~B7wse\ 13/7 I OJnrneU Intoersttg Kthrari} Sttjata, JJetn ^ork BOUGHT WITH THE INCOME OF THE SAGE ENDOWMENT FUND THE GIFT OF HENRY W. SAGE 1891 ^r- «— ~£?I ne " University Library QE 262.B7W58 1917 Th< ; geology of the country around Bourne 3 1924 004 549 915 Cornell University Library The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924004549915 MEMOIRS OF THE GEOLOGICAL SURVEY, ENGLAND AND WALES. EXPLANATION OF SHEET 329. THE GEOLOGY OF THE COUNTRY AROUND BOURNEMOUTH SECOND EDITION. BY H. J. OSBORNE WHITE, F.O.S. [First Edition, in 1898, by CLEMENT EE1D (F.R.S.), F.L.S., F.G.S.] PUBLISHED BY ORDER OF THE LORDS COMMISSIONERS OP HIS MAJESTY'S TREASURY. LONDON: PRINTED UNDER THE AUTHORITY OF HIS MAJESTY'S , STATIONERY OFFICE ~^y JAS. TRUSCOTT and SON, Ltd., Cannon Street, E.C. 4. And to be purchased from ~ E. STANFOKD, Ltd., 12, 13 and 14, Long Acee, London, W.C. 2 ; W. and A. K. JOHNSTON, Ltd., 2, St. Andrew Square, Edinburgh ; HODGES, FIGGIS and Co., Ltd.. Grafton Street, Dublin. 'rom any Agent for the sale of Ordnance Survey Maps, or through any Bookseller, or from the Dire ctor-General, Ordnan ce Survey Office, Southampton. 1917. Pries 2s. Od. Net. Missing Page MEMOIRS OF THE GEOLOGICAL SURVEY. ENGLAND AND WALES. EXPLANATION OP SHEET 329. THE GEOLOGY OF THE COUNTRY ABOUND BOURNEMOUTH SECOND EDITION. BY H. J. OSBORNE WHITE, E.O.S. [First Edition, in 1898, by CLEMENT RE1D (F.R.S.), F.L.S., F.G.S.] PUBLISHED BY ORDER OF TBE LOBDS COMMISSIONERS OF HIS MAJESTY'S TREASURY. LONDON: PRINTED UNDER THE AUTHORITY OF HIS MAJESTY'S STATIONERY OFFICE By JAS. TRUSCOTT and SON, Ltd., Cannon Street, E.C. 4. And to be purchased from E. STANFORD, Ltd., 12, 13 and 14, Long acre, London, W.C. 2 ; W. and A. K. JOHNSTON, Ltd., 9, St. Andrew Square, Edinburgh ; HODGES, FIGGIS and Co.. Ltd.. Grafton Street. Dublin. From any Agent for the sale of Ordnance Survey Maps, or through any Bookseller or from the Director-General, Ordnance Survey Office, Southampton." . ' 1917. Price 2s. Od. Net. Ill Preface to the first edition. The Hampshire cliffs between Christchurch and Lyniington long ago attracted attention for the number and beauty. of their fossil shells. They were described by Gustavus Brander, in his " Fossilia Hantoniensia," published in 1T66, with descriptions of •the fossil shells by Dr. Solander, and beautiful plates of the species of which no less than ten are reproduced in the following explanation. In later years original researches among the same strata were carried on by Lyell, Prestwich, the Rev. Osmond Fisher and others, of whose labours further account will be given in a forth- coming general memoir on the Hampshire Basin. Especial mention, however, should be made here of the detailed observa- tions of Mr. J. Starkie Gardner, which are further referred to in the following pages. The area comprised within the map (Sheet 329, new series), of which the present pamphlet is explanatory, was originally surveyed by the late H. iW. Bristow and J. Trimmer, and was included in Sheets 15 and 16 (original series) of the Geological Survey Map of England, published in the years 1855-56. In this map the superficial deposits were not represented. During the revision of the Geological Survey of the South of England, the district here described has been re-examined by Mr. Clement Reid, who in 1892-4 mapped it on the scale of six inches to one mile, and traced the distribution of all its superficial deposits. T\^o editions of the map here described ,are issued, one showing the areas occupied by the Cretaceous, Eocene, and Oligocene strata ("Solid Geology" edition), the other displaying the distri- bution of the various surface deposits (" Drift " edition). The present explanation has been prepared by Mr. Reid- It is intended as a general guide to the use of the map until a more detailed account of the whole surrounding region can be issued. Thanks are due to Mr. Gardner, and to the Council of the Geological Society, for permission to use two illustrations from his valuable papers. The map embraces a considerable part of the Tertiary Basin of Hampshire. Till it came into vogue as a residential district, much of this part of the country was almos uninhabited; and the towns, though ancient, were very small. Bournemouth has entirely sprung into existence within recent years, the old Ordnance Map of 1811 showing on the site merely a single house and decoy ponds, in the midst of open moors. ARCH. GEIKIE, Geological Survey Office, Director-General. 28, Jermyn-street, London, SW., 23rd May, 1898. (4705.) Wt. 59180—45. 500. 7/17. .T. T. & S., Ltd. G. 14. A 2 IV PREFACE to the Second Edition. The first edition of this Memoir, which, was written by the late Mr. Clement Reid, was exceptionally brief, a general memoir descriptive of the Hampshire Basin as a whole having been at that time in contemplation. Circumstances have pre- vented the preparation of the larger work, and opportunity has now been taken of the exhaustion of the stock of the original pamphlet to produce a memoir on the lines of other New Series Sheet Explanations. For the writing of this volume we hare again to acknowledge our obligation to Mr. H. J. Osborne White, to whom we are already indebted for several memoirs dealing with adjacent districts. The history of the earlier geological investigations of the part of the Hampshire Basin illustrated by Sheet 329 has been traced in the preface to the first edition, and it will suffice now to add that since 1898 researches on the gravels have been carried on by Dr. W. Theophilus Ord and Mr. Reginald A. Smith (to both of whom, and Mr. A. S. Kennard, the author of the present memoir desires to acknowledge his indebtedness), and on the Tertiary strata by Dr. F. R. Cowper Reed, Dr. Ord, and others. The figures of fossils referred to in the preface to the first edition are not reproduced in the second edition. Much of the ground has been re-examined by Mr. White, in order to bring the memoir up to date, but the map remains unaltered as the edition published in 1895 and colour-printed (Drift) in 1904. A. STRAHAN, Director. Geological Survey Office, 28, Jermyn Street, London, S.W. 1. 15th March, 1917. LIST OF SIX-INCH MAPS. The following is a list of the Quarter-sheets of the six-inch Geological Map of the area, of which MS. coloured copies, prepared by the surveying officer, Clement Reid, are deposited for public reference in the Library of the Geological Survey and Museum of Practical Geology. Dorset. 25 S.W., S.B. 26 S.W., S.E. 34 N.W., KB., S.W., S.B. 35 N.W., KE., S.W., S.B. 43 N.W., KE., S.W., S.E. 44 N.W., N.E., S.W., S.E. 50 N.W., KE., S.W., S.E. Hampshire. 70 S.W., S.E. 71 S.W., S.E. 78 N.W., KE., S.W., S.E. 79 N.W., KB., S.W., S.E. 85 N.E., S.E. 86 N.W., KE., S.W., S.E. 87 KW.,KE., S.W., S.E. CONTENTS. PAGE Preface to the first edition by the Director-General m Preface to the second edition by the Director iv List or six-inch Maps Chapter I.— Introduction, Physical Features, Formations, Tectonic Structures, &c. „ II. — Upper Chalk " „ III. — Reading Beds 9 IV. — London Clay 15 „ V. — Bagshot Beds ■■■ 19 ., VI. — Bracklesham Beds 28 „ VII.— Barton Beds 41 „ VIIL— Headon Beds 45 „ IX. — Plateau Gravel 47 „ X. — Valley Ctravel 57 „ XI. — Recent Deposits 63 „ XII. — Economic Geology 68 ILLUSTRATIONS. Fig. 1 . — Sketch-map of the District 3 „ 2. — Section in the Reading Beds, near Combe Aimer 11 „ 3. — Apparent Relation of the Reading Beds to the Chalk west of Kingston Lacy Hall .. . ... ... ... ... . 13 „ 4. — Sections in the London Clay and Reading Beds near Lytchett Matravers ... ... ... ... ... ... 17 „ 5. — View of the Cliffs between Poole Harbour and Boscombe... 23 „ 6. — Western Termination of the Bournemouth Marine Beds ... 26 „ 7. — Honeycomb Chine, west side ... ... ... ... ... 32 ,, 8. — Profile-section of the Barton Beds, Barton Cliff ... ... 42 „ 9. — Terraces of the Plateau Gravel, east of the Avon ... ... 50 „ 10. — Silhouette of Hengistbury Head, from the north ... ... 52 „ 11. — Palaeolithic Implements from the Bournemouth Gravels ... 55 „ 12. — Section through Gravel Terraces south of Wimborne 56 „ 13.— Section at Cliff End, near Highcliff 58 THE GEOLOGY OF THE COUNTRY AROUND BOURNEMOUTH. CHAPTER I. INTRODUCTION. Location and Area. — Sheet 329 of the Geological Survey Map on the one-inch scale represents an area of 216 square miles, of which about 43 are occupied by the sea. The land area comprises, approximately, 71 square miles of Hampshire, including Bourne- mouth and Christchurch, with the villages of Burley and Brans- gore on the western confines of the New Forest, and about 102 square miles of eastern Dorsetshire, including Poole, Wimborne Minster, and Sturminster Marshall. The extent of the map and the geological structure are outlined in the sketch-map, Fig. 1 (p. 3). Physical Features. — (1) Land Area. — The highest point is &i Badbury Rings, 327 feet O.D., in the north-western angle of 'the district, but ground within 20 feet of this level occurs at Lytchett Matravers on the west, as well as at Castle Hill, near Burley, on the north-east, so that the prevailing seaward slopes have a conver- gent tendency, which is clearly indicated in the courses of the principal rivers, the Frome, the Stour, and the Avon. Although the local topography is too varied to be adequately summarised in a phrase, the general aspect of the country suggests some such systematic description as ' a coastal plateau of low relief, in mature to senile stages of dissection.' North of the Stour, near Wimborne, the district takes in a typical Chalk-down, with radiating combes, groups of tumuli, and a crowning tree-clump encircled by the prehistoric earthworks of Badbury Rings. Eastward the down declines to the valley of the Allen, beyond which come the clay-lands around God's Blessing Green, the gravelly ridges of Cole and White Sheet Hills, and then a six-mile stretch of rolling heath, with fir plantations, in the basin of Moors River. Near here the lower courses of the Stour and Avon are divided by the narrow ridge of Sopley Common and Ramsdown, recognisable from afar by its undulate sky-line, and terminating southward in the little plateau of St. Catherine's Hill, above Christchurch. 2 GEOLOGY OF BOURNEMOUTH. South of Wimborne, and beyond the valley of the Stour, the ground rises for about two miles to the crest of the Corf e Hills and Dunyeat's ridge, whence there is a wide outlook over the theatre of Canford Heath, to the lower broken country of flats and rounded knolls about Poole Harbour. Canford Heath is bordered on the east by the scarp of a low, gravelly upland, which, beginning as a group of flat-topped ridges between and around the branches of the Bourne, develops south and eastward into the gently inclined plateau supporting Branksome Park, Bournemouth, and South- bourne- This plateau is cut off at the coast in a range of cliffs, from 100 to 140 feet high, and breached at irregular intervals by the steep-sided chines that form so attractive a feature of the Bournemouth seaboard. Beyond Southbourne the cliffs are con- tinued in the detached ridge of Hengistbury Head, and, after an interval of two miles, occupied by low ground near Christchurch Harbour, they rise again at Highcliff, where they form the southern limit of an inclined gravel-plain that leads up to the plateaux overlooking the vale of Cranes Moor, near Burley. The chief rivers of the district have already been mentioned. Of these, the Frome is represented by its irregular estuary, Poole Harbour, in which it is joined by the Puddle or Trent and the Shefford River, from the west; by the Corfe River from the south, and by a number of minor brooks from other quarters. The Stour and Avon, confluent in Christchurch Harbour, flow briskly in shallow, meandering channels on the flat floors of valleys which possess a breadth and curvature in keeping with rivers of far greater volume than these attain under existing conditions. Within the limits defined by the Bournemouth map the Stour is joined by three tributaries — the Allen and Moors Rivers from the north, and the Winterbourne from'the west, the last running in just above Sturminster Marshall. The Avon receives locally no affluents on its right or western bank, as the crest of the bluff on that side of its valley coincides with the water- parting between it and the Stour. The Mude, the Ripley Brook, and two other rivulets which join in from the east, all run south- ward for two or three miles along the floor of the main valley before they effect their several junctions with the Avon. There remain a few streams with independent outlets on the coast, such as the Bourne at Bournemouth, the Walkford Brook which runs out at Chewton Bunny, the Avon Water rising on Cranes Moor, and the Ober Water, which passes north of Burley. The latter two are the only local brooks to reach the sea outside the district. (2) Sea Area. — From the mature shore-line, defined by the graceful, paraboloid curves of Bournemouth (or Poole) and Christ- church Bays, the bed of the sea falls away south and south-east- ward at a moderate inclination, the greater part of it, within the limits of Sheet 329, lying between the 5 and 10-fathom lines. Its gently undulate contours are diversified by low residual ledges of Tertiary rocks, and by banks of Recent shingle and sand, which have their chief development near the outlet of Poole Harbour. The topography of the sea-floor bears but little resemblance to that of the neighbouring land, save in the eastern part of the area, INTRODUCTION. 6 where the salient in the coast at Hengistbury Head is prolonged down to the 10-f athom line in Christchurch Ledge ; and the Stour Valley has its sequel in a broad, irregular channel in the bottom of Christchurch Bay. Hengistbury Head and Christchurch Ledge, it may. be remarked, have a common origin, in so far as both have been determined by the outcrop of a group of resistant Fig. 1.— Sketch Map of the District. V . 0" • Ba db u ny. /?/ n^rs^y^, W1MBORNE V~/\ MINgTEFt • 7-'-- "■ J • O "VLytcheft &Matr avers A* i Bur ley strata, but the genetic relation of the Stour Valley and its con- tinuing channel is doubtful, since there is nothing to show whether the latter feature be a submerged and much-modified river-valley, or merely one of those apparently fortuitous, sea- worn hollows which abound in the sub-littoral zone. The Geological Formations represented on Sheet 329 are the following : — Eecent ... Pleistocene Oligocene Eocene ... 'Cretaceous ('Blown Sand. ... •< Alluvium. (.Peat. ( Brickearth. .. < Valley Gravel. (.Plateau Gravel. .. Headon Beds. 'Barton Sands. Barton Clay. Bracklesbam Beds. Bagshot Beds. London Clay. .Reading Beds. Upper Chalk. 4 GEOLOGY Or BOURNEMOUTH. As no local boring has been carried below the Upper Chalk, the character and age of the underlying strata can be only inferred from observations made in other parts of the country. Under Bournemouth the descending sequence for the first 3,000 feet beneath the Chalk is doubtless similar to that revealed in the cliffs near Swanage, except that the Selbornian' Series (Upper Gre.ensand and Gault) is likely to be rather thicker, and the Lower Greensand, Wealden, Purbeck, Portland, and Kimeridge Clay, are all probably thinner, than on the Purbeck coast ; the falling- off being most marked in the Lower Cretaceous divisions. With respect to older Mesozoic formations, inference degenerates so quickly into conjecture that it would appear unprofitable to follow the succession to lower stages. The nearest known outcrop of Palaeozoic rocks occurs near Shepton Mallet, and in the interval of 40 miles between that place and Bournemouth the Palaeozoic platform may have fallen to a depth of 8,000 or 9,000 feet below sea-level. Tectonic Structure — In this connection there is not much to add to the brief summary given by Mr. Clement Reid in the first edition of the Bournemouth Memoir. 