shafts. —Shafts connected with the jamb, pier, respond, or other part against which they stand. Engrailed. —Edged with a series of concave curves (Heraldry). Ensigned. —The charge distinguished by a significant mark or ornament placed upon it (Heraldry). Entablature. —The parts directly resting on a pillar or column, composed of architrave, frieze and cornice. Entresol. —A subordinate storey between two main floors of a building. Erased. —Ragged, as if torn off, as distinct from couped or cut even (Heraldry). Escarpment . — The steep face of a fortified position. Escroll. —A scroll (Heraldry). Escutcheon or Inescutcheon. —A small shield usually in the centre of the large one (Heraldry). Extrados. —The exterior curve of an arch. Eucharistic. —Pertaining to the Holy Communion. Facade. —Front elevation. Fanon. —See Maniple. Feal-dyke. —A turf wall. Fess. —A band across the centre of the shield (Heraldry). Fibula. —A clasp, buckle, or brooch. Finial. —The ornament crowning a pinnacle, gable, canopy, etc. Fitchy. —Applied to a cross where the lowest arm is pointed (Heraldry). Fleche. —A little spire of timber or metal. 288 INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. Flory. —Ornamented with Fleurs-de-lis. Flory-counter-flory. —Flory on both sides (Heraldry). Focus. —The cup-like hollow on the top of a Roman altar. Foil (trefoil, quaterfoil, cinquefoil). —A leaf-shaped curve formed by the cusping or feathering in an opening or panel. Fosse. —A ditch or moat. Frieze. —The middle part of an entablature between the architrave and cornice. Gardant. —Applied to a beast full-faced (Heraldry). Garderohe. —Mediaeval sanitary provision. Gargoyle. —A stone gutter spout, often wrought as a grotesque. Glacis. —The sloping approach to a fortified place. Groined. —Having an angular curve formed by the intersection of two arches. Grouted.—Whore a hard-setting mortar is poured into the interior of a wall, binding the stones together. Gules. —Red (Heraldrj^). Helm. —Helmet. Heraldic Achievement. —The full armorial bearings. The coat of arms, together with all the exterior ornaments of the shield—helmet, mantle, crest, motto, etc. (Heraldry). Hood-moulding. —See Moulding. Hurcheon. —A hedgehog (Heraldry). Hypocaust. —A heating arrangement under the floor. Impost. —The member of a pillar or pier from which the arch springs. Discontinuous Impost. —Where the arch mouldings impinge and die out in the splayed jambs. Intrados. —The interior and lower line or curve of an arch. "Jethart Justice ” (Jedburgh Justice).—Summary punishment. “ Hang first and try afterwards.” Label. —(i) A narrow band on the chief of a shield from which hang three or five “points” at right angles—usually borne as a “ difference ” or mark of cadency (Heraldry). (2) A Hood-moulding (q.v.). Label-stop. —The projecting wrought stone terminating a Label or Hood-moulding (q.v.). Machicolation.- —Openings between corbels for the discharge of missiles, or other defensive material. Majuscule. —A large or capital letter. Maniple (or Fanon). —An embroidered scarf, originally a handkerchief, which is worn suspended from the left wrist as a liturgical vestment. Mantling. —A cloth protection over the helmet (Heraldry). Mascle. —An open lozenge (Heraldry). Mensal Church. —A church allotted to the Bishop ; so called because the revenue from the church was applied for the supply of his table (mensa). Mezzanine. —See Entresol. Minuscule. —A small letter. Mote. —A conical earthen mound surrounded by a ditch and originally surmounted by a wooden fortress within a palisade. Moulding : (1) Bead-and-hollou) moulding. —A combination of a moulding circular in section and a cavetto. (2) Billet-moulding. —Properly an enrichment resembling billets or cylinders of wood spaced at intervals on the concave surface of a moulding. (3) Cable-moulding. —A convex or rounded moulding the surface of which is cut into convolutions to represent the twisted strands of a rope. (4) Cavetto-moulding. —A small concave moulding of one quarter of a circle. (5) Edge-roll moulding. —A rounded or circular moulding, usually accompanied by flanking fillets (rectangular mouldings) or quirks, wrought on the rybat angles at a void. (6) Hood-moulding. —A projecting moulding on the face of a wall above an arch, usually following the form of the arch. (7) Roll-and-hollow moulding. —A roll-moulding along with one or more concave mouldings. (8) Roll-moulding. — Cf. Edge-roll. Mullet. —A five-pointed spur rowel ; like a star, but with a hole in the centre. Mullions. —Upright shafts dividing the lights of windows. Newel stair. —A winding stair, the steps turning round a central upright pillar. Offset. —The sloping ledge on a wall or buttress where the upper part is set back. Ogival .—Having the form or outline of an ogive or point eel (Gothic) arch. Oillet. —See Oylet. 289 19 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. Or. —Gold (Heraldry). Orphrcy. —An embroidered band or bands on the front of an ecclesiastical vestment from the neck down¬ wards. Oylet, Oillet. —An eyelet hole ; an aperture or loop-hole for observation, etc. Pcsyuila, Plancta. —A large cloak, originally with a hood, for covering the whole body, of which the Chasuble {q.v.) is a development. Pale. —One of the ordinary’ charges of a shield, a band in the centre running from top to bottom. Palmette. —An ornament with narrow divisions or digitations, somewhat resembling a palm-leaf. Paly. —DiHded into perpendicular diHsions like pales (Heraldry). Parapet Walk. —A walk fomred for defensive.purposes at the wall-head of a tower. Patera. —A shallow, saucer-like vessel. Paty, Patee. —See under Cross. Pediment. —The triangular or circular part over the entablature, etc., of a building. Peel, or Peel-Tower. —Peel: a wooden palisade which might be daubed with clay, enclosing other buildings for a garrison ; in later times, in some cases, surrounding a tower; finally, the name is transferred popularly to the tower itself, then to all such towers. Pellet. —A circular boss or raised part, rounded or flat, in decorative work. Penannular. —Of the form of an almost complete ring. Pent-house roof. —A lean-to roof wdth one slope. Pilaster. —A rectangular pillar placed on a wall and partly in it, only showing a fourth or a fifth of its thickness. Piscina. —A basin with a drain discharging into the thickness of the wall, set in a niche or recess S. of the altar, where the chalice was rinsed and the priest washed his hands. Plat. —A flat surface. Plinth. —The rectangular solid under a pedestal or wall; the base of a column. Portcullis. —An iron grating hung over the entrance of a fortified place, and capable of being let down to defend the gate. Quadripartite vault. —A vault divided into four compartments by ribs or groins. Quaterfoil. —(i) See Foil. (2) A four-petalled flower (Heraldry). Quern. —A hand-mill for grinding grain. Quillons. —The arms forming the cross-guard of a sword. Quirk. —A small acute channel or recess between mouldings. Quoins or Quoin Stowes.