NORTHWESTERN UNIVERSITY
LIBRARY
EVANSTON, ILLINOIS
LLAIíDAFF CATHEDEAL.
Htris Fl Alios
■ íi;:'ií)'Kíi)i5íHsii ivsritaIIII.ILW.D ..î.îV(a.MV'ii"i;'
E E M A R K s
ON THE
ARCHITECTURE
OF
LLANDAFF CATHEDEAL;
WITH
A N ESSAY
TOWARDS A
HISTORY OF THE FABRIC.
Br
EDWAED A. FEEEMAN, M.A.,
pt» •>
LATE FELLOW OF TRINITY COLLEGE, OXFORD, AUTHOR OF THE
" HISTORY OF ARCHITECTURE."
LONDON :
W. PICKERING, 177, PICCADILLY.
TENBY: R. MASON.
1860.
TENBY: R. MASON, PRINTER, HIGH STREET,
TO THE
VERY REV. W. D. COISTYBEARE, M.A.,
DEAN OF LLANDAFF^
&C.J &C.J
THIS NOTICE OF THE CHURCH
OVER WHOSE RESTORATION HE SO WORTHILY PRESIDES
IS INSCRIBED^
IN MEMORY OF THE KINDNESS AND HOSPITALITY
RECEIVED FROM HIM
DURING THE RESEARCHES
WHICH LED TO ITS COMPOSITION.
CONTENTS.
Chapter I.—-General Description of the Church
§ 1. Outline and Ground-plan .
Comparison with other Buildings .
§ 2. The West Front ....
West End of Nave ....
South Tower .....
North Tov/er .....
§ 3'. Nave and Choir
Present Appearance ....
Original Design.—Exterior
Interior ......
Division of Nave and Choir
Arcades ......
Roof ......
§ 4, Presbytery^ Chapter-House^ and Lady
Chapel ......
Exterior of Presbvterv
€/ •'
Interior of Presbytery ....
Lady Chapel .
Chapier-Housc .....
o 6 «
VIH
CONTENTS.
Chapter II.—History of tlie Fabric . . . . .42
§ 1. The Romanesque Church . . . . 44
Biitish Period , . . . . . ib.
Extent of the Norman Church . . . 48
Tower-Porch ? . . . . . .49
Supposed Clerestory . . . . . 53
§ 2. Early English Additions . . . .61
Date of the Nave . . . . . ib.
External Walls . . . . .64
Eastern Towers . . . . . . 66
South Aisle of Presbytery . . .69
Chapter-House , . . . . . 71
§ 3. Decorated Repairs . ... . . ib.
The Lady Chapel . . . . . ib.
Aisles ....... 72
Aisles of Presbytery . . . . . 73
Presbytery . . . ' . . .74
Aisles of Choir and Nave . . . . 75
Perpendicular.—The North-west Tower . 77
§ 4. Decay and Restoration of the Cathedral . 79
LIST OF ILLÜSTEATIONS.
General View . . . .
Ground-plan
West Front
West Doorway ....
Doorways in Nave
Interior of West End
Bay of Nave . . . .
Bay of Clioir ....
Arcli into Lady Chapel .
Details of Lady Chapel
Arch and Window in Presbytery .
facing Titlejpage
'page
1
99
13
99
14
99
99
22
99
99
25
99
99
27
99
99
28
99
99
36
99
99
38
99
99
50
PEEFACE.
The present volume lias grown up in a manner which I
may venture to call analogous to the history of the
building of which it treats. Happening to be present
at the Cardiff Meeting of the Cambrian Archaeological
Association in 1849; when Llandaff Cathedral naturally
formed one of the most prominent objects of attention,
I was induced to make some remarks at one of the
evening meetings on some of the more singular pecu¬
liarities of the fabric. This was after only a very cursory
examination of the building, and was as much to point
out a few of the many difficulties connected with, as to
offer any solution of them. My casual speech next de¬
veloped into a paper for the Journal of the Association,
the Archœologia Cambrensis, This stage required a more
minute and diligent investigation of the church, which I
had the pleasure of performing in the presence and with
the aid of Mr. Prichard, the architect of the restoration.
xii
PREFACE.