1 "The whole area," he writes, "is occupied by the southward slope of the main Hamp- shire Basin syncline, the rise towards Studland and the Isle of Purbeck commencing just outside the district. The inclination of the strata is never great, only occasionally approaching 2°, and usually about 1°. The synclinal trough has an almost flat bottom, with but minor undulations. This flattening is due to the arrangement of the subordinate folds en echelon, the west- ward disappearance of one syncline near Wimborne coinciding with the increase of another fold which passes under Poole Harbour." The flexing is so gentle, and the solid rocks are so much obscured by superficial deposits, that it is seldom, possible to determine the position of the tectonic axes. The downfold which fades out west- ward near Wimborne, is traceable eastward, across the valley of the Avon near Bisterne and Whitefield Hill, into the area of the Lymington map-sheet, where it appears to be continuous with the principal syncline of the northern part of the Isle oi Wight. A line drawn from Lytchett Matravers through St. Catherine's Hill and Chewton Common roughly indicates the course of the adjoin- ing anticline, which brings up the Chalk as an inlier near Corfe Mullen. The flattening of the northward dip between Colwell Bay and Calbourne, in the Isle of Wight, is possibly due to the same fold, but more doubt attaches to a suggested correlation of the Poole syncline with the flexure at Headon Hill. As will be seen later on, there are local disturbances which give rise to higher dips than those mentioned by Mr. Reid in the passage quoted above. The dominant or regional dip is south-eastward, and it is this that has determined the general direction of the drainage. 2 The ' 1898. pp. 1,2. 3 See A. Strahan. ' On the Origin of the River Systems of South Wales &c ' Quart. Journ. Geol. Soc, vol. lvii, 1902, p. 221, INTKODUCTION. 5 Avon, it is true, runs southward for much of its course, and so do the Moors and Allen Rivers, but, as these streams tend to follow the strike of the stratification, it is probable that they are in part of 'subsequent' origin. 1 The Moors River, especially, shows adjustment to the outcrop of weak Eocene strata, although its higher branches in the Chalk country, like those of the Avon and the Allen, follow the direction of the dip. 1 Mr. C. Reid has adduced some evidence of the Avon's development at the expense of the earlier, south-eastward drainage-system (Ringwood Memoir, 1902, pp. 30-32). GE0L0G1' OF BOURNEMOUTH. CHAPTER II. UPPER CHALK. The Chalk appearing- at the surface in the north-western angle of the district belongs to the Zone of Belemnitella mucronata, a sub-stage which is known to be over 250 feet thick near Studland, about 11 miles southward of Wimborne, and is estimated to in- clude approximately 200 feet of beds next below the Eocene in the vicinity of Cranborne, some 8 miles to the north. Unfortunately, the Mucronata Zone _ seems not to admit of more than local sub- division, so that no satisfactory correlation with distant sections can be established. This is the more to be regretted as the zone is the highest member of the Cretaceous system so far recognised in the South of England. It would be interesting to know in what measure, if any, an original deficiency of sediment contributed to the northward and eastward thinning and disappearance of the zone, beneath the Eocene Beds, in Hampshire. It is presumable that all the zones of the Chalk below the out- cropping Mucronata Zone are also present underground in the area, but there is no direct evidence as to the total thickness of the formation. Judging from its known thickness in the Isle of Wight and in other parts of the South of England, the full thick- ness of the Chalk in the district is likely to be about 1 ,000 feet. Near Wimborne the chalk is soft, of fine texture, and pure white, save for occasional greyish seams of a slightly marly character. Well-preserved tests of foraminifera are plentiful, but the chalk on the whole is poor in fossils that can be readily seen with the naked eye. The flint-nodules, which are usually of small Size, occur but sparingly ; in some exposures there are hardly any to be seen in a depth of 10 to 15 feet; in others they are widely spaced in single or multiple layers, which seldom persist, as conspicuous features, for many yards at a stretch. Bands of almost flintless chalk occur at different horizons, and the highest parts of the formation to be seen near Wimborne are mostly of this type. Notes of Exposures- There are several sections, but as they show few features of interest it is unnecessary to describe all of them in detail. From a consideration of the form of the ground in relation to the dip, it is inferred that the oldest beds open to inspection crop out near the left bank of the River Stour above Sturminster Marshall. In that part of the district a road-side quarry half a mile south of Crab Farm shows 15 feet of chalk with scattered ' rust-balls ' (oxidised pyrites), three indistinct courses of rather UPPER CHALK. 7 small thick-rinded flints, and a few thin, oblique veins of tabular flint. Guards of Belemnitella rnucronata v. Schloth. are readily found in the upper part of the section. Other fossils noted include Axogaster cretacea Lonsd., Echinocorys scutatus Leske (a small variety, and a large form with an acuminate apex; both charac- teristic of the zone), Salenia granulosa Forbes, 1 some bryozoa of wide range {e.g., Stomatopora granulata (M. Edw.), Onychocella lamarcki (v. Hag.) ), Dimyodon nilssoni (v. Hag.), Ostreae. A larger quarry by White Mill Bridge shows a weather-stained, crumbling face of about 16 feet. Flints are rare save for an impersistent layer at the top. Mr. Jukes-Browne, who probably found the section in better condition, states that Belemnitella rnucronata is not uncommon here, but the only other fossil he mentions is Kingena lima (Defr.). 2 In open grass-land rather more than half a mile east of Badbury Rings there is a quarry which has been worked in two tiers. The lower tier gives a section about 15 feet deep in soft chalk containing a few, small scattered flints, and one well-marked undulate band of rather large flints, with thick grey rinds, near the top. Rusty pyrite-nodules of irregular shapes are plentiful. The most common fossils are the small calci-sponges Porosphaera globularis (Phill.) and P. nuciformis v. Hag., and ossicles of Asteroids referable to Metopaster parkinsoni (Forbes), Mitr aster rugatus (Forbes), and P entagonaster quinque-lobus (Goldf.). Other forms observed were Cidaris hirudo Sorig., C. subvesi- culosa ? d'Orb., Helicodiadema fragile (Wilts.), Echinocorys scutatus Leske (large, tall), Serpula plexus J . de C. Sow., Rhyn- chonella plicatilis (J. Sow.), Inoceramus inconstans ? Woods. The upper tier shows 10 feet of soft, apparently flintless chalk immediately overlying the beds seen in the lower section. The only fossil observed by the writer was Belemnitella rnucronata. Chalk like that in the lower part of the quarry just described, and probably at about the same horizon, is exposed in a pit 3 fur- longs north-north-west of the tumuli on King Down, and some almost flintless beds, yielding Belemnitella rnucronata, are dug on the south-eastern side of King Down Farm, 3 at a level about 30 feet lower than the adjacent outlier of Eocene Beds. Barren chalk with few flints was exposed in temporary trenches on the 200 feet contour north-west of Badbury Rings- The chalk exposed at intervals along the boundary of the Eocene Beds presents few notetvorthy features. The characteristic belemnoid and Echinocorys are usually forthcoming, but other remains are rare and difficult to detect in the stained and closely- jointed chalk in this position. North of the Stour, chalk with few flints has been worked in a small pit west of the high-road at Hinton Parva ; in another, much overgrown, just on the boundary of the Reading Beds a quarter of a mile north-west of High Hall, ' Recorded by A. J. Jukes- Browne, 'Cretaceous Rocks of Britain,' Mem. Geol- Swv., vol. iii, 1904, p. 122. 2 Loc. cit. 3 Probably the "High Wood" section mentioned by Prof. C. Barrois in his '' Reclierehes sur le terrain cretact super ievr,^"C," 1876, p. 65. 8 GEOLOGY OF BOURNEMOUTH. and in a third and larger excavation at the same distance south- east of Chilbridge Farm. At the Water Works near the bridge at the northern end of Wimborne a well, completed in 1899, was sunk to a depth of 160 feet beneath the Eocene cover. In the records 1 flints are mentioned only as occurring in a thin band of hard chalk about 85 feet below the top of the formation. The few fossils met with included Belemnitella mucronata. South of the Stour, beds with few flints can be seen in several places at or close to the boundary of the Reading Series near Combe Aimer and Henbury Barrow; also on the eastern side of the Chalk inlier south of Corfe Mullen. At the last-mentioned spot, where the junction with the Reading Beds is just discernible in an overgrown dell, the chalk at the contact is converted into firm limestone ; iron-stained, spotted with dendrites of manganese oxide, and traversed by tubular perforations from 0"25 to 2 mm. in diameter, the larger of which are filled with rusty sand. 1 ' Water Supply of Hampshire,' Mem. Geol. Surv., 1910, p. 145, and ' Country around Ringwood,' Mem. Geol. Surv., 1902, p. 7. GEOLOGY OF BOURNEMOUTH. CHAPTER III. READING BEDS. The abrupt change, from pure, deep-water limestone to the muddy sand of a shallow sea, which is encountered on passing from the Cretaceous to the Eocene, marks a break in the stratigraphical succession and a lapse of time, during which the Chalk was raised by earth-movements to within reach of the agents of erosion, and subjected to a planation that involved the destruction of a variable thickness of its higher beds. To what extent sub-aerial detrition was concerned in the removal of the missing strata, it is impos- sible to say. Where newly stripped of its Eocene cover the top of the Chalk resembles a wave-worn foreshore, and, as the lowest Eocene sediments are marine, one must infer that the planing- down process, in its final stages, at least, was performed by the sea. In the country dealt with in this memoir the Reading Series varies in thickness, apparently from about 70 to 1-00 feet. The marine Bottom-bed comprises a few feet of roughly laminated, speckled, loamy sand, having a greenish tint, due to the presence of glauconite, where unweathered, and containing small black or bluish-grey pebbles or pieces of flint which are mostly of sub- angular form. Larger flint-nodules with a dull green or black glaze, and unrolled though commonly having a gnarled appear- ance, which seems to be due to dissolution of the original white cortex, occur in an impersistent layer at or just above the junc- tion with the Chalk. The seams of oyster-shells often present at this horizon are but feebly developed at the outcrop near Wim- borne. The rest of the formation — probably of freshwater origin — is made up of current-bedded sands, loams, fine silty clays, and stiff red and purple mottled clays, arranged in overlapping lens-shaped masses, which may vary in thickness from a few inches up to 30 or 40 feet, and from a few yards to a mile or more in diameter. About Portsmouth and in the Isle of Wight the clays make up the bulk of the formation, and to judge from a well-boring at Christ- church, they still predominate in the eastern part of the Bourne- mouth district; but as the Reading Beds range westward they become decidedly arenaceous, and in the neighbourhood of Wimborne approximately half of their outcrop is occupied by sands. These latter are mostly coarse and ferruginous, and, besides pieces of clay torn by currents from slightly older parts of the series, they include seams of quartz-grit mixed with small chips and flakes of flint. Farther west, in the Dorchester dis- trict, the sands contain, in places, much subangular flint and 10 GEOLOGY OF BOURNEMOUTH. chert gravel of fluvial aspect, indicating the proximity of the western shore of the wide and shallow body of water in which tL'. Reading Beds were laid down. Notes of Exposures. Taking the main outcrop first, and beginning on the north, one notes a small exposure of the greenish Bottom-bed in the banks of I Jane 350 yards north-west of High Hall, and som« remnants d± the same bed, with .green flints and bits of oyster- shell, farther out on the Chalk north-east of Lower Barnsley Far*- Around and to the east of High Hall the Reading Beds are netu all mottled clay, but at the top there is some coarse sand, wh.Msu. appears to thicken northward, and has been worked to a depth^of about 18 feet in a pit 3 furlongs east of Stanbridge, just within the boundary of the London Clay, according to the map. Between Higher Honeybrook Farm and Wimborne there are surface indi- cations of coloured clay near the upper limit of the series, the lower beds being hidden by the alluvium of the River Allen. A well, made in 1896, at the Water Works near Walford Bridge, at the northern end of Wimborne, gave what appears to have been a complete section of the formation. The descending succession noted was as follows 1 : — [Valley Gravel] [ ? London Clay] [Reading Beds, 84 feet]' [Chalk, 159i feet.] Ballast ... Sand-ballast ... 'Coloured clay... Yellow elay ... Blue clay Grey sand Blue clay Sandy blue clay Grey sand Mints Soft chalk i Hard chalk with flints (Chalk Feet. H H 8 •2 " 4J> -■'•> 4 10 40 0} 84* 0| 74 256 In an adjacent trial-bore, the site of which is marked on the map, the Chalk was reached at a depth of 101 feet, or 4^ feet farther from the surface than in the well. Red mottled clay, just below the junction with the London Clay, is exposed for about 20 feet in the lower workings at the brick- yard on the right of the Allen, and a similar deposit, in the same stratigraphieal position, occurs in the higher ground about Hill- butts and Pamphill, between the Allen and the Stour. Sandy beds with clay-seams, probably representing the 40 feet of grey sand " met with in the Water Works well, form much of the middle and lower parts of the Reading Series hereabouts : they have been dug on the northern slope of Hound Hill, where the road-banks afford small exposures. About Abbott Street there are signs of a mottled clay close above the Bottom-bed, which can be seen in the banks of the lane north of Cowgrove. 1 Taken from the ' Water Supply of Hampshire,' Mem. Genl. Surv. 1910 p 146 with some alteration in the arrangement and interpretation of the record. ' ' HEADING BEDS. 1L In the Stour Valley, the Reading Beds are hidden by gravel <*',*nd alluvium for a space of about three-fourths of a mile. Where (hey reappear about Corfe Mullen, they consist almost wholly of red mottled and brown sandy clays. The highest beds, composed of these materials, were formerly well displayed in the railway cutting east of the village, and clays lower in the series appear in the lane-banks and brick-yard near Hill Farm, to the south. „The brick-yard affords the best existiAg section of .he mottled clays; red, purple, and grey; worked in a face of aoout 25 feet, pfti overlain by a few feet of pale grey silty loam. Dr. W. T. T-'kfd has a bit of resin said to have been found in the clay at the 4 ,- Jtitom of the pit. A small exposure of the Bottom-bed, resting "o-i hardened chalk, can be seen in an overgrown pit about a ^ ..alter of a mile south of the brick-yard. Under Henbury Plantation the ^Reading Beds suffer one of the rapid lateral changes so characteristic of this formation. The clays thin away south-westward into narrow bands and lenticles among a series of sandy beds which occupy nearly all the space between the London Clay and the Chalk. The ascending sequence, made out from a number of exposures and indications in this part of the district, is as follows : — Above the Bottom bed, to be seen in a road-bank at the boundary of the map west of Henbury Barrow, and at the top of a chalk-pit north-west of Higher Combe Farm at Combe Aimer, comes a ferruginous sand, hardened in places into sandstone, which has been excavated for building-stone, by the parish boundary south-east of the same farmstead. The iron-sand appears to pass up into yellow and buff sands, perhaps 40 feet thick, with included masses of clay and loam. A large pit by the side of the Blandford road south- east of Combe Aimer shows 20 feet of these sands, strongly -current-bedded and varying much in texture. Part of a lenticular mass of laminated silty clay (fire-clay), of pale grey tint, appears on the western side of the pit (fig. 2). The original intricacy of Fig. 2. — Section in the Beading Beds, near Combe Aimer. Scale, about 1 inch = 8 feet. a a. Fine sand ; b. Coarse sand, iron-stained at base; o. Silty clay. the bedding in the sands has been complicated at this spot by settlements in later times- The sands are followed by clays, which are not clearly seen, and these in turn by fine white and yellow sands extending up to the base of the London Clay, the