—Dressed comer stones. Raggle. —A groove cut in stone, especially on a wall, to receive the end or edge of a roof. Rain-water head. —Enlargement at top of a rain-water conductor. Raking cornice. —A comice which returns upwards, forming a pediment. Rampant. —The attitude of the lion, as in the Scottish shield, standing on the sinister hind-leg with both fore-legs elevated, the dexter above the sinister, and the head in profile (Heraldry). Redoubt. —A strongly fortified post or position. Re-entering angle. —An angle pointing inwards. Regality. —A feudal right of lands granted by the Sovereign, with jurisdiction conferred upon the vassal which was equal to Royal in criminal matters and to that of a Sheriff in civil cases. Regardant. —Applied to a beast looking backward (Heraldry). Reveal. —The side of an opening or recess at right angles to the face of the work. Roll-atid-fillet steps. —Steps with a rounded nosing, beneath which is a supporting fillet. Rubble. —Coarse walling, constructed of rough stones irregular in size and shape. Rusticated. —In masonry a dressed stone with a rough surface. Rybats. —The dressed stones forming a Reveal {q.v.). Sacellum. —A small temple or shrine. Sacristy. —The vestry of a church in which the sacred utensils, vestments, etc., were kept. Safe-lintel. —A wooden lintel behind a stone lintel or arch. Saltire. —A St. Andrew’s cross (Heraldry). Scarcement. —A projecting ledge in a building. Scarp. —The inner slope of the ditch of a fortified place. Sceattas or Sceat-coins. —Anglo-Saxon coins. Scoinson. —Inner edge of the side or jamb of a window or door. Sedilia. —Seats for the officiating clergy, usually on the south side of an altar. Sejant. —Said of a lion or other animal when in a sitting posture (Heraldry). 290 INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. Servery. —A room from which meals, etc., are served. Skew-put. —The lowest stone of the skew or coping of a gable, invariably projecting over the wall. Slipped. —Applied to a flower, branch, or leaf, plucked—not cut off (Heraldry). Soffit. —The under-side of a staircase, lintel, cornice, arch, canopy, etc. Soles. —Sills. Spandrel. —The irregular triangular space between the outer course of an arch and the corner of the rectangular figure within which the arch is contained; or any similar space. Squint. —An opening at an angle. Steimrtry. —Lands specially erected by the Sovereign into a Stewartry, or Regality-lands on their falling to the Crown. The jurisdiction was equal to that of Regality {q.v.). Stole. —A long narrow scarf fringed at the ends, worn by the Clergy. Stops. —Surfaces worked to receive a moulding. (See also Label-stop.) Stoup. —A vessel to hold holy water, usually set in the porch of the church and to the right hand of the person entering. String-course. —A horizontal line of projecting mouldings carried along a building. Surmounted. —In Heraldry having one charge placed upon another. Syse. —Assize = trial. Tempera. —A form of distemper, i.e. mural painting in colours with a medium soluble in water. Terp. —A mound. Throxigh Stone. —A flat-lying or table tombstone. Tincture. —Heraldic colour. Titulus. —An earthern mound in front of an entrance through a rampart. Tonsure. —The shaved crown worn by priests and other religious. Torso. —The trunk of a statue, especially one having the head or limbs mutilated. Trance, Transe. —A corridor ; a passage between buildings. Transepts. —The arms of a building on a cross plan. Transom. —A horizontal bar across a window opening, doorway, or panel. Traverse. —A cross piece. Trefoil. —(i) See Foil. (2) A flower of three petals (Heraldry). Tressure. —A narrow border within the shield (Heraldry). Triglyph. —An ornament consisting of a grooved tablet repeated at equal intervals in the frieze of the Doric order. Tusking. —Stones left projecting from the surface of a building in order that an aditition may be bonded into the existing wall. Tympanum. —An enclosed space in the head of an arch, doorway, etc., or in the triangle of a pediment. Urchin .—See Hurcheon. Vallum. —An earthen rampart. Voussoirs. —The wedge-like stones forming an arch. Withies. —A band of twigs twisted together. Zobmorphic. —Of animal form. 291 INDEX. (The Roman nuynerals refer to the pages of the Introduction, the Arabic numerals to the Articles of the Inventory.) Airswood Moss. " King Schaw’s Grave,” 648. Alais Knowe, Enclosure, 623. Allangillfoot, " Tumulus,” 202. Alms Dish, 58. Altars, Roman, 266, 273, 462. Amisfield : Carved Oak Door from Tower, in National Museum of Antiquities, 578. Colour decoration in Hall of Tower, Ixv, 578. Enclosure, 583. Tower, Ixii et seq., 578. Ancient Marches or Dykes, xix, Ivi, 48, 61, 80, 163, 176, 411, 566. Annan, xxx et seq. : Church of, xxx, xlvi, Ixiv, Ixvii. Inscribed Stone, 4. Mote of, lix, 3. Parish of, 1-8. Tower (Castle) of, xxxi, xl, xlvi. Annfield, Fort (site) near, 588. Applegarth : Market Cross (site) ,31. Monastery (site), 30. Mote, 13. Parish of, 9-32. Archbank, Fort (site), 507. Arkland : ” Fort,” 526. Tower (site), 536. Arkleton : Carved Stones, 226. Enclosure, 220. Armorial Bearings (see Heraldry). Auchen (see Auchencass). Auchenage Burn, 328. Auchenbainzie Hill, Cairns, 530, 534 (site). Auchencairn, Cairns, xlviii, 74, 75. Auchencairn Height, Cairn (site), 83. Auchencass Castle, lx, 384. Auchencat Burn, Fort, liv, 485. Auchencheyne, Cross-Slab at Woodlea Farm from, 253. Auchengassel Castle (site), 535. Auchengruith, Castle Gilmour (site), 571. Auchengruith Craig (supposed site of Church), 574. Auchenrivock Castle, 44. Auldton : Enclosure, 491. Fort (site), 508. Mote, lix, 483. Bail Hill, Cairn (site), 336. Bailie Knowe, 37. Bailiehill, Fort, xx, 640. " Baitablers,” xviii. Balgray Cleuchheads, Construction, 27. Ballaggan, Mote Hill, 160. Bankburnfoot, Fort, 633. Bankend, or Isle Castle, 34. Bankhead: Chapel (site), 206. Entrenchment in Old Graveyard, 200. Bankhead Hill : Enclosure, 190. Entrenchment, 201. Forts, 176, 634. “ King Schaw’s Grave,” xlvii, xlviii, 648. Barclay (and Murray), Arms of, 539. Barjarg Moor, Small Cairns, 328. Barjarg Tower, Ixiii, 327. Barmkin, or Barnekin, Ixi. “ Barnkin of Craigs,” 135. Barntalloch Mote (Barntalloch Castle). 