The result of our inquiries was to bring to light much
that explained previously existing difficulties, much also
that did little more than start new ones, to some of
which I fear I have not yet found the key. The subject
growing upon me, as the speech developed into a paper,
so the paper developed into a book ; and a casual visitor
to Llandaff has gradually found himself in a position
only too like that of the historian of its Cathedral.
My paper, as communicated to the Archœologia
Camhrensis^ was at once imperfect, and to a certain
extent controversial. It consisted chiefly of remarks on
points which had not been noticed, or on which I found
myself differing from other speakers at the Meeting.
The form which it assumed, as was natural under the
circumstances, was that of remarks, chiefly on the
history of the fabric, addressed to persons already
acquainted with its appearance ; it consequently con¬
tained no direct description of the building at all. But
when a separate publication of the paper was thought of,
in common with several others which had appeared in
the same Journal, it struck me that an opportunity had
occurred for supplying a desideratum in architectural
literature. Instead of a mere reprint of a magazine-
paper, I thought, if anything on Llandaff Cathedral
were published at all, it should be something that might
PREFACE
• • •
Xlll
make some pretensions to the character of a descriptive
and historical account of the building, a work which, as
far as I am aware, has not previously been attempted in
our times. A second sojourn at Llandaff, undertaken for
the purpose, has enabled me to produce, in the present
volume, an attempt to supply the deficiency. I have
worked in all the critical and historical remarks con¬
tained in my paper in the Journal, which form a great
part of the second Chapter ; but much the greater
portion of the work, and all the illustrations but one,
now appear for the first time.
I have not however ventured to call my conjectures,
for they are often little more, as to the several dates of
the building by the ambitious name of an "Architectural
History." Besides an unwillingness to enter so directly
into competition with the elaborate productions of Pro¬
fessor Willis, it of itself hardly merits the title, as it is
not the fruit of any documentary inquiries at all. I was
informed by the Dean that little or nothing was con¬
tained in the Cathedral archives at all bearing on the
history of the fabric. And I have had the less scruple
in making my own work exclusively architectural, as all
documentary and topographical questions cannot be left
in better hands than those of the accomplished antiquary
who now fills the office of Chancellor of the Church of
xiv
PREFACE.
LlandaíF, and from whom we may perhaps jet hope to
see a complete History of Glamorganshire.
I may however mention that I am at present engaged,
in conjunction with my friend the Rev. W. B. Jones, on
a far more extensive portion of the highly important and
comparatively unexamined antiquities of South Wales;
namely, a complete History of St. David's. Here we
hope not only to describe and conjecture, but to give a
truly historical account, grounded on documentary evi¬
dence, of the Cathedral and its appurtenances, as well as
of the Bishopric, Chapter, and City. And I cannot but
recommend the wide field of Welsh antiquities to all
whose tastes lead them to such pursuits. Even in the
respect in which the Principality is generally thought
most deficient, remains of ecclesiastical architecture, its
*
southern counties—for of North Wales I can say but little
from personal examination—contain very much more
than is often suspected. Besides a few really grand
structures scattered here and there, among which Brecon
Priory may almost dispute the second place, with Llan-
daif, the village churches of Pembrokeshire and Glamor¬
ganshire will afford no small amount of interest to any
who can dispense with the presence of much actual
detail. The immense importance of South Wales to the
military antiquary it is superfluous to mention.
PREFACE.
XV
In conclusion, I have to return my best thanks to the
Dean of LlandafF, the Rev. Chancellor Traherne, and the
Rev. Richard Prichard, B.D., Senior Vicai^, for much
courtesy and hospitality received during my researches
at LlandafF. I have peculiar pleasure in connecting my
work more immediately with the distinguished name of
the Dean, on account of the numerous points connected
with the history of the church, in which I have the mis¬
fortune to entertain a view different from his. And I
have still more special obligations to acknowledge to
John Prichard, Esq., the architect of the restoration, for
the zeal and interest with which he entered into all my
investigations, and for the extreme kindness and libe¬
rality displayed by him in furnishing gratuitously the
original drawings of all the wood-engravings.
OAKLAÑDS, DURSLEY,
June 1, 1850.
LLANDAFE CATHEDRAL.
= Parpendicular
A Lady Chapel
B Chapter-house
0. Signs of destroyed
arches
DD Towers.