junction with which is well shown in pits on either side of the 12 GEOLOGY OF BOURNEMOUTH. Blandford road and of Sandy Lane, near Warmwell Farm. Details of the sections in these places are given in the next chapter (p. 17). Outliers — The three little outliers at and about Badbury Rings differ to some extent in their composition from the mam body of the Reading Beds to the south and east. Judging from sur- face indications, the two patches due south of King Down Farm consist of loamy sand and sandy clay full of flint-pebbles of the well-rounded form more characteristic of the London Clay than of the Reading Beds, in the neighbourhood of Wimborne. Similar pebbles, mingled with flat bits of iron-sandstone, abound in the soil on the surrounding Chalk. It is questionable whether these deposits be true outliers, or merely superficial accumula- tions of material derived from Eocene strata- They are from 50 to, 100 feet lower than the adjacent outlier of Badbury Rings, which forms a knoll on the highest part of the down and pro- bably stands upon a remnant of the Eocene basal platform. The banks of a pond on the western slope of the Badbury outlier show greyish, brown-mottled, sandy clay of a common Reading Beds type, but the top of the outlier, 10 to 15 feet higher, is covered by brown sandy loam with abundant flint-pebbles, like those just noticed. On the whole, the lithological facies of this group of outliers is as suggestive of the Basement-bed of the London Clay as of the Reading Beds, and the writer is of opinion that both are represented. The occurrence of the London Clay here is by no means improbable, in view of tie fact that this member of the Eocene system irregularly oversteps the Reading Beds near Hinton Martell and Cranborne, a few miles distant to the north-east- 1 The remaining outlier, west of Kingston Lacy Hall, also pre- sents problematic features, though in this case the stratigraphical affinities are less in doubt. The constituent beds, while of an uncommon type, resemble those of the Reading Series. The chief difficulty is to account for their anomalous position, to which Mr. Clement Reid calls attention in the first edition of this memoir. The outlier is situated in the side of a shallow combe which runs down from Bradbury Rin^s to the Allen valley near Chilbridge Earm, and at a level about 40 feet below that at which the Reading Beds might be expected to occur, having regard to the dip in adjacent parts of the district. On approach- ing the spot from the south, by the green lane on the western side of the park, one observes that the bare Chalk, which appears underfoot in patches up to the boundary of the outlier, there abruptly gives place to sand; the northward slope of the ground simultaneously becoming more pronounced, and the surface of the lane hollowing into a broad rut with slight ledges that mark the outcrop of the firmer sand-beds- Continuing northward, the slope soon decreases, bits of chalk appear in the soil, and' the sand seems to thin away to a feather-edge at a distance of about 150 yards from the point where it is first encountered. The i See C. Keid, ' Country around Eingwood,' Mem. Geol. Surv., 1892, pp. 13 17 19 HEADING BEDS. 13 accompanying diagram-section (fig. 3) illustrates the apparent relation of the outlier to the Chalk along this line. The sand appears to die out on the east as it does on the north, while on the west boundary is hidden by turf. Fig. 3. — Apparent relation of the Reading Beds to the Chalk, west of Kingston Lacy Hall. N. c. Chalk, r. Reading Beds. In an irregular sand-pit on the eastern side of the lane the following succession can be made out: — Soil ; thin, sandy. Feet. b\ Brown loamy sand, with inclusions of finer greyish loam ... ... ... ... ... ... . ... 2 to 7 S. Roughly-laminated gre3 - -brown sandy clay, thinning out eastward ... ... ... ... .. 0£ 4. Coarse brown loamy sand of flint and quartz, full of rounded and angular bits of brown clay and iron- stone, small angular chips of flint, and a few unworn flints. The materials are loosely cemented in places by iron-oxide,- and the structure is confusedly current-bedded ... ... ... ... ... ... 3 3. Coarse, black-stained, flint and quartz sand seen for 2 'I. Light brown and reddish sand with bands of soft sand- stone, not well shown ... ... ... ... ... 3? 1. Yellow and brown sand, becoming very coarse in the lower part, where it contains many small angular pieces of flint ... ... ... ... ... seen 6 There is no appreciable dip. The succession above described differs in some respects from that noted by Mr. Iteid in 1892, when the lowest bed (1) probably was not exposed. This bed alone was being worked in 1915, in a hole about a yard from the side of the lane. As far as it had then been dug, it gave no clue to its precise position with reference to the Chalk below. Bed (6) is traceable, in degraded workings, to within three yards of the line where the southern boundary of the sands crosses the lane. It is evident from the surface-relations of the sand and the Chalk that their junction at this spot is inclined at a high angle to the horizon, and a small opening made, by Dr. W. T. Ord .and the writer showed the contact to be nearly vertical. The sands here have a selvage of stiff grey loam, containing pellets of chalk and small concretions of ironstone, and exhibiting a brecciate appearance on pared surfaces. The characteristic pebbles, unworn flints, and perforations of the Reading Bottom- bed are wanting. In reviewing the possible explanations of the position of this outlier, it appears safe to reject the hypotheses of tectonic infold- ing and of piping (i.e., settlement into a solution-funnel in the Chalk) : for, apart from the considerations that neither folds of sufficient intensity to meet the case, nor pipes of comparable b 2 14 GEOLOGY. OF BOURNEMOUTH. dimension's, are known to exist in this part of the country^ there is the absence of any appreciable dip or sag in the bedding of the sands, even in the immediate vicinity of the Chalk at their southern limit. 1 The absence of the usual Bottom-bed features furnishes an additional objection to the fold hypothesis. To Mr. Eeid, the curious position of this outlier, "taken with the coarse, almost gravelly nature of the deposit, suggests that the Reading Beds here rest on Chalk unconformably — as do the Reading gravels [in the Dorchester district] a few miles further west." 2 Broadly speaking, the Reading Beds are unconformable to the Chalk throughout their geographical extent, so that the expression ' ' unconformably ' ' in the above passage doubtless means "unevenly," and appearances do give the impression that the Reading sands here abut against a cliff, or steep b an k of a channel, which already existed when they were laid down. But an undoubted example of a buried chalk-cliff of Eocene age Las yet to be recorded from the South of England, and, having regard to the prevailing evenness of the Cretaceous-Eocene con- tact, one hesitates to accept this explanation of the phenomena. A small normal fault, with a downthrow to the north-north- east, seems to bs the most reasonable interpretation. Disloca- tions of this type are widely and capriciously distributed over the South of England, and instances of their association with outlying masses of Lower Eocene strata, whose preservation they have in a large measure served to ensure; are not infrequent in some parts of the country — as, for example, in the Chiltern Hills. 1 Owing to the scarcity of flints, a considerable excavation would be needed to show the attitude of the Chalk at this spot. 2 ' Country around Bournemouth,' Mem. Geol. Surv., 1898, p. 2. GEOLOGY OF BOURNEMOUTH. 15 CHAPTER IV. LONDON CLAY. The loamy facies presented by this member of the Eocene system over the greater part of the Hampshire Basin becomes modified, in the direction of increased sandiness, as the forma- tion ranges westward ; the lithological change being accompanied by an attenuation which brings down the thickness from about 320 feet, or rather more, near Portsmouth, to about 100 feet at the outcrop in the neighbourhood of Wimborne. Near the latter town the London Clay is made up of hnely- sandy, laminated, pyritous clays of bluish-grey tint (where un- weathered), alternating with thinner, loamy, glauconitic sands, which, by chemical reactions, are often partly converted into firm or friable red iron-sandstone, near the surface of the ground. Flat concretions of cement-stone and clay-ironstone occur in the clays, but they rarely exhibit the calcite-filled fissures of the typical septarium. There is a well-developed Basement-bed, 6 or 7 feet thick, less evenly stratified than the rest of the forma- tion, and consisting of glauconitic loamy sand with seams of clay and one or more bands of rounded flint-pebbles. The London Clay hereabouts is not well provided with fossils. Rolled bits of lignite and other remains of drifted vegetable matter are not uncommon, but marine shells, which abound at several horizons in the country to the east and north, are seldom, if ever, encountered in the shallow sections ordinarily available. As far as the. writer is aware, a well at Christchurch, and another at AVimborne, have furnished the few examples hitherto recorded. There is, indeed, a decided falling-off in the frequency of this sort of fossils towards the western limit of the formation — a decrease not adequately accounted for by the decalcifying action of percolating rain-water ; for the London Clay of Dorset, though arenaceous, is not more permeable than some other, and richly fossiliferous, sandy clays of Eocene age, such as those of the Bracklesham Series. It would seem that the local conditions in Cuisian times were in some way unfavourable to molluscan life. At the outcrop between Lytchett Matravers and Hinton Parva the London Clay appears to maintain its usual, broadly-conform- able relations with the Reading Beds. The junction, where it can be seen, is sharply defined and even; its disregard for the current-bedding in the Reading Series, while indicating some erosion, does not by itself imply any stronger discordance between the two formations than might arise from the levelling of sand- 16 GEOLOGY OF BOURNEMOUTH. banks or other constructional irregularities in the upper surface of the Reading Beds, at about the time when the London Uay began to, be accumulated. Yet, in the small outliers about Badbury Rings, only a few miles to the north-west, there are indications of an overstep, whose existence is more clearly demon- strated in the adjacent area of the Ringwood map-sheet where tne London Clay in places -comes in contact with the Chalk., Here, then, near the existing limit of the Eocene strata in north-eastern Dorset, an abrupt change occurred in the character of the erosion which preceded the deposition of the London Clay : the Reading Beds were not merely smoothed and planed to a fairly uniform thickness, as seems generally to have been the case elsewhere ; they were more or less deeply channelled, and to some extent replaced, by the marine sediments which succeeded them- The fact that this transgression on the part of the London Clay took place in a north-westward direction, contrary to the existing local dip, suggests that the older strata hereabouts had been slightly raised or tilted into an exposed position by one of those ill-defined, wide-spread, crustal warpings of Eocene date which, in the South of England, partially foreshadowed the well-marked folding of later Tertiary times. Notes of Exposures. Little of interest is to be seen on the expanded part of the out- crop north-east of Wimborne. There are signs of the basal pebble-beds along the boundary south-east of Stanbridge, and grey and brown sandy clays appear in the lanes about Pig Oak and God's Blessing Green. A loamy sand with grains of glau- conite, near the middle of the formation, shows in a road-bank 350 yards north-west of Onslow. From the Basement-bed, traversed in a boring at Wimborne, Mr- Clement Reid records 1 an indeterminable bivalve, apparently Cyprina or Cytherea — one of the few fossil shells so far found' in the London Clay of the district. In the upper part of the brickyard north-west of the town, stratified grey and brown silty clays are worked for about 10 feet. These belong, to the lower beds, which extend north-westward in the rising ground about Hillbutts. Indications of a small outlier of the Basement-bed, beneath the Plateau Gravel on Pamphill, are seen in the lanes south-east of the Manor House- South and south-east of Corfe Mullen the London Clay con- tains much thin ironstone, pieces of which abound in the soil. The stone, composed in some cases of sand, and in others of fine loam, cemented by red and brown ferric oxides, occurs at several horizons, but is best developed near the lower limit of the forma- tion. Mr. Reid notes small exposures on the ridge west of the . Knoll, and at the northern boundary of Stony Down Plantation. In the thin capping of London Clay on the ridge above Warm- well Farm, north of Lytchett Matravers, ironstone occurs in bands sufficiently thick to be workable for building-stone. West of the farm large slabs and blocks of it project from the turf 1 'Eocene Deposits of Dorset,' Quart. Journ. Gaol. Soc, vol. lii, 1896, p. 491. LONDON CLAY. 17 oil the floor of one of the stone-pits, which have long been dis- used. On the same ridge the Basement-bed is clearly exposed in Sandv Lane, a quarter of a mile north of Warmwell Farm ; also on both sides of the cutting in which the ridge is crossed by the Poole-Blandford road, south-west of Rotting Hill. The salient features of these sections, which lie in a nearly straight east-and-west line, are represented in Fig. 4. Fig. 4. — Sections in the London Clay and Reading Beds, near Lytehett Matravers. Vertical Scale, 1 ineh = 20 feet. Poole-Blandford Road. 1, 2, Reading Beds; 3, 4, 5, Basement Bed of (6) London Clay. The first section (a) is shown in a sand-pit by the eastern side of the Blandford road, where the succession is as follows : — London Clay. Soil ; thin, sandy and gravelly. Feet. "6. Mottled grey-brown sandy clay with clay-ironstone nodules . . . seen 5 5. Seam of small flint-pebbk s ... ... 4. Fine sand with bands of grey laminated loam ; thin seams of ironstone'in the j upper part ... ... ... ... 7 ! 3. Band of flint-pebbles and small quartz l_ pebbles in coarse sand ... ... ,.i_. j-.. ("2. Fine buff and white sand, with fine e „ j ° \ seams and specks of lignite near the top; current-bedded... ... seen 8 s. I Inches. 4 The even bedding in the London Clay shows a dip of about 10° north-east. A persistent seam of iron pan in the upper part of bed (4) contains impressions of fragmentary leaves and small bits of lignite. The banks of the road-cutting (fig. 4 b), at the turning of the lane to Warmwell Farm and Lytehett Matravers, show about 6 feet of evenly interbedded, light-coloured sand and grey loam (1), the latter containing much ironstone in thin lenticles. At first sight this group might be taken for part of the Basement- bed of the London Clay, but it actually belongs to the Reading Series, of which the topmost, lignite-speckled sand (2), with crimson patches, appears at a level about 10 feet higher in a pit a few yards farther west. Here the lower pebble-bed (3), dipping 5° west, is succeeded by interstratified sands and loams (4), as in section (a) ; the upper pebble-bed probably being repre- sented by some rounded flints scattered in the subsoil at the top 18 GEOLOGY OF BOURNEMOUTH. of the excavation. About a quarter of a mile westward, a pit on the eastern side of Sandy Lane gives a section (fig. 4, c) rather different from the last. The Reading sand (2) is slightly car- bonaceous, as before, but is less brightly tinted; while the pebble- bed (3) has divided into two bands, with a foot of loamy sand between them; bed (4) has acquired a markedly lenticular struc- ture, and the dip has changed again, to about 10° east. The variable dip in these sections, though more apparent in the London Clay than in the Reading Beds, is due to disturbances affecting both formations alike, and not to original irregularities in their junction. Some of the upper beds of the London Clay are shown in shallow brick-pits east of Lytchett Matravers. In a yard at the eastern end of the village a few feet of brown loam overlies a more sandy loam with nodules of ironstone. Another brick-yard just south of Barrow Hill shows — Brown loam with seams of ironstone ... ... ... 