431. Barr Burn, 171. Barr’s Hill, Fort, lii, 581. “ Bartizan,” meaning of, 33 (note on p. 24). Basins, Stone (see Stone Basins). Battie, Ally, 308. Beak-head, 58. Beattock Hill: Enclosure, 412. Forts, 398-401. Beatty, 308 (see Battie). Becks, 442. Belfrys, 58, 96, 155. Bells, Ixviii, 58, 96, 134, 227, 285, 452, 453, 496. Bell’s Tower, xxxii. Belt (the). High Townhead, Stone Fort, liii, 342. Bennoch, James, Martyr Stones, 232, 251. Bent Mote, Tumulus (site), 541. Benthead : ” Cairn ” (site), 82. Fort, 64. Bentpath, Enclosure, 646. Berryscaur, Heraldic Stone, 308. Bessie’s Hill, Forts, 173, 174. Birkshaw, Mound, 254. ” Birren,” meaning of, Ivi. Birren Hill, 434. Birren Knowes, 205. 292 INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. {The Roman numerals refer to the pages of the Introduction, the Arabic numerals to the Articles of the Inventory.) Birrens, Roman Station, xx, li, 462. Birrens Hill, Carruthers, Fort, Hi, 464. Birrenswark : Enclosures, Iv, 272. Fortifications, xx, xxiii, li, 272. Relics found at, Iv, 272. Black Bull Bank, 434. Black Hill, 163. Black Loch, Crannog (site), Ivii, 568. Blackaddie, Inscribed Stone, 565. Blacket House (see Blackwood). Blackball, Enclosure, 213. Blacklaw Tower, 477. Blackshaw, 39. Blackwood, or Blacket House, Ixii, 460. Blatobulgium, xxx, 462. Blindhillbush Hill, Fort, 12. Block, 434. Blountfield, 519. “ Blue Cairn,” Dupple Burn, 363. ” Bogle Walls,” Fort, Ivii, 638. Bogrie : Fort, 140. Tower, 138. Bogrie Moor, Small Cairns, 147. Bogs Burn, Cairn (remains), 561. Bonshaw Tower, Ixiv, i. Boonies, Enclosure, 645. Border Peel, lix, Ixii, Boreland : Enclosure, 413. Name, meaning of, lx. Tower (remains), 387. Bow Hill, 15. Breckonside Tower, Cairn Valley, 234. Breckonside Tower, near Moffat, 475. Bridge, Dumfries, 131. Bridgend, Earthwork (site), 498. Brieryshaw Hill; Enclosure, 219. Fort, 210. Broadshaw Rig, Small Cairn, 417. Brock Linns, Fort, 432. Broomfield, 567. Broomhill, Fort, 297. Broomhill Bank Hill, Forts, 10, ii. Broomrig, " Tumulus,” 585. Bruce Stone from Annan, 4. Bruntshielbog, Cairns and Standing-Stones near, 47, 48. Brydekirk Mains, 2, 6. Buccleuch Arms Hotel, Langholm, 430. Bucklerhole, Tower (site), 521. “ Burrain Skelton,” Fort, 18. Burn Farm, Cairn, 513. Burnbrae, Chapel (site) and Chapel Well near, 122. Burnfoot (Annandale), 474. Burnfoot (Eskdale), Enclosures, 641, 647. Burnhill, Fort (site), 520. Burnswork (see Birrenswark). Burrance Bridge, Fort near, 358. Byeloch, 524. Caerlaverock : Castle, xxxi-xxxiii, xxxvii, xliii, lx, 33 (i), 33 (2). Inventory of Furniture in 1640, Ixv. Parish of, 33-41. " Cairn of Creca ” (site), 8. Caimey Knowe, Small Cairns, 613. Cairns: xlvii, xlix (see also Tumuli). Long Cairns, 47, 249, 329, 351 (remains), 415. Round Cairns, 71-76, 99, 108, 222, 363,371, 372, 421- 424, 513, 529, 560-562, 610-612. Sites, 8, 49, 82-84, 88, 255, 256, 336. 379, 427, 440, 524, 534 . 569. 573 - Small Cairns, 47, 68, 69, 70, 75, 140, 145-147, 243-248, 328,343-348, 361, 362, 417-420, 437, 558, 559, 613. Caldwell Burn, 314. Calfield : Enclosure, 435. Tower (site), 443. Camp Hill, Bailiehill, Fort, xx, 640. Camp Hill, Torthorwald, Fort, 591. Camp Hill, Trohoughton, Fort, 130. Campknowe, Chapel Hill, Fort, 393. Campknowe, Gardenholm, Fort, 390. Canonbie : Grave Slabs in Churchyard, 42. Parish of, 42-57. Priory (site), xxxiv, 51. Tomb in Churchyard, Ixvii, 42. Capel Glen, Cairn, 71. Capenoch : Foundations of Tower, 330. Long Cairn, xlvii, 329. Carlyle, William of, 134. Car’s Wood, Enclosure, 360. Carruthers: Fort, Birrens Hill, lii, 464. Old Graveyard, 470. Village Site, 473. Carruthers, George, of Holdmains, 96. Carruthers, Sir Simon, supposed Effigy of, 517. Carterton : Chapel (site) near, 312. Forts, 294, 295. Carterton Knowes, Hut Circle, 195. Carthur Hill, Fort, liii, 291. Carved Cross and Mark, 386, 590. Castellated and Domestic Structures, Iviii et seq., i, 2, 33 (i), 33(2), 34 . 43. 44 . 59 . 60, 89, 90, 106, 107, III, 136-138, 156, 157, 164, 233-235, 280-282, 287, 288, 315, 327, 330, 337, 338, 357, 368, 384, 385, 387, 388, 429, 430, 445 (i), 445 (2), 446, 447, 460, 475-480, 494, 510, 515, 537, 551, 552, 563, 578.590, 615. Sites, 7, 53-57, 103, 123, 135, 167, 169, 170, 259, 260, 268, 269, 276, 278, 334, 354, 364, 365, 380-382, 441, 443 . 458, 521. 535 . 536. 542, 546. 547. 570. 571. 572. 575. 576, 608. Castle Dykes (see Dumfries). Castle Gilmour, Auchengruith (site), 571. Castle Hill, Dalwhat Glen, Fort, lii, 236. Castle Hill, Dryfe Water, Fort, 289. Castle Holm, Langholm, Castle, 429. Castle Loch, Lochmaben, Iviii. Castle O’er, Forts, etc., xx, liv, Iv, 175-190. Castlehill (Duncow) : Fort, 341. Standing-Stone, 350. Standing-Stone (“ Peerie’s Kist ”), 349. Castlehill (Durisdeer), 170. Castlehill (Eskdalemuir), 187. Castlehill (Lochmaben) Castle (site), 458. Castlehill (Pilmuir Common), Fort and Enclosures, liv, 113, 117, 118. Castleknowe, Saughtrees, 618. Castlemilk (site), Ixi, Ixii, 546. 293 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. {The Roman numerals refer to the pages of the Introduction, the Arabic numerals to the Articles of the Inventory.) Castle Robert (site), Corsebank, 334. Catherine’s Hill, Poldean, Fort, 620. Cauldkinefoot, Enclosure, 647. " Celtic Dike," the (see “ Deil’s Dike ”). Chalkyhill Wood, Fort, 369. Chapel, St Cuthbert’s Chapel at, 3S3. Chapel Croft, Chapel (site), 455. Chapel Hill, Fort, Campknowe, 393. Chapel Lea, Chapel (site), 632. Chapels (see Ecclesiastical Structures). " Chapman’s Stone,” Newbigging, 629. Chaxteris of Amisfield, xxix, 578, 586. Churches (see Ecclesiastical Structures). Cists, xlt-ii, 71 (supposed), 75, 222 (supposed), 272, 363, 371, 611, 648. Clackleith, Fort, 553. Cleuchfoot, 444. Cleuchhead, Fort, 161. Cleuchheads Hill, Fort, 18. Clochmabanestane, 263. Clochmabon, 261. Clonfeckle, Long Cairn (remains), xlvii, 351. Closeburn : Alms Dish (brass) at Manse, 58. Beak-head, 58. Bell, Ixviii, 58. Castle, Ixi, 59. Church, 58. Cross-Shaft, etc., 58. Cross-Shaft at Grierson Museum, Thornhill, Ixviii, 514. Font from Dalgarnock, 58. Gravestones, 77. Parish of, 58-88. Closs, 313. Coats Hill: Fort, 394. Mote, lix, 395. Small Cairns, 418. Cocklawrig, Enclosure, 300. Cockpool Castle (site), xix, Ixi, 542. Cogrie, Enclosure near, 414. Comlongon Castle, xix, Ixi, Ixv, 537. Comyn’s Castle, Dalswinton (site), 354. Conder, Gilbert, Tombstone of, 540. Connel’s Well, 616. Conrig Hill, 561. Coped Stones, 109, 