GROUND-PL A.N,
J Prichard, del
O. Jewitt, sc.
ïliiïïîiaff Ctttijfîtml.
CHAPTER L
GENERAL DESCRIPTION CE THE CHURCH.
Llandaff Cathedral is a building which, in many
respects, both of its history and architecture, stands
quite alone among English churches. To one who had
never visited St. David's, it would probably appear, in
all its circumstances, nearly the most striking of their
number. Elsewhere we are accustomed to find our
greater churches, those especially of cathedral rank,
sometimes in the densest parts of our great cities, but at
all events in towns of considerable size, rising as a
witness above the din and bustle of busy life. But the
founders of the Welsh cathedrals would seem, as indeed
is directly recorded of the greatest among them, almost
to have fled from the presence of men, and to have fixed
their dwellings in places adapted rather for retired
contemplation than for any active government of the
church, in sites suited rather for Cistercian abbeys than
for cathedral churches. The English sees have been in
B
2
LLANDAFF CATHEDRAL
several instances, after their original seats had fallen
into decay, removed to cities of greater importance ;
while the Welsh remain to this day in the small places
where they were originally founded. For there is no
reason to suppose that the " cities" of St. David's and
LlandaiF were at any time of greater intrinsic importance
than at present. A somewhat greater amount of popu¬
lation and industry was doubtless induced in past ages
by the presence of a greater number of resident ecclesi¬
astics; but there could hardly have been anything
beyond what was thus factitiously introduced. No
military or commercial consequence ever belonged to
them.
This air of desolation and remoteness from man is
however much less strongly felt at Llandaif than at St.
David's. The greater size of the latter cathedral, the
immense extent of surrounding ruins, the character of
the country around, produce an effect infinitely more
striking than that of LlandaiF. The richer country
around the latter, and the near neighbourhood of a large
town, take away much from its solitary character ; and'
the peculiar outline of the building helps to diminish
the effect. St. David's is eminently a cathedral in the
wilderness : Llandaff might almost pass for a village
church of unparalleled size. With no cathedral character
in any part of its exterior except the west front, with all
traces of collegiate buildings demolished, there is nothing
whatever to mark its peculiar purpose ; while the un¬
paralleled neglect which, till lately, had overwhelmed
alike the fabric and its services, has reduced the whole
LLANDAFF CATHEDRAL.
3
to a state into which St. David's, with all its deficiencies,
has never fallen.
In fact, the nearest parallel to Llandaif which I know,
is a building not of cathedral rank at all, at least not for
nearly eight centuries ; the Abbey Church of Dorchester
in Oxfordshire. They are the two greatest village
churches, existing and used as such, that I have seen ;
and several points of resemblance may be found both in
their architecture and their history. Both exhibit the
same vast length, unbroken by tower or transept, so
unusual in churches of their scale and rank. In both
an originally small church, by enlargement in different
directions, has swelled into a vast pile, but without
acquiring, either within or without, many of the dis¬
tinctive features of a large church.^ Llandaif, however,
has acquired far more of those distinctive features. The
individual parts, the nave, choir, &c., are quite cathedral
or abbatial in their character, though the general effect
is not ; at Dorchester even the parts taken singly exhibit
only an exaggeration of the parochial type. Both—
though Llandaif more completely and more disgracefully
—have fallen into utter neglect and decay; both have
come in for their share of the happy spirit of restoration
which does so much honour to our own day.
Llandaif Cathedral on the whole stands well. The
finest part of the building, the beautiful west front, is
1 At the present moment the resemblance between Llandaif and
Dorchester in a distant view is, from an incidental cause, extremely
striking. The towers rise pretty much the same height above the
main building, and the small portion of roof raised at the east end
produces a most singular eifect in both cases.
4
llandaff cathedral.
indeed concealed from any distant view ; but this defici¬
ency is quite counterbalanced by the singular and strik¬
ing approach from the " city the steep descent coming
down almost immediately upon the grand western portal.
The rest of the church stands open, and veiy good views
from the south, both nearer and at some distance, may
be obtained in a walk across the fields from Cardiff. I
know of no cathedral, from which the subordinate build¬
ings have been so completely removed, in which their
loss is so little felt ; probably because, as the character
of the building does not so distinctly proclaim its rank,
the deficiency is not so painfully forced upon the eye.