1 ft. to 5 ft. Coarse buff sand with chips of flint ... ... ... 5 ft. to 2 ft. ; while a third brick-pit, a furlong south-east of the cross-roads near Barrow Hill, is in thin-bedded sandy clay, just below the Bagshot Beds, which come in close by with an ill-defined boundary. In the south-eastern part of the district the London Clay was traversed in a well-boring at Christchurch. The record 1 is diffi- cult to interpret, but it appears probable that this formation is comprised in the 184 feet of beds next above a " red mottled clay," which may be safely assigned to the Reeding Series. The beds referred to are described, in descending order, as — Feet. " Loamy sand ... ... ... ... ... ... ... 37 Clay, claystones, shells and pebbles ... ... ... ... 112 Green sandy clay .., ... ... ... ... ... 15." No satisfactory line of demarcation can be drawn in the over- lying 531 feet of sands, clays, and lignites, the greater part of which undoubtedly belongs to the Bagshot and Bracklesham Beds. 1 ' Water Supply of Hampshire,' Mem. Geol. Surv., 1910, pp. 72, 73. The thickness of the London Olay is there given as 325 feet, or about the same as at Portsmouth, an improbably high figure. GEOLOGY OF BOURNEMOUTH. CHAPTER V. BAGSHOT BEDS. Near Bournemouth, the London Clay is separated from, the Barton Clay by some 700 or 800 feet of predominantly sandy strata, which, with their wide outcrop, are mainly responsible for the distinctive scenery of the district. That these strata belong in part to the Brackieshani Beds, and in part to a lower series, indifferently termed Bagshot Beds, Lower Bag- shot Be,ds, or Bagshot Sands, has long been recognised, but, despite the close attention they have received during the last forty years; a satisfactory plane of demarcation between the two formations in this part of the country has yet to be discovered. Fossils of known determinative value are rare or wanting, and the difficulty of correlation due to this defect is aggravated by the lithological . changes which the Bagshot and Brackiesham Beds undergo as they range westward from the neighbourhood of Brackiesham Bay, and by the lack of compre- hensive sections in the critical area between the eastern and western ends of the Isle of Wight. A recapitulation of the views expressed by early writers in regard to the stratigraphical position of the beds exposed on the Bournemouth coast, since these were first described by Sir Charles Lyell in 1826, would serve no useful purpose in the present work, for comparatively little advance in knowledge was registered prior to 1877, when Mr. J. Starkie Gardner published the first of a number of papers in which the beds are described in detail, and their features discussed. Gardner distinguishes three broad divisions or series of the local Bagshot-Bracklesham Beds, viz. : — 1, the oldest, which he terms " Lower Bagshot," and which includes the pipe-clay and sands of Corfe and Poole; 2, the Bournemouth Freshwater Series, cor- related with the Lower Brackiesham Beds of Brackiesham Bay, and taking in the sands and lignitic loams between Poole Head and Bournemouth Pier ; 3, the Bournemouth Marine Series, in- cluding the sands, clays, and pebble-beds between Bournemouth Pier and Highcliff. Although the nomenclature is unsatisfactory, the above grouping is conceded to be the most natural hitherto proposed. Unfortunately, however, the limits of the Freshwater Series are marked by no persistent features which allow of their being traced inland, so that Mr. Clement Reid, when revising the geological map in 1893-4, was constrained to adhere to an older scheme of classification, according to which the Brackiesham Beds are restricted to Gardner's " Bournemouth Marine Series''; and the Freshwater Series, together with the underlying Pipe-clay 20 GEOLOGY OF BOURNEMOUTH. (or "Lower Bagshot") Series, are assigned to a mainly non- marine formation, which is designated " Bagshot Beds" in the index to Sheet 329 of the new-series Geological Survey Map. The present memoir necessarily follows the latter classification, but the distinction will be drawn, as far as practicable, between the two divisions of the Bagshot Beds, which are separated by Mr. Gardner chiefly on the ground of the dissimilarity of their fossil floras; and the reader should bear in mind that the higher of these divisions is, in all probability, the stratigraphical equivalent of part of the Lower Bracklesham (Lutetian) of Bracklesham and Whitcliff Bays. The Bournemouth Marine Series is dealt with in the next chapter. I. Lower or Pipe-clay Division. Near Dorchester, where the Bagshot Beds assume a gravelly facies, they cut into older members of the Eocene system, much as the London Clay does near Wimborne and Cranborne, but within the limits of the Bournemouth map no such discordance is observable. There appears, on the contrary, to be a gradual passage from the loams of the London Clay into the coarse, current-bedded sands, which, with their intercalated pipe-clays and subordinate loams, make up the lower 200 to 250 feet of the Bagshot Series'. The sands are usually white and yellow, with local patches of orange, pink, and crimson, while the clays, which occur mostly in lenticular masses up to 20 feet or more in thickness, vary in tint from white or palest grey to rather dark blue-grey, and, like the sands, frequently show ruddy mottlings. These lower beds appear to be the delta-deposits of a river which gathered most of its finer load of waste in some distant upland of granitoid rocks, situated, presumably, to the south-west. Their structure somewhat resembles that of the Reading Beds : there are similar indications of a rapid accumulation of sand- banks, alternating with slow deposition of mud in the inter- vening hollows. Casts of detached leaves of sub-tropical plants are met with in the clays, but they seem to be much less common in the Bournemouth district than in the vicinity of Studland and Corfe, immediately to the south. A " snlall round oyster," denoting salt or brackish water, has been observed in Furzebrook clay-pit, south-west of Slepe Heath, and there is a doubtful record of the occurrence of Cardita planicosta and Turritella imbricataria in " a sand-pit at Lytchett, near Poole." 1 Mr. J. S. Gardner remarks that the distinctive character of the Pipe-clay flora (of Studland and Alum Bay) " is due to the size and variety of the leaves ascribed to the genus Ficus and to the Leguminosse, in a scarcely less degree to a deeply cleft palmate Aralia, a tri- lobed leaf resembling Liquidambar, a deeply serrate Banksia, and other leaves referred to Comptonia, Dryandra, and Myrica. 1 O. Fisher, ' On the Bracklesham Beds of the Isle of "Wight Basin,' Quart. Jotirn. Geul. Sue, vol. xviii, 1862, p. 83. The latter two fossils occur together in the passage ben's between the London Clay and Bagshot Beds near Fordingbrido-e. Sue C. Rpid, 'Country around Eingwood,' Mem. Gecl. Sun-., 1902, p. 15. BAGSHOT BEDS. 21 .Few, if any, of these have been found in the Middle Bagshot [or freshwater] division of Bournemouth." 1 The Lower Division crops out in an irregular belt of ground which takes in the greater part of the area mapped as Bagshot Beds west of a line drawn from the mouth of Poole Harbour northward to West Moors Station. The pipe-clay beds are most fully developed around Poole Harbour : northward of Parkstone and Hamworthy they become thin and silty, and seem not to have been dug, as pipe-clay, farther north than Lytchett Beacon. Along the southern shore of the harbour the low cliffs or banks are cut mostly in white and yellow sands and friable sandstone, occasionally including rolled and angular pieces of elay, as on the western side of Arne Heath and the south-eastern shore of Brand's Bav. White and red-mottled pipe-clays appear, among other places, on the northern side of Arne Bay, at Pound Island, and at Fitzworth Point. Long, Green, and Furzsy Islands are composed mainly of sand and sandstone, the last-named islet, however, showing about 20 feet of laminated, carbonaceous, sandy clay, above the sand, in the cliff on the northern side- This dark clay, which also appears in the bluff on the southern side of Brownsea Island, and has been dug there to a depth of about 25 feet at the disused pottery, belongs to the upper division of the Bagshot Beds. Pipe-clay was formerly mined beneath it near the northern extremity of Brownsea, and is said to have yielded well-preserved leaves. North of Poole Harbour, excavations at Lake Clay Works and near Hamworthy Junction afford good exposures of a massive pipe-clay bed, 20 to 30 feet thick, overlain by coarse sands. Clear sections are to be seen also at Kinson Pottery and 'other works on the western scarp of the ridge north of Parkstone. The junc- tion with the overlying current-bedded sands, of which about 50 feet is shown at Constitution Hill, is even or slightly undulate, the pipe-clay being brightly reddened at the contact, and mottled with the same tint to a depth of 12 feet or more below it. An inlier of pipe-clay has been worked in Bourne Bottom, to the east. Numerous other sections are to be seen near the high road from Parkstone to Lytchett Minster, and in the heath-land between Beacon Hill and Broadstone. North of Broadstone the pipe-clays give way to pale silty clays in thin beds and partings, such as are to be seen in brick-pits at Forest Hill, and by the roadside at Elder Moor, east of Lytchett Matravers. Near Wimborne the argillaceous inclu- sions in the lower part of the Bagshot Beds are merely grey and brown loams, of which there are exposures in the brickyard on the western slope of Cole Hill, north-east of the town- 2. Upper Division, or Bournemouth Freshwater Beds. This series may be succinctly described as a complex of white and yellow sands, laminated carbonaceous clays and loams, and thin impure pipe-clays, arranged for the most part in overlapping 1 ' British Eocene Flora,' pt. 1, p. 16. Palceontog i-apliical Sue, 1879. 22 GEOLOGY OF BOURNEMOUTH. lenses of all sizes up to a few hundred yards in diameter. The clays, in particular, have yielded an abundance of leaves and other remains of plants. Usually impressions only of the leaves are met with, but in some cases the substance is preserved, show- ing the variations in texture, and even faint indications of autumnal colouring. The flora, which has yet to be fully worked out, includes representatives of dicotyledonous forest trees, conifers, ferns, palms, marsh vegetation, and parasitic fungi,, and appears to be much more varied than that of the loweri division of the Bagshot Beds. In the Freshwater Series, now under consideration, the lower beds are especially characterised by the presence of willow, the middle beds by palms and ferns, and the upper by araucaria, net-veined ferns, and eucalyptus.. Physically, the Upper or Freshwater division conforms, even, more closely than the Lower, to the delta type of deposits. The same river, as Gardner thinks, may have been responsible for the constituent beds of both divisions. The feeble recurrence of pipe-clays in the upper one, while it implies that the sources of sediment were to some extent the same as before, also suggests that the change in the character of much of the associated material may have been brought about by normal stream- adjustments, involving extensions of the original drainage-system' to new terrains of different lithplogical constitution. The principal section of the Freshwater Series is presented in the range of cliffs between Poole Head and the Lift east of Bournemouth Pier (fig. 5). A gentle dip brings in successively higher beds towards the east, but, owing to the lack of continuity in the stratification, the total thickness ean be only roughly estimated at about 400 feet. The complete succession is not shown, the lowest beds, probably between 50 and 100 feet thick,, having been worn away to below sea-level and covered by the recent shore-deposits near the mouth of Poole Harbour. Some of the missing beds are represented in the sands and dark sandy clays overlying the pipe-clays in Brownsea Island, and appearing at intervals along the shore of the harbour near Lilliput and The Elms. They have yielded unbored drift-wood and a few leaves,. e.g., of Anemia subcretacea (Sap.). The following account of the cliff-section is mainly a resume of observations recorded in Mr. J. S. Gardner's papers. 1 Con- siderable changes have occurred in the cliffs since these observa- tions were made, and some, of the specially fossiliferous seams- described by Gardner have long since been worn away or become obscured, but his description, on the whole, is still applicable, as- far as it relates to that part of the cliff which lies to the west of Bournemouth Pier. East of the Pier the section has been greatly obscured by the construction of the Undercliff Drive, and the planting of part of the slopes above it. Between Poole Head and Bateman's (or Flag Head) Chine a dark fissile loam with bored pieces of lignite, at the base of the cliff, is overlain by white and yellow sands containing flat lenses- 1 More particularly his • Description and Correlation of the Bournemouth Beds," Quart. Journ. Geol. Soe., vol. xxxv, 1879, pp. 209-228 ; vol. xxxviii, 1882, pp. 1-15. BAGSHOT BEDS. 23 03 W a x < GO ^ *« e ~~ In 0) hr i- br o 01 s M ca 01 as C3 C SA, , 8 C3 o s S I— I Cs 3 a? P3 S3 a s g S ^ 8 S a Hfq Bi 2 W 3° B n O *£ turn '3 .§ -s a 24 GEOLOGY OF BOURNEMOUTH . of black and grey to fawn-coloured clay with leaves, some forms of which appear to be restricted each to a single lens. ^ In the lower part of the largest of these clay-inclusions, between 300 and 400 yards in length, Gardner observed that the prevailing leaves were of laurel and willow-like forms, and often attached to the twigs; while certain leaves, like those of Stenocarpus and Acacia, were confined to a particular patch in this clay, at a spot about 100 yards south-west of Canford (or Sugar-loaf) Chine (fig. 5). Other fossils obtained from the same clay included a small oval Smilax with a thin twining stem, Lygodium kaulfussi Heer, and a fossil feather. Another patch, at the southern angle of Canford Chine, is characterised by the preponderance of a large hornbeam-like leaf, and large stipules and serrate leaves, apparently of willow. Between Canford Chine and Branksome Chine (otherwise Brank- some Glen or Watering Chine') 1 the black clays intercalated in the sands are mostly replaced by smaller lenticles of light-coloured sandy clay and white pipe-clay, the latter being here unfossili- ferous. About 100 yards east of the Watch Tower a small mass of laminated light clays and sands, about 25 feet across, yielded " remains of insects, flowers, leguminous pods, small detached willow-like leaves with smooth or serrated margins, stipules, laurel and Diospyros leaves." 2 Truncated lenticles indicative of contemporaneous erosion, like those figured by Gardner, are still to be seen in the lower part of the cliff. Here, and at the mouth of Branksome Chine, reefs of ferruginous sandstone, with pyrites and bored lignite, appear on the foreshore. In a dark clay overlying the first of these reef-inaking beds Gardner observed a palm-trunk, and a layer of hard lignite with a root-bed beneath it. Bast of Branksome Chine the bedding becomes less irregular,, and several distinct horizons can be traced. Black clays like those near Poole Head are succeeded upwards by sands with a marked band of black sandy clay, another of ironstone, and a discontinuous seam of impure pipe-clay. In the cross section afforded in the side of the chine these beds have a dip of 4° or 5 a south-east. Nearer Bournemouth they fall gradually to the shore, the pipe-clay band becoming more defined and yielding remains- of date-palm. A dark clay which comes in above is remarkable for the abundance of a Myrica-like leaf and a large pinnate palm. In a carbonaceous sand near beach-level west of Alum Chine several coniferous tree-trunks, bored by Teredo, were lately to- be. seen. 3 Between Great Durley Chine and Bournemouth Pier the clay- bands distributed through the sands are exceptionally rich in plant-remains, notably Gleichenia hantonensis (Wankl.)^ Hewardiaregia Stt. & Gard., Osmunda lignitum (Giebel), Chrys- odium lanzaeanum (Visiani), Sequoia couttsiae Heer, Iriartea, 1 The "Branksome Chine "of Fig. 5 is a shallow indentation of the cliff north- east of the chine now known by this name. 1 Gardner, Quart. Journ. Geol. Soc, vol. xxxviii, p. 6. 