265. Corbet, John, Martyr’s Tomb, 586. Corehead, Enclosure, 492. Cornal Burn, 501. Comal Tower, 476. Corncockle, Crannog (remains) and Relics, Iviii. Corncockle Plantation, Fort, liv, 449. Corrie Common, Fort (remains of), 293. Corrie Old Churchyard, Gravestones, etc., 309, 310. Corriehills, Corrie Church (site) ,311. Corsan Monument, St Michael’s Churchyard, Dumfries, 13.8- Corsebank, 334. Coshogle Castle : Doorway of, 164. Heraldic Stone, 165. Site of Castle, i 6 g. Cot, Stone Circle, xlviii, r98. "Covenanters’ Graves,” 314 (site). Cowhill : Entrenchment, 283. Tower, -xl, 282. Craes Hill, Small Cairns, 243. Craig Hill, Fort, Iv, 637. Craig Wood, Fort, 36. Craigdasher, Small Cairns, 145. Craigencoon, Cairns, 610, 611. Craighaugh Holm, “ Hislop’s Grave,” 203. Craighoar Hill, 419. Craighouse, Enclosure, 303, Craighousesteads Hill, Fort, liv, 600. Craigieburn, Fort (site), 500. Craigielandshill, Cairn (site), 427. Craigshiels, Small Cairns and Hut-Circle, 361. Crannog, Ivii, 568 (site) (see also Lake Dwelling). Crawfordton, Peel Tower (site), 259. Crawick Water, Entrenchment or Mote, 553. Crawthat Cottage, Fort, liv, 595. Crawthat Hill, Fort, 602. Creca, Cairn (site), 8. Crichope Linn, Fort, 63. Crichton, William, Rector of Sanquhar, 565. Crofthead, 501. Crooks, Enclosure, 644. Crossford Hill, Small Cairns, 248. Crosses, Ixvii. Cross-slabs : As Boundary Mark, 253. In Graveyards, 274, 310, 439, 469, 470, 518, 550. Formed of Soil, 564. Free-standing, 374, 378, 538. Incised on rock, 386. Market Crosses, 252, 563 (remains). Shafts, Sockets, and other Fragments, 5, 58, 78, 250, 273, 275. 332, 333 . 514. 531 - Sites, 31, 502, 574. Crowdie Knowe, 470. Crunzierton Wood, Enclosure, 322. Cuil Plantation, Enclosure, 642. Cummertrees : Church (site), 94. Parish of, 89-95. Cumstone Burn, Fort, 15. Cup Marks, 284. Dalbate : Enclosure, 605. Fort, 602. Dalgarnock : Church and Well (sites), 85. Font of the Old Church, 58. Gravestones, 78. Dalmakethar, Forts, 19, 20, 21, Dalswinton, xxxiii. Chapel (site), 353. Comyn’s Castle (site), 354. Old House, 338. Dalton : Bell, Ixviii, 96. Fort (site), loi. Little Dalton Church, 97. Little Dalton (site), 104. Mickle Dalton Church, Ixvii, 96. Parish of, 96-105. Standing-Stone, 100. Dalveen House, Heraldic Stone, 166. 294 INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. (The Roman numerals refer to the pages of the Introduction, the Arabic numerals to the Articles of the Inventory.) Dalveen Pass, 163. Dalwhat, 244. Dalwhat Glen, Fort on Castle Hill, lii, 236. Davington, Fort (site) near, 207. " Debateable Land,” xviii, xxxiv et seq. Defensive Constructions (see Forts and Motes). " Deil’s Dike,” Ivi, 61, 80, 163, 566. *' Deil’s Jingle,” Ivi, 176. Dinning, Mote, 65. Dinwoodie Castle, Heraldic Stone at Dinwoodie Mains, from, 28. Dinwoodie, Enclosure, 25. Dinwoodie Mains, Heraldic Stone, 28. Dod (the). Fort, 621. Doe’s Hill, Fort, 461. Door, Carved, 578. Dormont Rig, Fort, 295. Dornock : Coped Stones, 109. Parish of, 106-110. St Marjory’s Church (site), no. Dornock, Battle of, xxxiii. Douglas Farm, Fort, 26 !j. Douglas and Johnstone, Arms of, 497, 631. Douglas, John, of Stenhouse, Gravestone of, 614. Douglas, Robert, of Coshogle, 165. Douglas, of Dalveen, 166. Douglas, of Drumlanrig, 156. Douglas, of Mouswald, 517. Druidhill Burn, Mote, 525. Drumlanrig : Castle, xlvi, 156. Fort, 158. Sundials, 156. Yett, 156. Dryfe Lodge, Enclosure, 299. Dryfeholm, 115, 126. Dryfesdale : Fort, 114. Parish of, 111-126. Dryfesands, last Clan Battle on the Borders, xliv. Duff Kinnel Bank, Enclosures, 323, 324. Dumfries ; Architectural Fragments, 132. Bridge, 131. Castle Dykes (Dumfries Castle), xx, xxxi, xxxii, lix, 128. Gravestones, St Michael’s Churchyard, 133. Midsteeple, 127. Mote of, 129. Parish of, 127-135. Duncow Burn, 351. Dundoran, Fort, 619. Dunnabie, Fort, 467. Dunscore, Parish of, 136-151. Dupple Burn, " Blue Cairn,” 363. Durisdeer ; Castle (site), xxxii, 170. Church, Ixvii, 152. Earthwork near, 162. Fort, 163. Grave-slab, 153. Martyr’s Tomb, 154. Mural Monument in Church, 152. Parish of, 152-170. Wrought-iron Stand and Bracket in Church, 152. Dyke, Standing-Stones, 426. Dykes, xix, Ivi, 48, 61, 80, 163, 176, 566. Earshaig Lake, 424. Earthworks, Mounds, and other Constructions of doubtful character, 27, 79, 91, 148, 162, 175, 185, 188, 189, 200, 201, 237, 240, 254, 283, 317, 396, 409, 472, 494. 526, 527, 553, 585, 618, 626. Sites, 38, 39, 50, 498, 532. East Earshaig : Cairns, 424. Fort, 402. East Morton, Fort, 512. Ecclefechan, St Fechan’s Church (site), 277. Eccles House, Entrenchment, 527. Ecclesiastical Structures, Ixvi, 58, 96, 97, 152, 155, 224, 229, 271. 332, 355 . 367. 383. 438, 543 . 548. 577 - Sites, 6, 29, 30, 40, 41, 50-52, 85-87, 94, 95, 104, 105, no, 119-122, 150, 151, 168, 206, 228, 257, 258, 267, 277. 279, 286, 311, 312, 326, 331, 352, 353, 442, 444, • 455-457. 459 . 474 . 574 . 594 . 616, 632, 649. Edgar, Robert, Martyr Stones, 232, 251. Edgemoor, Enclosure, 325. Effigies, 152, 154, 517, 549. Elbeckhill, 623. Elfknowe, 522. Elliock, Fort, South Mains, 555. Elliock House, 552. Elliock, Small Cairns on Moor, 559. Elliot, Adam, of Meikledale, Gravestone of, 225. Elliot, Walter, of Arkleton, Inscribed Stones, 226. Ellrig Wood, 244, 245. Elshieshields Tower, Ixii, 447. Enclosures, Iv, 22-26, 46, 117, 118, 179-194, 212-221, 299- 306, 322-325, 360, 412-414, 433-436. 468, 482, 484, 488-493, 545, 583, 605, 623, 624, 633, 641-647. Sites, 204, 205, 501. Enoch Castle (site), 167. Enzieholm, 638. Ericstane : Forts, 486, 487. Sundial, 495. Eskdalemuir : Enclosures, 187, 204 (site). Parish of, 171-208. Eskrig, Forts (sites), 125. Euchan Cottage, Fort near, 554. Euchan Water, " Kemp’s Castle,” Fort, Ivii, 557. Ewes : Bell, Ixviii, 227. Parish of, 209-228. St Cuthbert’s Church (site), 228. Fairholm, Fort (site), 124. "Fairy Craig,” Birrenswark, 272. Farthingwell: Church (site) near, 151. Laird of Lag’s Tomb, near, 149. Fellend, Small Cairn, 69. Ferguson of Isle, 337. Fiddleton, Enclosures, 212. Fingland, Enclosure, 179. Fir Tree Hill, Fort, 16. Fireplaces, i, 33 (2), 90, 446, 447, 510, 537, 578. Flecket Hill, 502. Fleming, Adam, Gravestone of, in Kirkconnel Churchyard, 373 - 295 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. (The Roman numerals refer to the pages of the Introduction, the Arabic numerals to the Articles of the Inventory.) Fleuchlarg, " White Cairn,” xlvii, 249. Fogo Famih', Burial Place of, 439. Font, 5S. Forrester, Arms of, 132. Forts, 1-lv, 3, 9-12, 14-21, 35-37, 45, 61-64, 98, 112-116, 130. 139-143. 