The castellated gateway of the old episcopal palace is a
fine object, and comes well into the grouping from
several points ; but it has no particular reference to
the cathedral. The palaces at Wells and St. David's,
especially the latter, magnificent in its ruins, could be
mistaken for nothing but what they are ; they are parts
of a whole, the largest and most splendid portion of the
collegiate buildings ; but the remains at Llandaff have
nothing distinctively episcopal about them ; they might
as well have been the stronghold of any Norman robber,
the lair of the wolf of the flock, rather than the dwelling
of its shepherd.
§ I.—Outline and Ground Plan.
The first aspect of the cathedral is not a little perplex¬
ing, and it requires considerable familiarity with the
building both within and without, fully to grasp the
LLANDAFF CATHEDRAL.
5
principle of its arrangement, and to recognize its com¬
ponent parts. Looking down on the church from the
rising ground to the south, the best point for obtaining a
view of its whole extent, the aspect is confusing indeed ;
the appearance of the building resembles a perfect chaos.
A deserted ruin at the extreme west ; the eastern por¬
tions fresh from recent restoration ; the centre reduced
to the likeness of a conventicle or a third-rate town-hall—
perhaps no more incongruous assemhlance of discordant
elements was ever brought together.
But it is not merely to these accidental circumstances
that the difficulties alluded to are owing. A thorough
restoration would diminish, but it would by no means
entirely remove them ; they are inherent in the design
of the fabric. Its ground-plan, outline, and arrange¬
ment are altogether unique. It consists—speaking of
the appearance which it presented when complete, and
which we may fairly hope it will, before many years,
present again—of a long unbroken body, comprising
under an uninterrupted roof, nave, choir and presbytery,
with a large Lady chapel projecting from the east end at
a somewhat lower elevation. Aisles extend along the
whole length of the main body and along one bay of
the Lady chapel ; the west end is flanked by low towers
terminating the aisles ; a square building, forming the
chapter-house, projects from the south aisle of the pres¬
bytery, having somewhat the air of a low transept.
All this is widely different from the oi'dinary design
of an English cathedral. The first and most marked
peculiarity is the absence, in a church of so great a size.
6
LLANDAFF CATHEDRAL
not only of a central tower, the usual crown of our large
churches, but of transepts in any form. In this respect
it is unique among the cathedrals of South Britain, and
has but few parallels among churches of equal size, even
when not designed as episcopal sees, as the present Ca¬
thedral of Manchester, Dorchester Abbey, Boston, and
St. Michael's, Coventry. And even among these, the
distinction of the several parts of the church is generally
more strongly marked than at Llandaif, where there is no
constructive difference whatever between nave and choir,
the only perceptible external change in the main body
of the fabric being between the choir and the presbytery,
and that consisting only in the different arrangements of
the clerestory.
The plan and arrangement of the church is alto¬
gether singular ; there can be no doubt but that the
constructive nave included both the true nave and the
ritual choir, the only architectural mark of distinction
being a slight change of detail, and that confined, as far
as we can judge, to the interior. What would ordinarily
be called the choir is really the presbytery. This is not
very uncommon ; the grand peculiarity is the absence of
transepts, which usually divide either the nave from the
choir,^ as in most of our large churches, or else, as at
Westminster, the choir from the presbytery ; or again
(where there are two pair of transepts) discharge both
functions in the same building. Thus, from the extreme
west to the east end of the presbytery, the only break of
1 With these we must reckon architecturally the churches where
the ritual choir is beneath a central tower.
llandaff cathedral.
7
any importance—there not being so much as a porch-
is that produced on the south side by the position
of the chapter-house. The general external appear¬
ance of the church, viewed especially from the east,
can never have been really beautiful, though highly
interesting from its unique character. The entire
want of any central point to produce harmony and
pyramidal effect, the long unbroken line of roof, run¬
ning between the two low western towers, and the want
of buttresses and general plainness of design, must have
always produced a great appearance of heaviness and
flatness. The want of a central tower and transepts
hinders all external cathedral effect from any point but
the direct western view ; the general notion suggested
is that of a large parish church of extraordinary length,
an idea strengthened by the large and beautiful Lady
chapel, which, projecting, as it does, at a slightly
lower elevation than the main body, has quite the ap¬
pearance of a parochial chancel. Yet, viewing it as a
parish church, we miss the predominant western tower,
which is precluded by the only really cathedral feature
of the exterior, the superb west front. In short, its
general appearance is a mixture of two altogether diffe¬
rent types, neither of which is allowed to appear in any
degree of perfection.