3 W. T. Ord, in 'A Natural History of Bournemouth, &c.' p. 312. (Bournemouth* 1914.) BAG SHOT BEDS. lo and Smilax. The most fossiliferous of these bands occur in the cliff below and a littie east of the Coast-guard Station. The fol- lowing passage relating to this part of the section gives a good idea of the complex structure of the Bournemouth Freshwater Series, and also illustrates Gardner's imaginative inter- pretation of the phenomena they exhibit: "The section of the beds just west of the pier presents at its base fine river-sand, becoming carbonaceous as the current diminished, then choked with the fallen leaves, and then a nearly stagnant pool. The deposition of mud was abruptly ended by an influx of coarse grit several feet in depth. The same process was repeated in the second leaf -bed ; the third pool was formed more suddenly, for the clay rests immediately on clean grit and is not discoloured, though leaf-impressions abound in it. It is margined, like a few other leaf-beds here, with a white marl penetrated by rootlets. The succeeding 5 feet of coarse granitic grit mark the passage of a considerable body of water, and of some swiftness, and the ironstone a period of stagnation. The white clay shows that water again trickled in, charged with enough sediment to bury the leaves before they decayed. This passes gradually into the black clay, small light patches with actually skeletonised leaves in them . . . penetrating the black stagnant clay, which is dark with decomposed vegetation and charged with sulphide of iron. In some layers ferns and seeds can still be traced; but in most the vegetable matter is thoroughly decayed. Layers of compact clay are separated by what are now films of carbonaceous matter, to the number of hundreds, each layer indicating possibly only a year. Over this is lighter clay, as if the water again began to trickle in more freely, and then once more follows the sudden change to grit. The reimbedded lumps of clay above show that many such sequences may have been broken up and swept away . . ." 1 On the farther side of the Bourne gap, and about 400 yards east of Bournemouth Pier, a bed of dark clay, which comes into the cliff about 40 feet above the beach, is underlain by light- coloured sands with angular blocks of clay containing remains of a flora differing from that of any of the lower beds seen to the west. The most abundant forms are an aroiid, a Eucalyptus , and an ArauCaria, the last-named being restricted to the beds east of the Pier, where it is the " prevailing fossil." The remain- ing beds of the Freshwater Series are mostly coarse sands with scattered logs of bored drift-wood, but the last of the fossiliferous bands definitely referred to this division is a thin pinkish clay with sandy laminae, formerly visible east of the Zigzag path in the recess till lately known as Steps Chine. This yielded many fern-fronds referable to Chry sodium lanzaeanum (Visiani), Osmunda lignitum (Giebel), Pteris eocaenica Ett. & Gard., Podoloma afline Ett. & Gard. (Phymatodes) . The fern-bed is overlain by a thin black clay (fig. 5 b), which Gardner considers to be of marine oriarin. Op. supra ett., pp. 11, 12, 26 GEOLOGY OF BOURNEMOUTH. Q ■pq I o Pq 02 Cb >-» £ 1')^, « Wliat with accumulations of talus, and municipal plantings and pro- hibitions, little evidence is now ob- tainable of the lateral passage of the upper members of the Freshwater Series into the Marine Bracklesham Beds that takes place near this spot. Mr. Gardner gives an unsealed dia- gram (here reproduced as fig. 6) which shows the Marine Beds inter- digitating with the Freshwater, and also thinning out between them and the apparently overlapping Bos- combe Sands. In the accompany- ing text 1 he ascribes the westward disappearance of the marine beds "partly to the rise of lower beds and partly to their passage into the Freshwater beds " ; and elsewhere he mentions ' ' three separate inter- calations of marine and Freshwater beds overlapping each other in rising succession," in the cliff "just be- neath the Bath Hotel." The lateral transition appears to be more gradual than one might infer from Gardner's statement that the marine beds give place to freshwater beds 1 ' within about 100 yards " ; a state- ment that can apply to a fraction only of the total thickness of strata (below the Boscombe Sands) exnosed in East Cliff. According to the Geological Survey map, and to ob- servations made by Dr. W. T. Ord 3 .before the construction of TJnder- cliff Drive, the Freshwater (Bag- shot) Beds were traceable beneath the Marine (Bracklesham) Beds for a distance of about 700. yards east of the Bath Hotel grounds, or 500 yard,s east of Steps Chine. The Marine Beds, in which marine fos- sils seem to be restricted to imper- sistent bands, themselves include leaf-bearing clays and loams that give no more evidence of having been laid down in sea-water than do 1 Op. cit., part I. Quart. Juurn. Geol. Sue, vol. xxxv, 1879, p. 223. 2 ' Excursion to Bournemouth,' Proo. Geol. Assoc, vol. xiii, 1894, p. 276. » 'The Geology of the Bournemouth to Boscombe Cliff Section,' pp. 4,, 6. Reprint .from Proc. Bournemouth Nat. Sci. Soc. for 1912-13, vol. v, pp. 118-135. BAG SHOT BEDS. '27 their analogues in the cliffs to the west of Bournemouth Pier. Inland, the Freshwater Beds are dug for sand and brick-earth in a large number of pits (some of them marked on the map) on the outskirts of Bournemouth, and in the heaths to the north. Few, if any, of these workings show features which may not be studied to better advantage on the coast, and only a small selec- tion will be noticed. Good exposures, to 20 feet or more in depth, are to be seen in pits in the sides of the combe east of Winton. These are opened in current-bedded sands with bands' and lenses of brown and greyish loam. A black loam, full of lignite, is worked in a pit on Poors Common, north-east of Heath Farm. Carbonaceous and light- coloured sandy clays overlie the pipe-clay in Bourne Bottom, and are now worked near the railway station at Bourne Valley, and in several yards west of New Town. In the brickyard half a mile south-east of Broadstone Junction Mr. C. Eeid observed impressions of leaves in ironstone nodules. Here, and in the cutting south of Broadstone Station, masses of white clay occur in the sands, which latter are well displayed, to a depth of 50 feet, in the Corfe Hills cutting. Dark, car- bonaceous clays, overlain by sands with irregular bands of iron- stone, are dug in the yard north of Dunyeat's Hill, and appear in the lower part of the railway cutting near Higher Merley. At Cudnell, west of Kinson, bored lignite occurs in red and grey mottled clay in the brick-pit north-east of the cross-roads. On the heaths and "commons north of the River Stour the clayey beds of this Series have been but little exploited. There are, however, some brick-pits in grey and white loams on both sides of the Pdngwood road north of Longham. 28 GEOLOGY OF BOURNEMOUTH. CHAPTEE YI. BEACKLESHAM BEDS. The strata mapped as Brackleshani consist chiefly of dark sandy clays and light sands with pebble-beds, accumulated for the most part under marine conditions, and possessing a simpler structure than the deposits described in the last chapter. They are an outcome of a movement of subsidence which, beginning at an earlier age, and becoming more pronounced as time went on, at length so far exceeded the local rate of sedimentation as to allow the sea gradually to extend westward over the fluvial beds of the Bournemouth Freshwater Series. That this sub- mergence was widespread, is evident from the present distribu- tion of marine sediments of Bracklesham age in the South of England alone; but that the limit of the submerged area lay not far westward of the Bournemouth district is hardly less evident from the character of the deposits here in process of formation at the time- At intervals throughout the Bracklesham Marine Series, and especially in its lower beds, the debouchure of a copious land-drainage somewhere in the neighbourhood, and on occasion close at hand, is indicated in the profusion of drifted vegetable matter, ranging from trunks of palms and conifers down to the fine carbonaceous mud that gives so much of the Bracklesham clay its dingy colouring. The Bracklesham Beds near Bournemouth are divided into four stages, which are named below in descending order, with their estimated thicknesses on the coast: — Feet. Highcliff Sands 67 Hengistbury Beds ... ... ... ... ... 61 Boscombe Sands ... ... ... ... ... 100? Bournemouth Marine Beds 150-200? Casts of mollusca are not uncommon in places. Such as are of known value in the correlation of the above-named divisions with the stages recognised in the Bracklesham Series in other parts of the country, have been obtained mostly from the High- cliff Sands and the Hengistbury Beds. These two divisions each contain a .small fauna of Upper Bracklesham-Lower Barton facies, and are classed with the higher of the beds characterised by the presence of Nummulites variolarius at Whitecliff Bay, in the Isle of Wight. The Boscombe Sands, unfossiliferous save for scanty plant-remains, also probably belong to the Upper Bracklesham ( Auversian) ; while the Bournemouth Marine Beds, which have yielded a few species having a wide range in time, BRACKLESHAM BEDS. 29 are referred to the upper part of the Lower Bracklesham Beds, with Nummulites laevigatus. Bournemouth Marine Beds. Though styled 'marine,' in order to emphasise their distinc- tion from the Bournemouth Freshwater Series, these feeds probably have less right to that title than the three other Bracklesham stages which succeed them. ' Fluvio-marine ' were a fitter epithet, for feut few definitely marine episodes have so far been observed, and the beds not only contain much drifted debris of land and swamp vegetation, but also preserve vestiges of marshy ground, penetrated by the roots of small plants. The Bournemouth Marine Beds present the seaward facies of a deltaic formation such as the Freshwater Series, if not of that series itself. There is evidence of a lateral passage between- the two sets of beds, as already stated, but the measure of their con- temporaneity cannot be. gauged with the data at present avail- able. Possibly a comparison of their respective floras, when these have been fully described, may shed more light on the subject. The grey and yellow sands and dark laminated clays of this division are exposed in the cliffs east of Bournemouth Pier for a distance of four miles. They come in as a thin wedge near the Bath Hotel, and, after expanding rather quickly east of the Zig- zag path (Steps Chine), theyoccupy roughly the lower half of the cliff between East Cliff Lift and the mouth of Boscombe Chine, beyond which place their upper limit declines to the beach south of Cellarfield Farm, near Southbourne. At the Zigzag the Boscombe Sands near the top of the cliff are underlain by about 50 feet of interbedded loams and sands, but how much of this thickness belongs to the Marine Beds the writer could not determine. Marine fossils seem rare; indeed, Dr. W- T. Ord, who has repeatedly examined the beds at and about this spot during the last 20 years, states that the only thing of the kind he has found is a doubtful example of Callianassa. Mr. J- S. Gardner, however, observed near here an exposure of fos- siliferous beds, 30 to 50 feet above the beach, that gave indica- tions of a vertical transition from freshwater, through brackish, to marine conditions. His description 1 of the descending sequence (beginning some distance below the top of the cliff) is given here, in a slightly condensed form: — 1. Dark sands, with preen grains containing masses of Ostrea, dorsata ? (coated with Flustra), Area appendicwlata ?, Modiola nysti ?, Tellina tenuistriata, Galyptraea trochiformis ?, Phorus agglutinans (Xenophora), Natica labellata, Gerithium. [Thickness not stated.] 2. Liver-coloured clay (turning black on exposure), with abundant remains of Gallianassa, of a Cnio-like shell, and of Bryozoa, Ac, 15 feet. (The surface of the bed beneath is eroded for about 6 inches, and filled in with the overlying clay.) 3. Stiff black clays passing into lighter liver-clays at bottom, and, after a break, into liver-clays with ferns. Very dark sandy clay. White or ash-coloured sand with lignitic bands. 30 feet. 1 ' Description and Correlation of the Bournemouth Beds. Part I. Upper Marine Series.' Quart. Journ. Geol. Soc, vol. xxxv, p. 224. c 2 Bournemouth Marine 30 GEOLOGY OF BOURNEMOUTH. In a bed of dark clay (3 feet), seen close by to the east, the same author found abundant seeds and fruits, associated with various dicotyledonous leaves, including a Dryandra, and also Cactus spines. About 200 yards east of the Lift the cliff shows the following succession, beginning at the top : — Feet. Plateau Gravel ... ... ... ... ... ■•■ ■•• •> Boscombe (9. Yellow and orange sand ... ... about 20 Sands. \_8. Tine white sand ... ... ... ... 14 ' 7. Drab and yellow sands and loams 4 6. Greenish sand with oysters ... ... ... 2 5. Light grey sand ... ... ... ... 15 4. Carbonaceous sand with rolled pieces of J clay containing leaves 9 tj™ s 3. Grey carbonaceous sand, full of angular lumps of dark clay ... ... ... 13 2. Dark clay with much lignite, pyritized plant-atems, leaves of palm, &c.... ... 6 1. Coarse sand with seams of clay ... ... 2 Bed 2 is a lens of coarse breccia, having a rude oblique bed- ding, and containing blocks and slabs of clay up to a foot at more in diameter. The brecciation is contemporary with the bedding, but an explanation recently advanced, 1 attributing it to the direct action of storm-waves, is not in keeping with the angular shape of the fragments of clay. It is possible, however, that a violent storm gave rise to the breccia in another way — namely, by so altering the set and velocity of the bottom-currents as to cause them to undercut, and so induce block-foundering of, the bands of clay contained in the sands previously laid down. The greensand (6), which has yielded casts of a small Ostrea, appears to be on the horizon of the similar bed (1) of the pre- ceding section. Farther east the bedding becomes rather more regular, and the lower half of the cliff consists mainly of grey, laminated sandy clay. At intervals along the base of the cliff as far as Boscombe Chine small patches or pockets of seeds and fruit, ranging up to about the size of a hazel-nut, have been observed. Dr. Ord has lately called attention to a prolific patch of the kind, in stiff bluish-white clay, about 300 yards east of the Lift. Near by, another patch, no longer discernible, yielded to Mr. Gardner many seeds and fruits doubtfully referred by him to Hightea, Cucumites, and Petrophiloides. About 250 yards eastward Nipadites makes its first observed appearance, in dark lignitic loamy sand, in the lower part of the section, where it is associated with fruits resembling those of Anona and Hightea minima Bowerb., all being more or less rare at this point. For the ensuing 400 yards or so much of the section is obscured by slips, and nothing of particular interest is seen. Nearer Boscombe, casts of a small Ostrea are observed at intervals in dark grey and greenish sandy clay, and have been traced as far as the bluff on the eastern side of Boscombe Mouth. In the last 300 yards, 1 W. T. Ord, 'The Geology of the Bournemouth to Boscombe Cliff Section' (reprint), p. 7. BRACKLESHAM BEDS. 31 ending near Boscombe Pier, the lowest beds in the cliff are mostly dark lignitic sands, rich in fragments of wood and pine-stems. Gardner mentions a drab clay with impressions of leaves and fossil roots or sedges, at a spot about 200 yards west of Boscombe Chine. At the eastern angle of the chine a patch of greenish clay, filled with branches of conifers and tufts of leaves referred to Dry- andra, was formerly to be seen. Stouter branches bored by Teredo, and seeds, were observed in an underlying ash-coloured sand, and masses of small oysters in a liver-coloured clay close by to the east. These features are no longer discernible. The oyster-bed is believed to be the easternmost of the Bournemouth Marine Beds that has yielded remains of marine or brackish- water mollusca, excepting the borings referred to Teredo or Pholas. At Honeycomb Chine (Fig. 7), 250 yards east of Boscombe Pier, Nipadites reappears in force. The descending section here is : — Feet. Plateau Gravel 8 Boscombe Sands 7. Yellow and white sand 50 ''S. Pale grey sand with specks of HgDite ... 