158, 159, 161, 163, 171, 172, 173, 174, 176, 177, 178, 209-211, 236, 237, 240, 242, 261, 262, 272, 289-295, 297, 298, 318-321, 339-342, 358, 359 . 369. 370. 389-394. 398-408, 410, 411, 432. 449-451. 461-467. 481. 485-487. 511. 512, 516, 528, 544, 554, 555, 557, 579-581, 591, 595-602, 609, 619-622, 633-640. Sites, 32, loi, 102, 124, 125, 207, 270, 366, 428, 454, 499 . 500, 503-509. 520, 521, 533, 567, 588, 589, 593. 607. Fourmerkland Tower, 280, Frenchland : Foundations, 494. Tower, Ixiii, 480. Friars’ Carse ; Artificial Island and Relics, Iviii. Fort, 142. Monastery (site), 150. Frieze, painted, Ixv, 578. Furniture, Ixv. Gallaberry, Fort, liv, 115. Gardenholm, Forts, 390, 391, 392. Garpol Water : " Camp,” 396. Mote, lix, 397. Garrs Hill, Fort (site), 593. Garvald : Church, Ixvii, 355. Gravestones in Churchyard, 356. Garwald, 180, 181. Gates, Iron (see Yetts). Gateside, Fort, 143. Gawin Moor, Cairns, 75. Gibson, John, Martyr Stones, 232, 251. Gillesbie Tower, 287. Gillrig, 360. Gilnockie: Castle (site), 50. Mound and Ditch (site), 50. Roman Camp, xx, xxx, li, 45. Tower (see Hollows Tower). " Girdle Stanes,” Stone Circle, xlviii, 198. Girharrow, Hut-Circles and Small Cairns, 246, 247. Glass: From Church Foundations (Mediaeval), bcvii, 271. From Roman Fortifications and Camp, 272, 462. Gledbrae, Cairn (remains), 612. Gledenholm, Small Cairns and Hut-Circles, 347, 348. Glenae Tower, 357. Glencairn : Church, Ixvii, 229. Martyrs’ Tombs, 232. Mural Monuments, 230, 231. Parish of, 229-260. Glencaple, 40. Glendinning Cleuch, " Fort,” 526. Glenesslin School, Chapel (site) near, 258. Glengenny Moor, Cairn (remains), 560. Glenmaid, Small Cairns and Hut-Circles, xlix, 343. Glenrae Castle (site), 570. Glentenmont, Enclosure, 436. Goose Hill, Castle (site) near, 575. Gordon of Lochinvar, 138. Gotterbie Moor, Fort, lii, 451. Gowanlea, St Thomas’ Chapel (site) near, 457. Graham of Blawatwood, io6. Grahams of Dunnabie, Burial-Place of, 309. Graham, William, and Margaret Irvine, 376. Granton, Enclosure, 484. Gravestones (see Tombstones and other Memorial Slabs). Greenhill Plantation, Fort (remains), liii. Grennan Hill : Entrenchment, 532. Fort, 528. Gretna, Parish of, 261-270. Gretna Green : Coped Stone, 265. Grave-Slab, 264. Gretna Hill, Fort (site), 270. “ Grey Mare’s Tail,” Forts near, 64, 481. " Grey Wether,” Standing-Stone, 223. Grierson (and Lawrie), Arms of, 235. Grierson (and Murray), Arms of, 93. Grierson Museum, Sculptured Stones at, Ixviii, 514. Grierson, William, Martyr’s Tomb in St Michael’s Church¬ yard, Dumfries, 133. Griersons of Lag, 136, 149, 517. Gufhill Rig, Cairn, 76. Haggy Hill, Fort, liv, 596. Halfmerk Hill, 335. Halldykes, 119. Hallgreen, Priory (site), 51. Hallguards (site), 278. Hallmuir, Fort, 112. Hamlin Knowe, Enclosure, 193. Hangingshaw, Enclosure, 24. Harelaw Tower (site), 56. Hartfell, Lord, Arms of, 496. Hartmanor : Stone Circle, xliii, 199. " Stone Rings,” 185. Hastings Hall, Moniaive, Cross-Shaft, 250. “The Haunches,” 49. Heathery Plantation, Small Cairn, 245. Hell’s Hole, Fort, 402. Hencastle Rig, Fort, 292. Henwell, Enclosure, 433. Heraldic Stones, 28, 93, 127, 132, 165, 166, 225, 308, 375, 497 . 517. 539 . 540. 584. 586, 592, 615, 631. Heraldry : Arms of : Carruthers, 96. Charteris of Amisfield, 578, 586. Douglas (and Johnstone), 165. Douglas (Killywarren), 615. Douglas of Coshogle, 165. Douglas of Dalveen, 166. Douglas of Drumlanrig, 156. Douglas of Mouswald, 517. Edmonstoun (?), 132. Elliot, 225. Ferguson, 337. Forrester, 132. Gordon of Lochinvar, 138. 296 INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. {The Roman numerals refer to the pages of the Introduction, the Arabic numerals to the Articles of the Inventory.) Heraldry— continued : Arms of— continued : Grierson of Lag, 149. Hartfell, Lord, 496. Irving, 375. Jardine (and Lockhart), 446. Johnstone, 165, 264, 446, 495, 496, 537, 606, 628. Johnstone (and Douglas), 497, 631. Johnstone (and Rollo), 495. Kennedy, William, Abbot of Crossraguel and Com mendator of Holywood, 285. Kirko, 137. Kirkpatrick, 77. Lawne (and Grierson), 235. Lawrie (and Riddell), 235, Mar, 33 (2), 165, 166. Maxwell, 33 (2), i38(?), 280, 282, 308, 578. Maxwell (and Douglas), 34. Maxwell (and Mundell), 584. Morton, John, Lord Maxwell, Earl of, 33 (2). Murray, 537, 539. Murray (and Barclay), 539. Murray (and Grierson), 93. Nithsdale, Robert Maxwell, First Earl of, 33 (2). Queensberry, Douglas, Marquis of, 156. Scotland (Scottish Arms), 33 (2), 127, 308, 537. Stewart of Dalswinton, 33 (2). Wilson (and Young), 540. Young, 540. Herries, Lord, xxvii, 89, 90. High Townhead, Stone Fort, liii, 342. Highmains Hill, Fort, 36. Hightown Hill, 580. Hillhead, Forts, 407, 408. Hillside, 120. " Hislop’s Grave,” Craighaugh, 203. Hizzie Birren, Enclosure, 643. Hoddom Castle, xxviii, xxxix, xl, Ixi, 90. Grave Slabs at, 92. Hoddom; xxi. Church, Foundations, Ixvii, 271. Cross-Slabs, 274. Cross-Socket, 275. Mediaeval Glass, Ixvii, 271. Parish of, 271-279. Peel Houses in, Ixii (k.). Roman-dressed Stones, 271. " Hoddomstanes,” 90. Hog-backed Stone, 377. Holehouse Linn, Forts, 389, 428 (site). Hollows Tower, xxxv, Ixii, 43. Spiral-marked Slab at, 1 , 43. Holmains : Cairn, 99. Fort, “ Range Castle," 98. Tower (site), 103. Holmhead Hill, Small Cairns, 244. Holmhouse, Cairn (remains), 612. Holywood ; Abbey (site), Ixvii, 286. Bells, Ixviii, 285. Parish of, 280-286. Stone Circle, " Twelve Apostles,” xlviii, 284. Honeyhole, " White Cairn,” 529. Horseclose, Fort (remains) in Greenhill Plantation, liii. Horse Park, Maxwellton, Fort (supposed), 242. Hound Rig, “ Blue Cairn,” 363. Howthat Burn, Enclosure, 22. Hume, Marie Menzies or. Gravestone of, 356. Hunterheck : Enclosure, 490. Fort (site), 505. Hut-Circles, xlix, 66, 67, 195-197, 247, 343-348, 361, 416, 461, 604, 605. Hutton, Mote of, li.x, 296. Hutton and Corrie : Cross-Slab in Corrie Old Churchyard, 310. Gravestone in Corrie Churchyard, 309. Parish of, 287-314. Ingleston: Lower Mote, lix, 238. Martyr’s Monument, 251. Name, meaning of, lx. Ingleston Moor, Small Cairn, 417. Inscribed Stones, 4, 127, 226, 271, 273, 375, 376, 425, 462, 565. 