Comparison with other Buildings.—Such then is
the wonderful and altogether unique appearance of this
remarkable, but, as a whole, certainly not beautiful,
church. It presents the strange phœnom'enon of a
8
LLANDAFF CATHEDRAL.
church of cathedral dignity, of a size fully entitling it
to a place among minsters of the second class, with
Southwell, Romsey, and St. David's, and still more
with its internal architectural composition quite corre¬
sponding to its size, but which yet exhibits in its exterior
only a single cathedral feature. The only approach to a
parallel that I am aware of, as I have already observed, is
Dorchester. But the case of Llandaff is less strong than
the latter, which in its absence of a clerestory and in its
distinct high gables, not only lacks all abbatial character,
but reproduces the smallest type of village churches.
Christ Church in Hampshire might be considered as pre¬
senting a slight approximation in its lack of a central
tower and the very slight importance of its transepts ;
but there is reason to believe that these peculiarities do
not form part of the original design ; and even at present,
though the two churches agree in the long extent of
building thus presented to the eye, yet they differ in a
most important respect. The peculiarity of Llandaff is
its unhrokenness ; nave, choir, and presbytery, as I before
said, are externally one. The grand characteristic of
Christ Church is that its parts, with great beauty each in
itself, are utterly incoherent. A nave, choir, and Lady
chapel, each magnificent when taken alone, are brought
together without any mutual connexion, or any attempt
to fuse them into a consistent whole.
But though we may vainly search through English
churches of equal size and dignity for an exact parallel
to the anomalies of Llandaff Cathedral, I am inclined to
think that it is only the grandest and most important
LLANDAFF CATHEDRAL.
9
instance of a tendency busily at work throughout Wales,
and indeed not unknown to England also. Every archi¬
tectural student must have observed that, of the two
types of a church, the parish church and the minster, it
is far more common to find a church entitled by con¬
ventual or collegiate rank to the latter, reduced to the
parochial type, than for the distinctive arrangements of
the cathedral and the abbey to be reproduced even in
the vastest parish churches. The contrary I conceive to
be the case on the continent. But in England, wherever
a collegiate or monastic church was also parochial, the
latter character often swallows up the former. Collegiate
churches especially, which were generally parochial, do
not, as a class, even when the present fabric is not older
than the foundation, differ much from common parish
churches ; even so large a building as Manchester does
but exaggerate the parochial type ; Beverley and South¬
well are quite exceptions to the rule ; Wimborne Minster,
but for its towers, would differ in nothing from many a
large parish church. Cathedral and conventual churches
were less commonly parochial; yet there are a good
many instances, and one can hardly fail to attribute the
peculiar character of Dorchester, in some degree at least,
to its twofold character in this respect.
In Wales the case is still stronger; of the four
cathedrals, three at least are also parochial, and St.
David's alone presents the cathedral type in its fulness.
And in several conventual churches which I saw in
South Wales I was struck with the absence of conventual
character. Excluding the very first-rate remains, like
c
10
LLANDAFF CATHEDRAL.
Strata Florida, I saw a type of large cross churclies
without aisles, of which Llanbadarn-fawr is a grand
specimen ; but it has nothing conventual about it. Such
also, as far as can be judged in their present ruined state,
was St. DogmaeFs near Cardigan, and the Priory at
Haverfordwest ; though the former does appear to have
had some part at least vaulted. Of the same type in
many respects is the admirable conventual church of
Leonard Stanley in Gloucestershire.^ Brecon Priory
may on the whole be referred to the same type, though
it has aisles, and rather more conventual character. But
I saw othér conventual churches at Cardigan, Kidwelly,
and Llangennith in Gower, quite like common parochial
structures, not even really cruciform—^Kidwelly alone
having transepts, but no central tower. Monkton Priory,
near Pembroke, is an example of the same use on a much
larger scale. And so is the collegiate church at Brecon,
notwithstanding the extreme beauty of its choir. I may
be generalizing from insufficient premises, but it struck
me that the peculiarities of Llandaif Cathedral were little
more than a more extensive development of this same
tendency.