5 Dark reddish ash-coloured lignitic sand with abundant Nipadites ... ... 1-3 4. "White and ash colouredlignitic sand with a few fruits resembling Cucwmites, Petrophiloides, and Hightea ... ... b' 3. Dark grey sand with rolled lignite ... 2 2. White sandy clay, bored by Pholas P ... 3 Ijl. Grey sand, to beach-level ... ... ... 12 Near the mouth of this chine,. and probably in bed 3, Gardner saw the base of a palm-stem of which the bark and woody struc- ture were in perfect preservation. Entangled in the roots was some of the white clay in which the tree had been embedded before it was washed, as he thinks, from the older beds nearer Poole. 1 The Nipa-lvxnts, which all seem to have germinated, are inferred to have floated to this spot, as they have escaped the attrition which the water-worn vegetable debris associated with them has evidently undergone, and it appears probable that the parent trees grew in a tidal swamp not far away. Like many other plants preserved in the Eocene Beds of this district, Nipa- dites (N. parkinsoni? (Brongn.) ) has a curiously restricted lateral range. Despite its abundance in the sides of Honeycomb Chine, no examples are known to have been found in the Bournemouth Marine Beds in the adjacent cliffs to the east, or within three- quarters of a mile to the west. Even more restricted in its occur- rence is the cactus Palmacites daemonorops Heer, of which many perfect limbs, associated with a Sequoia-like conifer, have been found in an obliquely-bedded lignitic sand in the lower part of the cliff 120 yards east of Honeycomb Chine. This cactus, first described from the Upper Oligocene of Bovey Tracey, has so far been obtained from no other part of the Bournemouth cliff 1 Op.cit., p. 221. Bournemouth: Marine Beds. 32 GEOLOGY OF BOURNEMOUTH. section. Gardner notes the occurrence of sharks' teeth in the same bed. Yig. 7. — Honeycomb Ohine, west side. (After a photograph by W. T. Ord.) X. Position of Nipadites Bed. Hence eastward, to their point of disappearance beneath the beach on the farther side of Southbourne, the beds present few features of interest, though dicotyledonous leaves (sometimes preserved in films of pyrites), bean- and pea-like pods, and bored tree-trunks are not uncommon in the grey-and-liver-coloured clays and loams in the lower part of the cliff. BKACKLESHAM BEDS. 33 As it is seldom possible to distinguish the Bournemouth Marine Beds from the sandy strata above and below them away from the coast, notices of the chief inland exposures of this and other local divisions of the Bracklesham Series are deferred to the end of the present chapter. Bos combe Sands. Litholog'ically, this is the least varied of the four divisions : it consists almost entirely of white and yellow current-bedded sands with included bands and thicker lenticular masses of shingle, composed of well-rounded flint-pebbles, with small pebbles of quartz, and, more rarely, of quartzite. Much of the sand is of fine texture, and where unprotected is quickly eroded by the wind. In the coarser sands and grits the subangular and rounded par- ticles of quartz that make up the bulk of the deposit are often mingled with conspicuous white chips and flakes of flint which, at first glance, look like comminuted shells, The Boscombe Sands are evidently of littoral origin, and are believed to mark the landward migration of bars and shoals during, and in consequence of, the subsidence mentioned at the beginning of this chapter. Their structural features seem to indicate tha.t the prevailing movement of material was westward, and some of the shingle-beds are banked up as if by waves running in from the east. Near their upper limit, however, a change to more tranquil conditions of deposition, in deeper water, is implied by a more orderly arrangement of the bedding, and by the incoming of seams of sandy clay. Like the underlying division, the Boscombe Sands make their first appearance on the coast, going eastward, near the top of the rise east of Bournemouth Pier; coming in gradually beneath the Plateau Gravel. At this spot they appear to have just over- lapped the Bournemouth Marine Beds, or to have overstepped them; it is hard to say which, owing to the perplexed structure and indefinite boundary of these beds. Pebbly bands recur in the lower part of the Sands, but have no great development to the west of the Lift. A thin bed of ferruginous sandstone occurs between the Lift and the Zigzag path. Nearer Boscombe the Sands are about 50 feet thick under the highest part, of the cliff. A few hundred yards east of Boscombe Chine the pebble-beds rapidly expand into a mass of coarse sandy shingle, upwards of 30 feet thick, which has been described as forming, "for its limited extent, the most important conglomerate-bed in the English Tertiaries." 1 Eastward, the pebble-beds continue with varying, but generally decreasing thickness, to Southbourne, where they die out, the base of the Sands passing shortly after below the level of the beach. In the low cliff between Southbourne and Hengistbury Head the upper beds of the Boscombe Sands are shown in an inter- rupted section, a few feet deep, beneath Valley Gravel. Here the Sands become more evenly stratified, and contain occasional 1 J. Preatwich, ' On the . . . Coast Section from Christchurch Harbour to Poole Harbour,' Quart. Journ. Geol. Soc, vol. v, 1848, p. 47. 34 GEOLOGY OF BOURNEMOUTH. seams of carbonaceous loam and iron-sandstone. Where the cliff rises again, at the western end of the Head (Warren Hill), it displays about 30 feet of the Sands, up to their junction with the Hengistbury Beds, which come down to the shore, in their turn, half a mile farther east. The highest parts of the Boscombe Sands exhibit abrupt variations in tint; white passing laterally into ash-grey, black, or coffee-colour, the darker areas often having lighter stripes and mottlings, which give the brown parts the appearance of tortoise-shell. Despite these colour-changes one can discern the succession of beds made out by Mr. J. S. Gardner, which is briefly as follows : — Feet. Hengistbury Dark sandy clay with a layer of flint-pebbles Beds. at the base. 3. White to grey lignitic sand, with 2 to 4 fairly persistent layers of flint-pebbles in the upper part ... ... ... ... 8-12 (?) The lower part indents, and contains Boscombe j detached masses of, the underlying Sands. \ darker sand. 2. Coffee-coloured and black sand (changing to white and pinkish in places), bored by Pholas? 17 1. White sand ... ... ... seen, about 7 In a black loam or loamy sand (probably Bed 2), at or near the foot of the cliff 150 yards south-west of the Watch House, Mr. Clement Reid found a specimen of Nipadites. Hengistbury Beds- These are sub-littoral marine deposits, laid down in water of moderate depth, but still decidedly deeper than that indicated bv any of the Bracklesham Beds described in the foregoing pages. They are divided into two unequal sub-stages, distinguished as Lower and Upper, respectively 12 and 49 feet thick. The Lower Hengistbury Beds consist of dark olive-green sandy clay; coarse, roughly laminated, and containing much glau- conite. A layer of flint-pebbles — some of large size — with small pebbles of white quartz, occurs at the base, and similar pebbles are thinly disseminated throughout the clay. There are numerous thin bands of lignitic matter, also coniferous fruits, and logs of bored drift-wood. Animal remains seem to be represented merely by occasional foraminifera (notably a Haplophragmium), indistinct casts of lamellibranchs, and sharks' teeth. The full thickness of this sub-stage is seen only on the south-western face of Hengistbury Head. About midway along that face the dip (2° or 3°) carries its lower limit to the shore, and thence, round the Point, the upper limit declines along the lower part of the cliff, and disappears close to the north-eastern angle of the head- land. The Upper Hengistbury Beds are of more interest. These are mainly composed of laminated silty clays, glauconitic in places and with partings of fine sand, the latter material also forming i Op «»«., p. 215. BRACKLESHAM BEDS. 85 irregular bands at the junction with the overlying Highcliff stage. A pale shade of chocolate is the dominant tint in the clays. The most remarkable lithological feature is the dark, red-brown, concretionary ironstone, which forms from two to five courses of flat 'doggers,' ranging up to 4 yards or so in diameter. These concretions are made up of minute granules of iron carbonate stained by iron peroxide, mingled with fine angular quartz-particles, 1 like those of the enclosing silty clay. Their composition can best be observed on the shore, which is strewn with masses of the stone, washed clean by the waves. A small proportion of them are septarian; nearly all show seams and specks of lignite ; and not a few contain flattened tree-stems, the bark of which has been converted into glossy coal. The broadest trunk hitherto observed measured 2| feet, but the usual limit is about 1 foot. Sharks' teeth and vertebrae are not un- common. Off-shore the Head is skirted by a reef of these fallen concretions, known as Beerpan Rocks; and Christchurch Ledge, farther out, most probably marks the south-eastern continuation of the ironstone outcrop towards Alum Bay, in the Isle of Wight. The clays of the Upper Hengistbury Beds contain casts of marine shells which in some cases are sufficiently well preserved to admit of determination. Their presence has long been known. Sir Joseph Prestwich, writing in 1848, 2 states that he had found a few Bracklesham and Barton Clay species above the "lower part of the clay " at Hengistbury; and a note engraved on Sheet 16 of the original Geological Survey map records Car- dium semigranulatum, Cytherea obliqua, Pinna affinis, Thracia sp., Lamna elegans, Otodus appendiculatus , and leaves of dico- tyledonous plants, but without indicating their precise horizons- Mr. J. S. Gardner, however, seems to have observed neither shells nor leaves in these beds on the coast, and it remained for Dr. F. R. Cowper Reed to make the chief contribution to our know- ledge of their fauna, so recently as 1913. 3 In the disused quarry which almost cuts the headland in two, Dr. Cowper Reed found a number of fossils between the highest and the second band of ironstone concretions, i.e., between 6 and 12 feet below the base of the Highcliff Sands. Six of these he identifies as " Nuculana (Leda) minima, Protocardium turgidum, Corbula pisum, Crassatella sulcata, Panopea (Glycimeris) inter- media, Calyptraea aperta." The remaining species, "about which," he says, " more or less uncertainty exists, or which are indeterminable," comprise, " Anomia lineataf, Area duplicataf, Pectunculus (Axinaea) dissimilis?, Cardium cf. formosum? , Cyrena cf. crassaf, Meretrix (Cytherea) cf. incurvata?, Callista (Cytherea) sp., Cardita sulcata'?, C. (small) sp, Corbula ficusf, C. cuspidataf, Tellina sp., Turritella sp., Callianassa? sp.," also a vertebra of a small fish, and plant-remains. About 8 or 9 feet below the second ironstone band Dr. Reed obtained "Mytilus affinis and 1 F. Chapman, ' On some Foraminifera from . . Hengistbury Head, Hamp- shire," Geol. Mag., 1913, p. B58. 2 Op. eit., Quart. Journ. Geol. Soc, vol. v, p. 45. ■ 'Note on the Eocene Beds of Hengistbury Head,' Geol. Mag., 1913, pp. 101-103. 36 GEOLOGY OF BOURNEMOUTH. Tornatellaea nysti " from dark chocolate clay with grains of glauconite ; and at the base of the cliff at the north-eastern corner of the headland he found an example of Schizaster d'urbani, at a horizon estimated to be 10 feet lower still. The fossils occur in patches, and the present writer could discover only rare and faint impressions of shells at the horizons indicated. Internal casts of a small Corbula (C. pisum? J. Sow.) are common, with those of some other lamellibranchs, in a band 2 feet above the lowest layer of ironstone nodules, about the middle of the south-eastern side of the Head. Four species of arenaceous foraminifera are recorded from the "Upper Hengistbury Beds by Mr. F. Chapman 1 , who infers, from their poor development and small size, that they had lived under estuarine conditions. Dr. Cowper Reed points out that the fossils found by himself support Prestwich's suggested correlation of the Hengistbury Beds with the Barton Series, 2 but his statement that " all the species identified with certainty belong to the Barton Beds, and some are typical of them," appears open to question, since of the six species (in the first list) -positively identified by Dr. Reed, five are common to the Bracklesham and Barton; Protocardium turgidum (Sol.) being the only one not hitherto recorded from beds below the Barton Clay. 3 Although the Bartonian element is more strongly marked in the doubtful species, and in the few other species (excepting the Bracklesham echinid, Schizaster d'urbani Forbes) found at lower horizons, it hardly warrants the transference of the Upper Hengistbury Beds from the Bracklesham to the Barton Series, as these series are at present defined ; nor does Dr. Cowper Reed urge that this transference should be made. It is generally acknowledged that the Upper Bracklesham and Lower Barton faunas have much in common, and that no great break occurs between them at the particular horizon selected as the joint limit of their respective stratigraphical stages. That horizon is the base of the Zone of Nummulites elegans, and there is little room for doubt that the Hengistbury Beds lie below it. Highcllff Sands. The Upper Hengistbury Beds are conformably overlain by yellow and white sands containing thin seams of pipe-clay near their base, which is marked by a loamy band of dark yellow or orange tint. These beds form the higher part of Hengistbury Head, and are exposed in the cliffs on the south-western and south-eastern sides. They attain a thickness of about 25 feet under the Watch House at the highest point, and are covered unevenly by Plateau Gravel. The sands of Hengistbury Head are separated by an interval of 1^ miles from those of the typical section between Highcliff 1 Log. cit. 2 Prestwioh, op. cit., p. 44. s See E. B. Newton, ' Systematic List of the F. E. Edwards Collection of British Oligocene and Eocene Mollusca,' Brit. Mus. (Nat. Eist.), 1891, and 0. Fisher, infra cit. BBACKLESHAM BEDS. 37 Castle and Mudeford, but their correlation with the latter is sup- ported by their general lithological similarity, by a consideration of the local dips, and by Mr. Gardner's recognition of the under- lying clay and ironstone in temporary exposures in the inter- vening ground. 1 It is unnecessary here to enlarge upon the difficulties that attend other possible readings of the local Bracklesham succession. One of the connective exposures observed by Gardner was at Mudeford Ferry ; the other, further north-east, was in the left bank of the Run, which at the time was sapping the base of the cliff, and had laid bare a " dark sandy clay with ironstone nodules" beneath the lowest bed of the High- cliff Sands now visible on this part of the coast. It would seem, therefore, that the cliff south-west of Highcliff Castle, which shows the junction with the Barton Clay, affords at times a com- plete section of the Highcliff Sands. Here, as at Hengistbury Head, the lower beds consist of little but sand, but near the upper limit of this stage the change to the relatively deep-water conditions of the Barton Clay is heralded by the appearance of glauconitie loams. The descending sequence is as follows (figs. 8, p. 42; 13, p. 58): — Barton Clay. 9 Highclife Sands. Feet. Inches. 7. Zone of Nummulitea elegans : Dark green sandy clay, with bright green glan- conite and large grains of quartz, 9 in. '6. Red-brown marly or clayey sand, loosely bound by iron oxide ... ... ... 5. Dark-green and pale chocolate sandy clay, with bright green glauconite and a few flint-pebbles ... ... ... 9 4. Band of rounded and sub -angular flint pebbles and small quartz-pebbles in dark loam ... ... ... ... 1 3. Light-grey and white current-bedded sand, becoming loamy and darker near base. Band of flint - pebbles (6 in.) at base ... ... ...about 2. "White and light buff current-bedded sand ... ... ... ... seen Level of beach Below this J. S. Gardner saw 2 : — ( Loose white sand. 10 ft. \ Hard dark clayey sandstone, with 1. < scattered pebbles at base, 1 ft. I Hard yellow mottled sand shading to \ { white, 6 ft. ) resting on [P Hengistbury Beds] Dark sandy clay with ironstone nodules. 