592. Irving, Arms of, 375. Irving Monument, St Michael’s Churchyard, Dumfries, 133- Irving of Bonshaw, i, 106. Irvings of Robgill, 107. Irvings, Stakeheugh, 44. Iron-Work, 127, 152, 156. Isle or Bankend Castle, 34. Isle Tower, Ixiii, 337. Jardine, Arms of, 446. Johnston, James, of Nether Castlehill, Tombstone of, 606. Johnstone: Enclosure, Johnstone House, 182. Parish of, 315-326. Johnstone, Ambrose, of Poldean, 631. Johnstone, Andrew, in. Johnstone, Arms of, 165, 264, 446, 495, 496, 497, 537, 606, 628, 631. Johnstone, John, Gravestone with Arms at Gretna Green, 264. Johnstone, Mungo, in. Johnstone, Robert, 495. Johnstones of Corehead, 388. Johnstonecleuch, Site of Chapel near, 326. Keir, Parish of, 327-331. Kelloside, Ivii. Kelwood, Fort, 37. " Kemp,” meaning of name, Ivii. " Kemp’s Castle,” Fort, Ivii, 557. Kennedy, William, Abbot of Crossraguel and Commendator of Holywood, 285. Kentigern, .xx, xxi, xxii. “ Kilblain” (site), 41. " Kilbride ” (site), 331. Killywarren, 615. Kilncroft, 286. Kilpatricks of Closeburn, 59. “ King Schaw’s Grave,” xlvii, xlviii, 648. Kingsholm, 128. Kinmont Tower (site), 55. Kinnelhead ; Cairns, 422, 423. Cross (incised on rock), 386. Small Cairns, 419. Structure, 385. 297 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. (The Roman numeral refer to the pages of the Introduction, the Arabic numerals to the Articles of the Inventory.) Kinnel Water, Fort, 404. Kirk, James. Mart3T’s Tomb in St Michael's Churchyard, Dumfries, 133. Kirk Burn, Fort, li, 163. Kirk (Kirko) of ISogrie, 13S. Kirk (Ivirko) of Sundaj'rvell Tower, 137. Ivirkbog, Standing-Stone, 81. Kirkbride Church, 155. Kirkbride Mains, 331. Kirkconnel: Church, Springkell, 367. Grave-slabs, 373. Kirkconnel (site), 380. Parish of, 332-336. IvLrkconnel Tower (site), xlii, 276. Kirkhill Cottage, Fort (remains), 319. KLrkhill, Stone Circle, xlviii, 625. Kirkholm HDl, Fort (site), 32. Kirkland, 364. ICirkland Hill, 332, 358. Kirklands, 85. Kirkleys, Fort, 159. Kirkmahoe : Fort near Manse, li, 340. Parish of, 337 - 354 - Kirkmichael, Parish of, 355-366. Kirkmichael Fell, Fort, 359. Kirkpatrick (site of Chapel), 87. Kirkpatrick : Inscribed Lintel, 375. Tower (site), 381. Kirkpatrick, John, Gravestone of, 77. Kirkpatrick, Major John, 614. Kirkpatrick-Fleming, Parish of, 367-382. Kirkpatrick-Juxta: Inscribed Stone at Manse, 425. Parish of, 383-428. Kirkstyle, Ewes : Bell, Ixviii, 227. St Cuthbert’s Church (site), 228. Kirkton (Dryfesdale) : Church (site), 121. Tower (site), 123. Kirkton (Kirkmahoe), St Quintin’s Church (site), 352. Kirktonhill, Church (site), 649. Knock Hill, Fort, 406. Knock Hill, Mound, 79. Knockbrack, Small Cairns, 70. Knockenhair, 562. Knockenshang : Cairn, 76. " Tumulus ” (site), 84. Knockespen : Glenae Tower, 357. Small Cairns, xlix, 362. Knockhill, Cross Fragments and Inscribed Stones at, Ixviii, 273- Knowe, The, (Castle O’er), Fort, 178. Lag Tower, Ixii, 136. Lag’s Tomb, 149. “ The Lake," 171. Laird’s Houses, Ixi. Lake Dwelling, Ivii, 144 (see also Crannog). Langholm : Buccleuch Arms Hotel, 430. Langholm Castle, xxxiv, xxxvii, xxxviii, xli, 429. Parish of, 429-444. Langknowe, Fort, 159. Laurie, of Maxwelton, Stephen, Mural Monument at Glen- cairn, 230. Laverhay: Earthwork, 626. Fort, 622. "Law Birren,” 192. Lawrie, Arms of, 235. Leithenhall, Enclosures, 624. Linnhall, Hut-Circles, 604. Little Dalton : Church, 97. Site, 104. Little Hutton Chapel (site), 119. Loch Hill, Fort, Iv, 211. Loch Urr : Entrenchment, 148. Lake Dwelling, 144. Lochan Burn, Cairns, 419, 423. Lochbank, 450. Lochhouse Tower, Ixii, 388. Lochmaben : Artificial Island in Castle Loch, Iviii. ■Bells, Ixviii, 452, 453. Castle, lix, lx, Ixi, 445 (2). Church (site), 456. Old Castle, xxxi-xxxiii, lix, 445 (i). Parish of, 445-459. " Lochmaben Stane,” xlviii, 263. Lochside : Crannog (site), 568. Small Cairns, etc., 558. Lochwood: Mote, "The Mount,” lix, 316. Mound, 317. Tower, xix, xxiv, Ixi, Ixv, 315. Lockerbie, Tower, in. Lockhart (and Jardine), Arms of, 446. Longerhallis Hill, 21. Longknowe, Enclosure, 181. " Loupin’ Stanes,” Stone Circle, xlviii, 199. Low Auldgirth, 60. Luce Church (site), 279. Luce, Heraldic Stones from Spedlin’s Tower at, 446. Lukup, William, Master of Works, Drumlanrig, 154, 156. Lunelly Tower, 288. Lyneholm Hill, Fort, 639. McMich(a)el, Daniel, Martyr’s Tomb at Durisdeer, 153. McMount, Tumulus (site), 84. Macrieholm Knowe, Enclosure, 46. Mains (Tower), in. Malls Castle (site), 547. Manse of Sanquhar, Cairn (site), 569. Mansfield, Earl of, 539. Mar, Earls of, 33 (2), 165, 166. Marches or Dykes, Ancient (see Ancient Marches or Dykes). Mark, Incised, 590. Market Crosses ; Applegarth (site), 31. Moniaive, 252. Sanquhar, 563 (remains). Martyrs’ Tombs, 133, 153, 203, 232, 251, 314 (site), 586, 614. Maxwell, Arms of, 33 (2), 34, 138 (?), 280, 282, 308, 578, 584 - Maxwell, John, Master of, 89. Maxwell of Barncleuch and Glengaber, 133. Maxwell House, Architectural Fragments from, 132. 298 INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. {The Roman numerals refer to the pages of the Introduction, the Arabic numerals to the Articles of the Inventory.) Maxwell, Robert, First Earl of Nithsdale, Arms of, 33(2). Maxwells of Caerlaverock, 33 (i), 33 (2). Maxwellton : Fort (supposed). Horse Park, 242. Mote, lix, 241. Maxwelltown Museum, Bell in, Ixviii, 134. Maxwelton House, 235. Meggat Glen, 644. “ Meg Tod’s Mote,” Moffat, 482. Meikledale : Enclosures, 216, 217. Standing-Stone, 223. Meikeholmside, Enclosure, 493. Mellingshaw Tower, 479. Mennock Pass, Cross, 564. Menzies, Marie, or Hume, Gravestone of, 356. Merkland Cross, Ixvii, 378. Mickle Dalton {see Dalton). " Mid Cairn,” 75. Mid Hill, Enclosure, 26. Mid Raeburn, Enclosure, 184. Mid Steeple, Dumfries, 127. Middlebie, Parish of, 460-474. Middlebie, Peels in, Ixii (w.). Middleholm Hill, Cairn (site), 440. Millbank, Fort, liv, 14. Milligan, John, 515. Minsca, Fort, 465. Miscellaneous Constructions and Ruins, 27, 50, 79, 80, 330 385. 472. 494. 526, 532, 583, 605. Mitchell, Robert, Martyr Stones, 232, 251. Moat Brae, Dumfries, 129. Moat Farm, 143. Moat House, Annan, 3. Moatland, Mote, 239. Moffat: Bell, Ixviii, 496. Forts (sites), 499, 506-508. Heraldic Stones, 497. Motes, lix, 482 (?), 483. Parish of, 475-509. St Cuthbert’s Chapel, Ixvii, 383. Mollin, Forts, liv, 320, 321. Moniaive : Cairn (site), 255. Cross-Shaft, 250. Mercat Cross, 252. Monteith, James, 227. Moorcleugh, Chapel (site), near, 168. Moodlaw, Enclosure, 183. Moreland, Fort (site), 504. Morrison House, Fort (site), 102. Morton Castle, xxxiii, l.x, 510. Canoe found near, Iviii. Morton Church (site), 52. Morton, Earl of, xxvi, 33 (2). Morton, East, Fort, 512. Morton Mains Hill, Fort, lii, 511. Morton, Parish of, 510-514. " Moss Castle,” 91. Mosshead Hill, Fort (remains), 597. " Mosskesso ” (site), 313. Mossknow : Cairns, xlviii, 371, 372, 379 (site). Hog-backed Stone (remains), 377. Inscribed Lintel, 376. Mosspeeble, Enclosure, Ivi, 215. Mote Cottage, Fort, 318. Mote Knowe, 409. Motes, Iviii et seq., 3, 13, 65, 128, 129, 160, 237-239, 241, 296, 316. 