How much the cathedral loses by all this in external
majesty hardly needs to be insisted on. One of the
greatest distinctive merits of English architecture is the
complete cruciform shape of our great Gothic churches,
marked in all its fulness by the predominant central tower.
No other outline makes a church so thoroughly a whole,
1 See Mr. Petit's Description in tlie Archceological Journal for
March, 1849, p. 44.
LLA.NDAFF CATHEDRAL.
11
and at the same time gives so much liberty in designing
the several limbs of the cross. The cruciform shape is
alike adapted to every scale above the very meanest,
and to every proportion ; the other hardly bears to be
employed in a church of more than very moderate size.
And all these faults will be more conspicuous when the
cathedral is restored to perfection than they are at pre¬
sent; now, in its chaotic state, the design is hardly
intelligible ; but let any observer continue in imagina¬
tion the roof of the presbytery to the extreme west end,
and he will speedily perceive that the church, viewed
from any point but the direct west, will be little better
than a shapeless mass of wall and roof.^ To illustrate
these remarks the more strongly by contrast, I may refer
to such churches as Stafford, Yatton, and Melton Mow¬
bray ; the western views of the two latter are among the
^ To illustrate this more fully, I have given, rather than a general
view of the church in its present chaotic state, one representing it as
restored. In this I have drawn as little as possible on my imagina¬
tion, nowhere indeed, except in the south-west tower, and the principal
features of that are taken from an old drawing of the cathedral in
possession of the Cambrian Archseological Association. The rest re¬
presents the building as it stands at present, introducing only the per¬
fectly certain continuation of the nave and aisles, the upper windows
in the east wall of the chapter-house, and the battlement of the north¬
west tower, which I have copied from that at Cardiff, which it is
always said to have resembled. I have even given the chapter-house
its present low roof, and the aisle its present unsightly battlement. I
have thought this, on the whole, the best course, as no otherwise
could the peculiar outline of the building, so well worthy the most
attentive study, be so distinctly laid before the reader ; while the
most important features of the church, as they originally stood, can
be given with absolute certainty ; and the present view, confused and
unmeaning as it is, is really hardly worth preserving.
12
llandaff cathedral.
finest to be found in any parochial buildings. All three
are much smaller than Llandaif, and not to be com¬
pared to it within, yet the effect of the bold transepts
and central steeples gives them, in their general aspect,
far more of the external character of cathedral or abbey
churches than can be claimed by their really more
splendid competitor.
I will now proceed to a somewhat more minute de¬
scription of the cathedral, as it appears at present, reser¬
ving for future consideration those points which bear
more directly upon its architectural history. And we
shall find that the awkwardness of its general outline
does not preclude very great beauty and merit in indi¬
vidual portions, while the internal effect must, when the
building was perfect, have been among the very finest of
its own style and size.
§ II.—The West Front.
We will begin with the west end, by far the most
beautiful feature of the church, consisting originally of
the gable of the nave between two flanking towers, but
of these the northern one alone remains entire. The end
of the nave is pure Early English, and is one of the very
best examples we have of an arrangement of lancet
windows. In fact this façade stands almost by itself
among English cathedrals—Ripon being, I believe, the
only other exception—as an example at this date of the
simple and beautiful arrangement so delighted in by the
architects of the previous eera. We have here no mask-
JLLAVMIIMFF
WE ST
CATHEBlliAF ,
Î:' RO.N T .
llandaff cathedral.
13
ing of the real construction by unnatural and ugly pieces
of wall, as at Lincoln, Salisbury, and Wells ; nothing
frittered away on unmeaning arcades, as in the two first
of those examples ; no sacrifice of the queen art to one of
its subordinates, as in the third ; the architectural con¬
struction is not disguised, but ornamented, and orna¬
mented in the most tasteful and judicious manner. The
nave gable and the towers rise simply and unpretend¬
ingly from the ground ; the original portion exhibits the
most perfect unity of design, and even the later recon¬
struction of the tower is harmonized therewith in a
manner displaying no mean effort of art and judgment.