33 6 L 67 1 Gardner, op. cit., Quart. Joum. Geol. Soc, vol. xxxv, p. 211. 2 Op. cit., p. 211. His brief account of the section is by no means clear. Though it is not explicitly stated that the lowest sands, now grouped as bed 1 , underlie those numbered 2, the words used hardly admit of any other interpretation. Yet, on the ground, it seems improbable that as much as 17 feet of beds could have been exposed below the existing lower limit of the clifl-section at or near the spot which Gardner indicates. Moreover, his description of bed 1 is so nearly applicable to bed 2 and the lower part of bed 3 as to suggest that it may indeed refer to them. If it does, then the thickness of the Highcliff Sands is probably about 50 feet. 38 GEOLOGY OF BOURNEMOUTH. Bed (2) is almost horizontal at Cliff End (iig. 13, p. 58), but dips below the beach at a short distance to the east. Many of the flint-pebbles in the loamy base of (3) are white throughout, and can be crumbled between the fingers. The purer sand above shows concentric banding. Bored wood is plentiful, and there aresome indistinct casts of lamellibranchs. Early observers mention a band of tabular ironstone-septaria in this bed, possibly in the eastern part of the section, now observed by the protective works below the Castle. The higher and more conspicuous pebble-bed (4) persists throughout the existing exposure of the Highcliff Sands, but Osmund Eisher 1 states that the pebbles almost dis- appear eastward, while their enclosing dark sandy clay or loam becomes "full of impressions of fossils," including those named below : — " Murex minax Oytherea lucida ? Fusus carinella (common) suberycinoides (common) Voluta nodosa , trigonula ? Serpula Crassatella sulcata Dentalium (large species) compresaa Area duplicata Sanguinolaria hollowaysi Cardium parile Corbula gallica porulosum pisum Cardita (ribbed) Panopaea." In the pretty green-speckled clay (5), which also contains casts of shells, Fisher notes a smaller and different assemblage, com- prising : — " Fusus pyrus Cardita (small, ribbed) Pyrula nexilis Cytherea (a Barton species) Voluta nodosa ? Crassatella costata Dentalium Corbula pisum." Cardium semistriatum The thin ferruginous bed (6) may be regarded as a passage-bed between the Bracklesham and Barton. Messrs. Gardner, Keeping, and Monckton, 2 who include it- in the Bracklesham Beds, describe it as "underlying and defining the Zone of Num- mulites elegans (7),' the "Nummulina Prestwichiana bed" of Eisher, though Fisher himself obtained the zonal nummulite from this "foxy-red" band (6) also. But the nummulite in ques- tion seems to be rare at this horizon, though it is common or plentiful in the greenish clay a few inches above. Bed 6 con- tains ill-preserved shell-casts, among which Fisher identified: — " Ancillaria canalifera Cardita (small, ribbed) Voluta (small, with distant ribs) Modiola nodosa? Corbula pisum Turritella imbricataria Thraoia Cardium parile Echinoderm." It may be observed that the majority of the molluscs speci- fically named in the above three lists are common to the Brackle- sham and the Barton Beds. The most distinctively Bracklesham forms, Cytherea lucida? ( = Meretrix nitidula (Lam.) ), C subery- cinoides Desh. (Meretrix), C. trigonula? Desh. (Meretrix), and 1 'On the Bracklesham Beds of the Isle of Wight Basin,' Quart. Joum. Oeol. Soc, vol. xviii, 1862, pp. 88, 90, 91. 3 ' The Upper Eocene, &c.,' Quart. Joum. Oeol. Soc., vol. xliv, 1888, p. 587. BRACKLESHAM BEDS. 39 Sanguinolaria hollowaysi J, Sow. (Gari 1 ?), are recorded from trie pebble-bed (4), which Fisher regarded as the highest member of the Bracklesham Series. Inland Exposures. The difficulty, already alluded to, of distinguishing the several stages of the Bracklesham Series inland, arises not only from the small dimensions of the sections and the weathered condition of the beds, but also from the changes in character and thickness which the latter undergo within a few miles of the coast. Collectively, the Bracklesham Beds become thinner northward, but the Highcliff Sands (assuming the identification to be correct) gain considerably in thickness for a few miles inland. It is hardly less difficult to determine the extent of the Bracklesham outcrop, and the line denoting the Bagshot-Bracklesham boundary on the Bournemouth map-sheet must be taken as largely con- jectural. Near Boscombe the Bracklesham Beds extend farther north than is suggested by the broken boundary-line on the map, for beneath the Plateau Gravel dug at the southern end of Little Down Common there is a markedly glauconitic sand, probably belonging to the Bournemouth Marine Beds. North of the Stour, brown loams show in the road-banks at Dudsbury, and there are small openings in grey, brown, and red- dish sandy clays in the heathland near Parley Barrow. Possibly these beds also belong to the stage just mentioned, and the flint pebbles freely scattered in the soil above may be relics of a pebbly episode in the Boscombe Sands. To the east, beds of white and light-grey clay crop out in a belt of ground roughly parallel with the Stour in the southern part of Barnsfield Copse, and can be seen in small openings north and east of Keys House. Carbonaceous loams alternating with sands are shown in a brick-pit half a mile west of David's Hill; and in the sandy ground about Matchams Plantation and Fox- bury Hill there are surface-indications of pebble-beds at higher horizons. Whether any strata younger than the Boscombe Sands are preserved in this part of the district, it is difficult to decide. To judge from the succession at Ramsdown, described below, it seems not unlikely that the sands and pebbles in the high ground near Matchams Plantation belong in part to the Highcliff Stage. In the prominent ridge of Ramsdown and St. Catherine's Hill the Hengistbury Beds and Highcliff Sands both appear to be represented. Mr. J. S. Gardner 1 gives the following section, seen by him in a brick-pit north of St. Catherine's Hill : — Feet. Inches. " Highcliff Sand P Orange sand and clay ... ... ... 3 ( Ash-coloured clay ... ... ... ... 3 Hengistbury" \ Drab clay with iron concretions exactly Beds. 1 as at Hengistbury Head ... ... 5 (. Dark and very lignitic sand ... ... 2 6 White and yellow sand with layers of light clay almost white at base ... 3 " 1 Op. cit., p. 216. 40 GEOLOGY OF BOURNEMOUTH. Gardner thinks that, in spite of the thinness of the supposed Hengistbury Beds, "there can be but little doubt of their iden- tity." The brick-pit has long been disused, but the succession above described can still be made out in its degraded sides. The chocolate (not drab) coloured clay now visible resembles that of Hengistbury, but is more carbonaceous and less silty. A few ironstone concretions of the Hengistbury type, though of smaller dimensions, lie on the floor of the working. Between the top of the pit and the summit of the steep knoll which rises imme- diately to the south-west of it upwards of 100 feet of higher beds " come in, consisting, as far as can be seen, of yellow sands with some intercalations of grey sandy clay. If these beds all belong to the HighclifE Sands, that division is at least twice as thick here as it is at Highcliff, about 5 miles distant. Light current-bedded sands with seams of coarse grit underlie the gravel on St. Catherine's Hill, and are exposed for 20 feet in the Watch Tower hillock at the southern end of the ridge. East of the River Avon the highest; sandy beds of the Brackles- ham Series crop out in a low bluff between Bransgore and Wood- end, and can be seen in pits at Nea Croft. GEOLOGY OF BOURNEMOUTH. 41 CHAPTER VII. BARTON BEDS. East of the River Avon the Bracklesham Beds are overlain by a series of marine deposits remarkable alike for the number, variety, and good preservation of their fossils. On a palaeonto- logical basis this series is divisible into three stages and a number of sub-stages or zones, but the Geological Survey map recognises only two, broad, lithological divisions, which it treats as separate formations under the names of Barton Clay and Barton Sands. Fortunately, the two systems of classification are not in conflict, for in the Hampshire Basin, at any rate, the common limit of the Barton Clay and Barton Sands is also that of the Middle and Upper stages of the whole Barton Series. As the type section in Barton Cliff lies partly outside the limits of the country shown on the Bournemouth map, and is reviewed at some length in the lately-published Geological Survey memoir on the Lynrmgton district, a brief summary of its leading features will suffice for the present work. A more detailed account of the Barton section, with references to earlier literature, will be found * in Messrs. Gardner, Keeping, and Monckton's paper on "The Upper Eocene, etc.," 1 from which fig. 8 is copied, with slight alterations. The base of the Barton Clay appears near Cliff End, south-west of Highcliff Castle, and is brought down to the shore by the gentle eastward dip at a point about 750 yards west of Chewton Bunny. The junction with the succeeding Barton Sands — plainly marked both by the contrasting colours of the clays and sands, and by a ledge in the cliff — comes into the section about 3 furlongs east of Chewton Bunny, and the upper limit of the whole series appears about 1J miles beyond the eastern boundary of the Bournemouth district, near Becton Bunny. The Barton Clay is composed of dark sandy, glauconitic clays and loams, with a total thickness of 102 feet. Of this thickness, rather less than half belongs to the Lower Barton stage, 2 , which, beginning below with the thin Zone of Nummulites elegans, in green sandy clay, takes in a set of loamy sands with Voluta athleta (Sol.) and Cassis ambigua (Sol.), and ends with rusty sand con- taining abundant Pholadomya margaritacea J. Sow. This stage is the richest in species of mollusca, a large proportion of them 1 Quart. J mm. Oeol. Soc., vol. xliv, 1888, pp. 578-633. 2 Sometimes (as in Fig. 8) termed " Highcliff Beds," and even " Highcliff Sands," the latter name being applied also to a division of the Lower Barton (Pig. 8), as well as to the highest local stage of the Bracklesham Beds. It is in the last sense alone that " Highcliff" should be used as a stratigraphical term. 42 GEOLOGY OF BOURNEMOUTH. Fig. 8. — Profile-section of the Barton Beds, Barton Cliff. (Adapted from Gardner, Keeping, and Monckton, Quart. Journ. Geol. Soc, vol. xliv, 1888, opp. p. 594). Voluta solandri, V.scabricula, V.scalaris V.subambigua.V.humerosa. 10 dark olive sandy clay < Batillaria pleurotomoides f etc.J Long Mead End Beds. White &.tawny sandsrp. * Brackish - Ytater fossils j Oliva branderi Zone , Earthy sane Becton full of casts t Bunny Beds. Whitish sand unfossiliferousr Voluta luctatr'n.V. ambigua. Septaria Drab clay Best fossil zone Voluta athleta, V.nodosa. Drab clay Drab sandy clay Septaria Voluta suspensa Zone 'holadomya Bed Highcliff f vi;; , sands ' Drab clay with sand drifts __ T Red band 'Stiff drab clay w ' Nummulina Zone [Nummulites eleians] Ironstone Band 33 Green sandy clay tin Pebble bed •V Whitish sand Vi-Line of pebbles BARTON BEDS. 43 being of minute size, and occurring in sandy pockets. It has been remarked that many of the shells here resemble living forms from Australia and Japan, and seem to indicate a considerable depth of water. 1 Small corals (Turbinoliae), echinoderms (Ophiura wether elli Forbes, etc.), claws of crabs, teeth of fish (Arius, Myliobates, etc.), turtle-bones, a few worn freshwater shells, and drift-wood, are among the other fossil remains present in these beds. The upper half of the Barton Clay (i.e., the Middle Barton Stage) is more decidedly argillaceous, and yields most of the large and handsome shells so noticeable in collections of Barton fossils. On the coast, and for some distance inland, this stage is characterised by the presence of large concretions of cement- stone (often septarian), which are widely spaced in a few layers- Two layers of these concretions, here of rounded form, occur just above the base of this stage near Chewton Bunny. Between two higher layers, which are farther apart, and in which the concretions are of paler tint and of tabular shape, there are two exceptionally rich fossil-bands, in the Zone of Voluta luctatrix. The top of the Barton Clay is marked by a reddish bed, full of broken shells, hardened slabs of which occur on the beach at the foot of Barton Cliff. The Barton Clay has yielded remains of a rare cetacean, Zeuglodon wanltlyni Seeley, some bones of which, together with a large collection of Barton mollusca, are in the possession of the Bournemouth Natural Science Society. At the close of the Middle Barton age obscure geographical changes diverted the mud-bearing currents responsible for the Barton Clay to other areas of deposition, and brought in drifts of sand. The Barton sea over the Hampshire region appears to have become shallower as well as clearer, and the concurrent changes that took place in the fauna were such as to suggest a response to altered physical conditions, rather than a great lapse of time between the deposition of Middle and the Upper stage of the Barton Series. The Barton Sands, or UjDper Barton, comprise about 90 to 100 feet of light-coloured, fine-grained sands, with some beds of a loamy and clayey character, which are developed chiefly at the base and near the top of this stage. In the bluish-grey loamy sand (Chama Bed) immediately above the Barton Clay, Chama squamosa Sol. occurs in profusion, often with the valves united. 2 Above come the Becton Bunny Beds — light, unfossiliferous sands overlain by earthy sands, and these by dark sandy clay with Oliva branderi J. Sow., etc., followed by sandy loam containing brackish - water shells (Cyrena, Dreissensia, Erodona, etc.), Lastly come the white and yellow sands of the Long Mead End Beds, with Lucina gibbosula Lam., Batillaria pleurotomoides Lam., etc., and a thin band of greenish clay at the junction with the Headon Beds. In the upper parts of the Barton Sands there is clear evidence of shoaling water, and an approach to the con- ditions that were to prevail in Lower Headon times. 1 Gardner, Keeping and Moncktrc, op. cit., p. 261. 2 Between Barton Court and Becton Bunny. In the area. of the Bournemouth map the fossils have been removed by weathering. 44 GEOLOGY OF BOURNEMOUTH. Inland Exposures. During the construction of the railway between Christchurch and Brockenhurst a good section in the JBarton Clay and Sands was opened in the cutting east of Hinton Admiral Station, a little more than a mile north of the coast- Messrs. Gardner, Keeping, and Monckton 1 observe that the shell-beds at the top of the Barton Clay are considerably thicker than in the. cliff section, and that the Chama Bed is unfossiliferous. If, as these authors think is probable, the greenish clay in the western part of the cutting belongs to the Lower Barton, the thickness of the Middle Barton stage is here reduced to a little over 30 feet, com- pared with 53 feet on the coast. Farther north, dark greenish and grey sandy elaj^s with iron- stone nodules, in the lower part of the Barton Clay, are exposed in brickyards at Bransgore and Crow ; and bluish sandy clay of the higher beds of this formation is worked in a brickyard half a mile north-east of the former village. The Barton Sands are dug in many places, but the sections afforded are of no particular interest. Near the surface of the ground the fossil shells have everywhere been dissolved away by percolating rain-water, leaving at best only indistinct impres- sions in iron oxide. Exposures can be seen on the western side of Poors Common near Bransgore ; beneath the Plateau Gravel on either side of the road from that village to Thorny Hill ; north-east of Charles Farm near Crow, and in the railway cut- tings through Cranes Moor. 1 Quart. Journ. Owl. Soc, vol xliv, 1888, p. 597, Fig. 5. GEOLOGY OF BOURNEMOUTH. 