395 . 397. 431. 448. 483. 525. 556. 582, 617. Mounds, 50, 79, 254, 317, 472, 526. ” Mount, The,” Lochwood, Mote, lix, 316. Mouswald : Armorial Stone and Effigy at Church, 517. Fort, Panteth Hill, 516. Mouswald Place, 515. Parish of, 515-524. Tombstones, etc., at Ruthwell U.F. Church, 518. Muckle Snab Hill, Enclosure, 468. Mullach, Vitrified Fort, liii, 339. Mumbie Tower (site), 54. Mundell (and Maxwell), Arms of, 584. Municipal Structures, 127, 563. Mural Monuments, 152, 230, 231, 592. Murray, Arms of, 93, 537, 539. Murray, William, and Jean Grierson, Arms of, 93. Murraythwaite, " Moss Castle,” 91. Murraythwaite House, Carved Lintel, 93. " Murthat Dike ” {see “ Deil’s Dike ”). Nether Dod Cairns, 73. Small Cairns, 68. Nether Hutton, Mote, 296. Nether Kirkcudbright, St Cuthbert’s Chapel (site), 257. Nether Mains, Chapel (site), 86. Nether Mill, Fort (site), 366. New Woodhead, 45. Newbie Castle (site), 7. Newbie Mains, 7. Newbigging, Applegarth, 22. Newbigging, Wamphray, Standing-Stone, 629. Newfarm, 409. Newfield, 39. Newhall Hill, Forts, 601. Newland Hill, Forts, liv, 599. Nith Bridge : Cross-Shaft, 531. Fort (site), Virginhall Plantation, 533. Nithsdale, Earl of, 33 (2). Norwold Farm, Malls Castle (site), 547. Nutholm Hill; Enclosure, 545. Fort (remains), 544. Old Auchenbrack, Small Cairns, 613 Old Barr, Fort, 554. Old Craighouse, Fort, 292. Old Crawfordton, 233. Cairn (site), 256. Old Gretna, 263. Old Hill, Fort, 209. Old Irvine : Enclosure, 434. Small Cairns, 437. Orchard, Cross-Socket, 333. Orchard (the), Snade, " Mote,” liii, 237. Outer Woodhead, Tower (site), 5-’ Over Cassock, Fort, Iv, 171. Over Rig, 175. Painting, Ixv, 578. Panels, Carved {see also under Heraldry), 584, 615, 631. 299 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. (The Rovian numerals refer to the Panteth Hill, Fort. 516. Park Hill. Hillhead, Fort, 408. Parkcleuch, Fort (remains), 293. Patersons in Skipmyre, Gravestones of, 587. Peat Hill, Enclosure, 301. Peat Hill (Kinnelhead), Structure, 385. Peel, lix, Ixii. " Peerie’s Kist,” Standing-Stone, 349. Pennershaughs Church (site), 474. Pennershaughs Graveyard, 469. Penpont, Parish of, 525-536. Peterburn, Cairns and Standing-Stones near, 47. Pilmuir Common, 113, 117. Pingle Bridge, Chapel (site) near, 51 a. Pinnacle Hill, Fort, 579. Place of Craigs, Tower (site), 135. Poldean : Fort on Catherine’s Hill, 620. Heraldic Panel, 631. Standing-Stone, 630. Portrack Castle, 281. Purdomstown, Fort, li, 466. Pyatshaws Rig, Enclosure, 304. Queensberr}', James, Second Duke of, and his wife. Mural Monument at Durisdeer, 152. Queensberr}-, Marquis of. Arms, 156. Quern, 272. Raeburn, Hut-Circles, 197. Raeburnfoot: Roman Camp, xx, xxx, li, Iv, 172. Tumulus (site) near, 208. Raecleuch Tower, 478. Raehills ; Enclosures, 322, 323, 324. Tower (remains), Boreland, 387. Raggiewhate, Fort, 116. ‘‘ Range Castle,” Fort, liii, 98. Range Castle Hill, Fort, 290. Red Kirk (site), 267. Redhall Castle (site), 382. Relics recovered, Iv, Iviii. From Caerlaverock ; Pottery, 33 (i). From Cairns : Charred wood, 70. Flint chips, 70. Flint implement, 75. Stone disc, 70. Um (beaker), 75. Urn (cinerary), 255. From Church Foundations at Hoddom : Cross-slabs (mediaeval), 274. Mediaeval glass, 271. Roman-dressed stones (one inscribed), 271. From Crannogs ; Canoes, Ivii, Iviii. Hammer-head of whinstone, Iviii. Oak bowls, Iviii. Oak mallet, Iviii. Pottery, Iviii. From Forts, etc. : Bronze Bridle-bit, 272. Bullets of lead (glandes), 272. Coins, 272. Doha or Amphorce, of Roman origin, 172. Glass : Armlet of glass, 272. Window glass, 462. Arabic numbers to the Articles of the Inventory.) Relics recovered— continued: From Forts, etc. — continued : Iron fragments, 272. Quern (broken), 272. Roman inscribed stones (altars, etc.), from Birrens, 462. Samian ware, 272. Stone ball, 272. Tiles (floor and roofing), 462. Repentance Tower, xli, Ixi, 89. Riddell (and Lawrie), Arms of, 235. Rig Hill, Small Cairns, 437. Rigfoot (Annandale), 410. Rigfoot (Meikledale), Enclosure, 218. Robgill: Cairn, 108. Promontory Fort, 370. Tower, Ixii, 107. Rock Sculptures, xlix. Rockhall, Mote, lix, 448. Rockhall Moor, 519, 523. Rockhallhead : Fort (site), 454. Rokele Chapel (site), 459. Roger Gills, Fort (site), 503. Rokele Chapel (supposed site), Rockhallhead, 459. Rollo, Isabella, 495. Roman Altars, 266, 273, 462. Roman Camps and Stations, 1 et seq., 45, 172, 272, 462. Roman-dressed Stones, 271, 462. Roseburrain, Fort, 17. Rosetta, Cairn (supposed), 421. Ross Castle (site), 365. Ross Mains: Fort, 318. Ross Castle (site), 365. Rough Castle Hill, Enclosures, 180. Rough Island, Loch Urr, 144. Roundstonefoot, Enclosure, 488. Royach Cairn (site), 88. Royal Arms of Scotland (see Heraldry). Ruins (see Miscellaneous Constructions and Ruins). Runic Monument, xxii, Lxvii, 538, and Appendix. Ruthwell: Cross, xxii, bcvii, 538. and Appendix. Heraldic Stones and Tombstones, 539, 540. Parish of, 537 - 542 - Ruthwell U.F. Church, Mouswald, Tombstones, etc., 518. Ryecastle, Enclosure, 23. Ryehill: Castle (site), 576. Mote, xxvi, 556. St Bride’s Chapel (site), 444. St Bridget’s Chapel (site), 105. St Bryde’s Kirk (site) and Well, 6. St Bryde’s Tower, 2. St Christopher’s Chapel, Architectural Fragments from, 132. St Christopher’s, St Mary Street, Dumfries, Architectural Fragments at, 132. St Columba’s Chapel (site) and Well, 40. St Connel’s Chapel (site) and Connel’s Well, near Kirk- connel, 616. St Connel’s Church (Kirkconnel Parish), 332. pages of the Introduction, the 300 INVENTORY OF MONUMENTS IN DUMFRIESSHIRE. (The Roman numerals refer to the pages of the Introduction, the Arabic numerals to the Articles of the Inventory.) St Connel’s Grave, Half Merk Hill (site), 335. St Cuthbert’s Chapel, Moffat, Ixvii, 383. St Cuthbert’s Chapel, Nether Kirkcudbright (site), 257, St Cuthbert’s Church (site), Kirkstyle, Ewes, 228. St Fechan’s Church, Ecclefechan (site), 277. St Marjory’s Church (site), no. " St Marjory’s Cross,” 5. St Michael’s Chapel (site), 120. St Michael’s Churchyard, Dumfries, 133. St Mungo (see also Kentigern) : Church, 543. Parish of, 543-547. St Ninian’s Well, 85. St Patrick’s Chapel (site), 87. St Quintin’s Church (site), 352. St Thomas’ Chapel (site), 457. Sanquhar: Cairn, The Manse (site), 569. Castle, Ixi, 551. Church, 548. Crannog in Black Loch (site), 568. Cross, 563. Cross-Slab, 550. Efhgy, 549- Old House, 563. Parish of, 548-576. “ Sean Caer,” (site), 567. Senchar Hospital (site), Ixvii, 572. Sanquhar Moor, Cairn (site), 573. Saugh Hill Plantation, " Stone Ring,” 188. Saughtrees, 618. “ Scots’ Dike,” xix, 48. Sculptured Stones, 43, 514, 628. ” Sean Caer ” (site), 567. " Seancastle Doon,” 240. Selcoth, Enclosure, 489. Senchar Hospital (site), 572. Sepulchral Monuments (see Tombstones and other Memorial Slabs). " Serjeant Know,” 308. ” Seven Brethren,” Whiteholm Rig, Stone Circle, xlviii, 603. " Shancastle Doon ” (see Seancastle Doon). Shankend, Enclosure, 306. Shaw, Enclosure, 305. Shaw Hill, 647. Shaw Rig, Fort, 633. Shaws Moor, Small Cairns and Hut-Circles, xlix, 345. Shiel Burn, Forts, 635, 636. Sibbaldbie, Church (site), 29. Sites, 6, 7, 8, 29-32, 38-41, 49-57, 82-88, 94, 95, 101-105, no, 119-126, 135, 150, 151, 167-170, 204-208, 228, 255-260, 267-270, 276-279, 286, 311-314, 326, 331, 334-336. 