45 CHAPTER VIII. HEADON BEDS. The strata of Tertiary Age that remain to be considered belong to the Lower division of the Headon Beds, so well displayed at Hordle Cliff, east of Barton (Sheet 330). In the area of the Bournemouth map these beds do not appear on the coast, but come in a few miles north of it, in the rising ground of Holmsley Walk, and the ridge of Burley Beacon and Castle Hill, farther inland. At Hordle Cliff the Lower Headon Series is made up of inter- stratified green marls and clays, and loamy sands, with sub- ordinate bands of ironstone, limestone, and lignite. The physio- graphic conditions indicated are such as obtain in delta-lagoons with restricted or occasional access to the sea, while the asso- ciated organic remains suggest a climate like that of the coastal region of North Carolina at the present day. Seed-vessels of aquatic plauts (Chara, etc.), and debris of land vegetation, occur in. profusion at certain horizons. Shells of fresh- and brackish- water molluscs, such as Erodona plana (J. Sow.), Dreissensia hrardi Fauj., Melanin acuta (J. Sow.), Cancelliria miryata Edw. MS., Limnaea longiscata Brong., and Yiriparux lentus"(,J . Sow.), also abound; a Unio {U. solandri J. de C. Sow.) characterising beds near the junction with the Middle Headon Series. The lower beds, especially, contain a remarkable assemblage of ver- tebrate remains, referable to alligator (Diplocynodon), terrapin (Ocadia), tortoise (Trionyjc), several genera of birds (Actiornis, Agnopterus, Grus, etc.), and of mammals (Antliracotheriwii , Viverra, Hyaenodon, Adapis, etc.). A fuller account of this section, with references to the literature, will be found in the memoir on the Lymington map-sheet (No. 330). x Excepting a dark lignitic clay or loam at the base, none of the well-marked horizons in the Hordle coast-section has been identified in the area covered by the present memoir, though it is not unlikely that close observation of the local pits would reveal the presence of the chief bone-bearing strata, known as the Croco- dile and Mammalia Beds. Near Hordle the Lower Headon Series is estimated to be about 80 feet thick. As it is better developed there than in the country farther east, one may assume that it was originally at least as thick in the adjacent part of the 1 ' Geology of the Country near Lymington and Portsmouth,' Mem. Oeol. Surv., 1915, pp. 37-41. 46 GEOLOGY OF BOURNEMOUTH. Bournemouth district, to the west, where now there seems hardly room at any spot for as much as 60 feet of beds between the Barton Sands and the Plateau Gravel. Probably the highest members of the series have been worn away, together with the marine Middle Headon, no indication of which has been observed. Notes of Exposures. On the west side of Poors Common, near Bransgore, a pit shows about 2 feet of the carbonaceous loam which almost in- variably forms the basal member of the Headon Beds, resting on evenly-stratified sandy loam of the Barton Sands. Bedded greenish clays in the lower part of the series have been dug in a brick-pit a quarter of a mile east of Purlieu, and in another yard at Thorny Hill. At the latter spot the clay, containing seams of Erodona and Cyrena, is succeeded by unfossiliferous brown sands and sandy clays, which are exposed for a thickness of about 20 feet in the higher part of the brickyard. On the northern slope of Thorny Hill a shallow pit by the side of the road to Burley is opened in green marl with bands of shells, mostly broken. The fossils present include Erodona plana (J. Sow.), Melanopsis fusiformis (J. Sow.), Melania acuta (J. Sow.), Viviparus lentus (J. Sow.), and- chips of bone. Other small exposures of shelly marl with Melania acuta were noted by Mr. C. Reid in the Valley of Walkford Brook at Plain Heath, and in the watercourses in Holmsley Inclosure. No sections were seen in the wooded outlier of Castle Hill, west of Burley. The ground is clayey, and marl or clay for land-dressing has been dug in an overgrown pit near the southern end of the ridge. -GEOLOGY OF BOURNEMOUTH. 47 CHAPTEE IX. PLATEAU GRAVEL. The sedimentary record is broken off with the Lower Headon Beds, but, having regard to the fortunate preservation of a relic of the Bembridge Limestone at Creech Barrow, about seven miles south-west of Poole, and to the marine character of the Upper Hamstead Beds in the Isle of "Wight, there can be little doubt that the Eocene formations of the Bournemouth district were formerly covered by a considerable thickness of Oligocene strata of more recent date than any now surviving there. Concerning the dark ages of the Miocene and Pliocene, there is not much to be said, for the records, as far as they are pre- served, are but vaguely expressed in terms of erosion and tectonic structure. In all probability the east-and-west undulations of the Tertiary Beds near Bournemouth are an outcome of the same stresses that give rise to the parallel, but stronger, folds which traverse the Isle of Purbeck and the Isle of Wight, and are in- ferred to be of Miocene date. It was during this period that the intermittent subsidence which had been in progress throughout the Eocene and Oligocene was definitively checked, and replaced by a movement in the opposite direction. The latter movement, which involved the sub-aerial exposure and destruction of strata previously laid down, seems to have given way to subsidence in late Miocene and early Pliocene (Lenhamian) times, but at the epoch of the Plateau Gravel, with which the stratigraphical history of the district reopens, the dominant movement was again one of elevation. Near Bournemouth the gravels of the Plateau group are well developed, and distributed with so little regard to the boundaries of the solid formations below as to make it evident that by far the greater part of the denudation suffered by the Tertiary Beds, since Lower Headon times, had been accomplished before any of the gravels were formed. Such superficial undulations as the gentle Miocene flexures gave rise to had almost or quite dis- appeared, and the edges of the tilted Tertiary strata had been bevelled off to a moderately even surface with a prevailing in- clination to the south-east. It is not here implied that the local Plateau Gravels rest on a definite plain of erosion : a short inspection of the map will show that the deposits in question have a fairly wide range in altitude within restricted areas, and that while some of them are isolated on the tops of the ridges along the water partings, others more numerous descend into the larger valleys, there to form sloping sheets and terraces down to within 60 feet of local river-level. Clearly, the deposits differ 4,8 GEOLOGY OF BOURNEMOUTH. much in age, though most, if not all, come within the limits of the Pleistocene. They are not known to contain contemporary organic remains, but have yielded many Palaeolithic implements, referable to the Chellean and Acheulian culture-stages, chiefly in localities near the coast. In section, the gravels normally display the structural features indicative of deposition in running water. Subangular flints and quartz-sand are their main constituents, and with these are associated Tertiary flint-pebbles, ironstone, and sandstone (sarsen), together with much c'herty sandstone from the Upper G-reensand, and pebbles of a variety of other rocks, such as vein- quartz, silicified Purbeck limestone, and Palaeozoic grits and quartzites — these, with part of the Greensand chert, probably being derived immediately from Eocene gravels of the kind developed near Dorchester. As far as the writer has seen, the local Plateau Gravels include no deposits that could be reasonably regarded as of marine origin. Their structure, composition, and distribution jointly proclaim them to be the coarse alluvia of streams which gathered the greater part of their burden of rock-waste in the country drained by the existing Avon, Stour, and Frome, whose lower courses lie within the district. Many of the gravel-spreads may be con- fidently ascribed to the former activities of one or other of these three rivers, but the parentage of the higher deposits, especially, is often in doubt. It would appear that in the early ages of the Plateau Gravel epoch the main lines of drainage in southern Hampshire and eastern Dorset were less strictly defined than they subsequently became, and that the various streams from the enclosing uplands of the Chalk and older rocks mingled their waters on extensive flats that had been opened out in the soft Tertiary strata near the axial line of the Hampshire Basin. At that time, and for a long while after, the drainage of the western part of the basin was carried off eastward by what has been termed the 'Solent River,' a continuation of the Frome; and it probably was not until after the Bournemouth Plateau Gravels had been deposited that the sea encroached far enough upon the area between the Isle of Purbeck and the Isle of Wight to break up this old drainage-system. 1 The contrast between the Plateau Gravel and the fine-grained modern river-deposits at lower levels is so striking that it might well appear to constitute a serious objection to the fluvial theory in respect to the former. The prevailingly coarse texture of the Plateau Gravel, however, is readily accounted for under the assumption that the drainage channels in former times frequently carried a much greater volume of water than they ever have to 1 Mr. Clement Eeid long entertained the idea that the effective breaching of the southern side of the Solent Valley by the sea took place in Pliocene times, but, latterly, a reconsideration of the problem led him to think — with the writer and Mr. Henry Bury — that this event occurred more recently, and probably towards the close of the Pleistocene period. Cf. C. Reid, ' Country around Ringwood,' Mem. (leal. Sum., 1902, pp. 31, 32 ; and ' Ancient Rivers of Bournemouth,' Bournemouth Nat. Sci. Soc, 1910. Also H. J. 0. White, 'Country near Lymington, &c.,' Mem. Geol. Surv., 1915, pp. 48, 49; H. Bury, 'Physical Geology of Bournemouth,' Proc. Geol. Soc, No. 984, 1916, p. 22. PLATEAU GRAVEL. 49 cope with at the present day; and that assumption is justified by what is known of the climatic conditions obtaining in the South of England during the Pleistocene period. Indisputable evidences of arctic temperatures have yet to be found in the Bournemouth district. Boulders like those of the Selsey pen- insula, which imply the former presence of floating ice, are wanting in the local deposits, and of the Selsey ' erratic gravel ' no representatives are likely to have been preserved, except, perhaps, in the form of scattered blocks on the sea-bed; for the westward continuation of the ancient foreshore upon which the Selsey erratics were stranded must have lain some distance sea- ward of the existing cliffs on the Bournemouth coast. The only phenomena suggestive of ice-action are those of contortion and trail, which are often to be seen in the superficial parts of the gravels, at various altitudes. Distribution and Exposures. Passing now to a more particular consideration of the deposits comprised in the Plateau group, it will be convenient to begin with the developments in the eastern part of the district. (a.) Gravels near and east of the Avon. — In his memoir on the country around Ringwood (Sheet 314), Mr. Clement Reid describes four stages of gravel-covered terraces and plateaux, on the eastern side of the Avon, which he distinguishes as : — 1, Palaeolithic Terrace (100 feet) ; 2, Eolithic Terrace (150 feet) ; 3, Highest Terrace (200 feet) ; and 4, High Plateau (300 feet) : the measurements in parentheses denoting approximate altitudes above the level of the river. Of these, the High Plateau stage does not extend into the area of the Bournemouth map. The Highest Terrace, next below it, is represented in the top of Castle Hill, (fig. 9), above 300 feet O.D., which is capped by gravel containing a good deal o'f Greensand chert ; and, more doubtfully, by Dur Hill Down and Whitefleld and Thorny Hills (all between 240 and 250 feet O.D.) farther south. East of the above, and partly outside the Avon basin, a group of gravel-flats, rather over 200 feet O.D., and including Durmast Hill, Holmsley Ridge, and Plain Heath, probably corresponds to Mr. Reid's ' Eolithic Terrace.' The gravel is exposed in pits and road-banks on the borders of the Durmast Hill plateau, and at Burley it is partly cemented by iron pan into masses of con- glomerate, called 'Burley Rock.' On Plain Heath there are small sections east of Hill Farm, and east of the Southampton road near the point marked 208 on the map. Eoliths are not known to have been found in these deposits, nor, for that matter, in many of the so-called Eolithic gravels of the Ringwood dis- trict either. South of Plain Heath (210 feet O.D.) a continuous spread of gravel slopes down to the top of the cliffs near Newtown and Barton (110 feet), at an average inclination of about 30 feet per mile, but in a belt of ground roughly indicated by a line drawn from Harrow Lodge through Hinton Park to Beckley Farm a slight increase in the gradient, between 170 and 150 fp -c o OO s ^ s 53 & > w .„ and J. R. Dakyns. 2nd Ed. By A. H. Grbbit and A. Strahan. Bs. Bd. < FALMOUTH AND TRUBO AND MINING DISTRICT OF CAMBORNE AND BEDRUTH. By J. B. Him. and D. A. MAOALISTBR. 7s. 6d. - FENLAND. By S. B. J. SKBRTOHLY. 36s. Bd. HOBDEBNESS. By 0. BEID. 4s. ISLE OF MAN. By G. W. LAMPT.TJOH. 12s. TERTIARY FLUVIO-MARINE FOEMATION of the ISLE OP "WIGHT. By EDWARD FORBES. 6s. ISLE OF WIGHT. By H. W. BRIST0W. New Ed. By 0. BEID and A. STRAHAN. 8s. Bd. GUIDE' TO GEOLOGICAL MODEL OF INGLEBOROUGH DISTBICr. By A. Straban.. id. ISLE OP-PURBEOK AND WEYMOUTH. By A. STRAHANi V)s. Bd. GUIDE TO GEOLOGICAL MODEL' OP ISLE OF PURBECK. By A. STRAHAN. Bd. KENT, ON THE MESOZOIO BOOKS INCOME OF THE COAL EXPLORATIONS IN. By G. W. Lampluoh and P. L. Kitchis. 3s. 6d. LAKE DISTRICT, NORTHERN PART OP. By J. 0. WARD. 9s. LANCASHIRE, SUPERFICIAL DEPOSITS OF SOUTH WEST. By C. JB. DE RANCE. 10s. Bd. LONDON AND PART OP THE THAMES VALLEY. By W. WHITAKER. Vol. I., 6s. Vol. II., 5s. . LYMINGTON AND PORTSMOUTH. By H. J. OSBORNE WHITE. In Bd. MIDLAND COUNTIES, TRIASSIC and PERMIAN BOCKS of the. By.E. Htjll. 5s. NORFOLK and: SUFFOLK, VERTEBRATA of the FOREST BED SERIES of. By E. T. Newton. „ 7s. Bd. NORWICH. By H. B. WOODWARD. Is. RUTLAND, &e. By J. N W. JUDD. 12s. 63. Iron Ores of Great Britain : — Part II. SOUTH STAFFORDSHIRE. Price 1». Part III. SOUTH WALlSS. Price Is. 3d. Coal Fields : — YORKSHIRE COALFIELD. By A. H. GREEN, B. RUSSELL (and Others). 42» YORKSHIRE and NOTTINGHAMSHIRE, CONCEALED COALFIELD of. By W GIBSON la Bd EAST SOMERSET and BRISTOL COALFIELDS. By H. B. WOODWARD. 18s. ' " WARWICKSHIRE COALFIELD. By H. H^ HOWELL. Is. Bd. LEICESTERSHIRE and S0U3SH DEBBYSHIRE COALFIELD. By C. FOX-STHANGWAYS Bb NORTH STAFFORDSHIRE SOALFIELDS. By W. GIBSON (and Others). 6«;- SOUTH WALES, Coals of. B§".*A. Strahan and W. Pollard. 2nd Ed. 2s. SOUTH WALES. See under New Series Maps and Memoirs. Sanitation and Water Supply: — SOILS AND SUBSOILS from a SANITABY POINT OP VIEW. By H-B. Woodward. 2nd Ed Is 6d NOTES on SOURCES of, TEMPORARY WATER SUPPLY in the SOUTH of ENGLAND and NEiGHBOURHSIG PARTS of the CONTINENT. 2d. ji«™uwujib« WATER SUPPLY OF BEDFORDSHIRE and NORTHAMPTONSHIRE. By H. B. WOODWARD and B THOMPSON, is. Bd. WATER SUPPLY OF BERKSHIRE. By J. H. BLAKE. 3s. WATER SUPPLY OP HAMPSHIRE. By W. WHITAKER. Bs. WATEB SUPPLY OP KENT. By W. WHITAKBR. 8s. Bd. WATER SUPPLY OF LINCOLNSHIRE. By H. B. WOODWARD. 4s. 6(7. WATER SUPPLY OF NOTTINGHAMSHIRE. By G. W. Lampluoh and B. Smith. 5s, WATER SUPPLY Of OXFORDSHIRE. By R. H. TlDDEMAN. 2s. 3d. WATER SUPPLY.OF SUFFOLK. By W. "WHITAKER. Ss. Bd. WATER SUPPLY OF SURREY. By W. WHITAKER. 7s. WATER SUPPLY OF SUSSEX. By W. WHITAKBR and O. RBID. Ss. . WATER SUPPLY OF SUSSEX (Supplement). By W. WjIITAKER. 2s. Bd. WATER SUPPLY OF EAST RIDING OF YORKSHIRE. By C. FOX-STRANGWAYS. 3s RECORDS OF LONDON WELLS. By G. Barrow and L. J. Wills. 4s. Bd. Museum Catalogues: — HANDBOOK TO BRITISH MINERALS,- By F. W. Rcdler. Is., and Others. GUIDE TO THE COLLECTION OF GEMSTONES IN THE MUSEUM OF PRACTICAL GEOLOGY Bv | • W. F. P. McLintock. 9d. ' HANDBOOK TO THE COLLECTION OF KAOLIN, CHINA-CLAY AND CHINA-STONE IN THE MUSEUM OF PRACTICAL GEOLOGY, By J. ALLEN Howe. 3s. id. m«mu»l,UM Special Reports on the Mineral Resources of Great Britain :— Vol. 1. TUNG'STEN AND MANGANESE ORES. ' 2nd Ed. Is. VOL. 2. BARYTES AND WITHERITE. 2nd Ed. 2s. VOL. S. GYPSUM, ANHYDRITE, CELESTINE AND STRONTIANITE. Is Vol. 4. FLUORSPARS 2nd Ed, ' 9d. Vol. 5. POTASH-FELSPAR, PHOSPHATE OF LIME, ALUM SHALES, PLUMBAGO OR GRAPTTTTF MOLYBDENITE, CHROMITE, TALC AND STEATITE (S0AP8T0NE, SOAP-ROOK AND PCTtrTOVPi DIATOMITE. 2nd Ed. Is. * fOiaiONH), v;'^ "ii ; ill Pi -Sill iHiniii ,:,;:;.:. iiifii .:■:■■■! ■■iiiiif. v:i;)!,!;;:'.:;f;(S(:?; ^ Oi ■ !i.i!ii(".-::.)!.)i-t! JiiliisiliiiliiP 'Slii ■■i !i ;# ■ . ill I ; Siiiiiliiiii ilii;!