352-354. 364-366, 379-382, 427, 428, 440- 444 . 454 - 459 . 473 . 474 . 498-509. 520-524, 532-536, 541. 542, 546, 547. 567-576. 588, 589, 593, 594, 607, 608, 616, 632, 649. Skipknowe, Standing-Stone, 627. Slacks, Fort (site) near, 589. Smith, William, Martyr’s Grave, 614. Snade: Castle (site), 260. " Mote,” the Orchard, liii, 237. Solway Moss, xxxvii. Sorbie Bridge, Cairn (remains), xlvii, 222. South Kilblain, 41. South Mains, Elliock, Fort, 555. Southburn Fort (site), 607. Spedlin’s Tower, Ixi, Ixv, 446. Heraldic, etc.. Stones at Luce, 446. Spiral-marked Slab, 1 , 43. Springfield Hill, Fort, 141. Springkell : Cross, 374. Fort, 369. Gravestones, 373. Kirkconnel Church, 367. Kirkconnel (site), 380. Stakeheugh, xxxiv, 44. Standburn Cottages, Stone Circle near, xlviii, 603. Standing-Stones, 47, 81, 100, 223, 263, 349, 350, 426, 471, 627, 629, 630. Stanshiel Rig ; Fort, 403. Small Cairns, 420. Staplegordon : Church (site), 442. Enclosure, Henwell, 433. Stapleton Tower, Ixi, Ixii, 106. Stewart of Dalwinston, Arms of, 33 (2). Stiddrig : Cairn (long), xlvii, 415. Fort, 405. Hut-Circle, 416. Stockbridgehill, Fort, 463. Stone Basins, 132, 332, 470; Stone Circle and St Marjory’s Cross, remains of, Woodhead, 5 - Stone Circles, .xlviii, 5, 81 (supposed), 198,199, 263 (supposed remains), 284, 307, 603, 625. Stone Rings, 185, 188, 189, Stonehouse Tower (site), 269. Strong Places on the West March in the i6th century, Ixiv, Sundaywell ; Fort, 139. Small Cairns on Moor, 146. Tower, 137. Sundials, 156, 495. Tail Burn, Fort (remains), 481. Tanlawhill, Enclosure, 191. Tanner’s Linn, Fort, 320. Tarnis Burn, 414. Tarrona, Fort, 209. Lassie’s Height, Fort, lii, 411. Templand, 449, 451. Temple Wood, Fort, 142. Templehill, Mound, 472. Thornhill : Cross at Nith Bridge, Ixvii, 531. Cross-shaft from Closeburn, Ixviii, 514. ” Three Brethren,” 272. Threip Moor, Cairn, 72. Libbers Castle, lix, lx, 157. Tinwald, xxiii. Gravestone in Churchyard, 586. Heraldic Slab, Lodge of Tinwald House, 584 Mote, XXV, 582. Parish of, 577-589. 301 HISTORICAL MONUMENTS (SCOTLAND) COMMISSION. (The Roman numerals refer to the pages of the Introduction, the Arabic numerals to the Articles of the Inventory.) Tombstones and other Memorial Slabs, 42, 58, 77, 78, 92, 133. 149. 15-. 153. 154. 155. 203, 225, 232, 251, 264, 274. 309, 310. 356, 373 . 439 , 469, 470. 518. 540, 586, 587, 606. 614. Todshawhill, 205. Torthorwald : Castle, Ixi, 590. Church (site), 594. Mural Slabs, 592. Parish of, 590-594. Torthorwald, William, Lord of, 134. Tower of Sark (site), xix, 53. Townfoot, Fort, 61. Townhead, Hut-Circles, 66, 67. Tracks, Ancient, 19, 163, 411. Trailflat : Church, 577. Gravestones in Churchyard, 587. Trail trow: Chapel (site), xxi, 95. Tower, xli (see also Repentance Tower). Trigony Wood, Fort, 62. Trohoughton, Fort, 130. “ Trj'al Cairn,” near Byeloch (site), 524. Tumuli, 139, 202, 417, 519, 530 (see also Cairns). Sites, 84, 126, 208, 522, 523, 541. Tundergarth : Castle (site), 608. Gravestones, 606. Parish of, 595-608. Tweeds Cross, Flecket Hill (site), xix, 502. “ Twelve Apostles ” (Stone Circle), Holywood, xlviii, 284. “ Twin Rings,” Eskdalemuir, 187. Tynron : Churchyard, Martyr’s Grave, 614. Killywarren, 615. Parish of, 609-616. Tynron Boon, lii, 609. Unthank : Church (foundations), 224. Enclosure, 214. Gravestones, 225. Upper Glendivan, Enclosures, 221. Upper Hill, Unthank, Enclosure, 214. Village Site, 473. Virginhall Plantation, Port (site), 533. Vitrified Fort, liii, 339. Wait Hill, Fort (site), 509. " Wallace’s House,” Fort, 358. Wamphray : Gravestone in Churchyard, 628. Parish of, 617-632. Sculptured Stone in Church, 628. Wamphray Place : Mote (site), 617. Sundial from Wamphray Place at Ericstane, 495. Wanlock Water, Entrenchment or Mote, 553. Wardens of the Marches, xxxviii et sea. Wardlaw Hill : Earthwork (site), 38. Fort, XX, 35. Watcarrick, Enclosure, 186. Watchman Hill, Small Cairns and Hut-Circles, 346. Water beck, 472. Waterhead, Enclosure, 302. Wauchope Castle (site), 441. Wauchope Church (remains), 438. Wauchope Churchyard, Cross-slab (Gravestones), 439. Well Burn, Moffat, Forts (sites), 499. Well Path, li, 163. Wells, 6, 40, 85, 122, 616. Welsh, William, Martyr’s Tomb in St Michael’s Church¬ yard, Dumfries, 133. Westerhall, Enclosures, 642, 643. Westerkirk, Parish of, 633-649. Westhills : Roman Altar, 266. Tower (site), 268. Westhills Moss, Fort, 261. Westside, Enclosures and Hut-Circles, 192-196. Westside, Heraldic stone at Berryscaur from, 308. Whitcastles : Port, 298. Stone Circle, xlviii, 307. “ White Birren,” Lyneholm Hill, Fort, 639. ” White Cairn,” Fleuchlarg, xlvii, 249. " White Cairn,” Honeyhole, 529. White Hill, Ericstane, Fort, 487. White Hill, Hightown Hill, Fort, 580. Whitecastle Knowe, Fort, 9. Whiteholm Rig, Stone Circle, xlviii, 603. W^hitestanes Moor, Small Cairns and Hut-Circles, xhx, 344. Whitstone Hill, Fort, 598. Wilson (and Young), Arms of, 540. Windy Edge, Long Cairns, Small Cairns and Standing- Stones, and Site of Cairn near, 47, 49. Winterhopehead, Standing-Stone, 471. “ Witches Wa’s,” Cairn (remains), 74. Woodhead, 5. Woodhouse : Cross, 378. Tower, Ixiv, 368. Woodlea, Cross-Slab, 253. Woodwork, 578. Woody Castle, Fort, liv, 450. Wreath’s Tower (site), 364. Yetts, Ixiii, 59, 156, 327, 337, 537. Young family. Tombstones of, 540. Young’s Burn, Hut-Circle and Small Cairns, 361. Printed under the authority of His Majesty’s Stationery Office By Neill & Co., Ltd., 212 Causeway side, Edinburgh. SHEET 31 (PARTS OF 32 34 ). ICitlKOONNEL kfffrPA TR/QJ nii-phf fi««r » 0 >.<.nb- WeSTIRK/RK ,K I ff:K M ! L i ; tOnbckarth ••S'-IN'e""" »oL'r\if»-oo i^- /, ' fUMiite TTi CUMMCRTliefl . . > IIC)V.4I. COMMlhSION ON .ANCIKNT A IIISTDUK'AI. MOM MENTS CARRUveHOCR^ Date Due ' RfP 1 • W- r [ Form 335. 45M 8-37. D00384384U qS424R 7th 914.1 355119