West End of Nave.—The west end of the nave is
divided into three horizontal stages, the lower containing
the great western doorway, the central the west window,
the upper one the gable-light, to air the roof above the
ceiling. The first of these is perfectly plain, containing
nothing on each side the doorway, which is one of the
most remarkable, though hardly among the most beauti¬
ful, features of the church. It is in reality pure Early
English, but while its round arch gives it the general
effect of an earlier style, some of its details suggest a
later, so that it has rather the air of an inconsistent com¬
pound. The round arch, it is needless at this time of the
day to observe, throws no doubt upon the date of this
front as a matter of history, though it detracts from the
ideal purity of the style. Yet for the position in which
it is actually placed, the round head gives a more suit¬
able proportion than a pointed one. The shafts on
14
LLANDAFF CATHEDRAL.
which it rests are detached, clustered, and handed, though
now somewhat mutilated ; the capitals are remarkable for
their extraordinary height, and are richly ornamented
with representations both of animal and vegetable life.
The mouldings are singularly arranged, but plain and
not pleasing, consisting chiefly of two large filleted rolls
placed near together, without a hollow between. The
effect is heavy, and a good deal resembles a coarse form
of the double ogee of a later style. The doorway has a
tympanum—if that term be correctly applied to a com¬
position of many stones—whose lower portion forms two
round arches, looking like a mutilated double doorway ;
but there is no reason to suppose that there ever was a
central shaft, and indeed the construction of the masonry
forbids the idea ; but there may probably have been
some kind of boss, as it is now cut off in a very abrupt
manner. In the head is a vesica containing an episcopal
figure.
The second horizontal stage contains the west window,
a fine triplet and most judiciously arranged, its lights
being scattered over the whole front, instead of being, as
in some instances, gathered close together. The whole
stage is thus brought together into one design, instead of
the triplet appearing like something merely cut through
the wall. The central light is somewhat taller and higher
than the others, and the space between each is filled with
a narrower and very acute blank arch; the masonry
within each of these curves outwards in a curious man¬
ner. All these arches spring from shafts doubly banded.
Over the side lancets is a range of six small roses, three
LLANDAFF CATHEDRAL.
WE8T DOORWAY
J Prichard del
O. Jewitt sc
LLANDAFF CATHEDRAL. 15
on each side the central light. The whole composition
is extremely eflfective, and fills up the space admirably.
The upper, or gable, stage is, perhaps, not quite so
well managed, but still is very beautiful.^ A single
large central window, whose arch is round or very nearly
so, has rather the eflfect of dividing the composition into
two parts. On each side are three imperfect round-
headed trefoil arches, adapted to the slope of the gable,
so that a cusp of each is lost on the side nearest the
central window. Over this last is a single arch of the
complete trefoil form containing another episcopal statue;
the whole is crowned by a mutilated cross. The original
gable was nearly equilateral, but it has been a little re¬
duced in pitch, without diminishing its positive height,
by a slight increase in elevation given at the shoulder,
probably when the north tower was rebuilt.
From the end of the nave we turn to the towers termi¬
nating the aisles, where we find Llandaif differing from
most English cathedrals, though agreeing with many
foreign ones, in possessing two unequal and dissimilar
towers. But while in many French examples this un¬
pleasant peculiarity^ is clearly the result of design, at
1 If we may trust the drawing of the cathedral referred to in a
former note, in which, though the rest of the building is not very
exactly given, the details of the west front appear scrupulously copied,
though in a very distorted proportion—this arch was filled in, as I
had indeed previously suspected from its width, with a two-light
composition of geometrical tracery. The drawing however seems to
represent it as blank, which can hardly have been the case, unless it
were subsequently blocked.
2 I have seen it somewhere stated, but I cannot lay my hand on
my authority, nor am I certain how far the statement is correct, that
16
llandaff catfiedral.
LlandaíF, as in the analogous case of Canterbury before
the late repair, it is simply owing to a subsequent re¬
building of one of them. Certainly, from the small
traces left, the details of the two do not appear to have
been precisely identical, but we should require some
further evidence to make us believe that they violated
the ordinary English rule of being perfectly similar in
proportion and general design.
South Tower.—The southern tower was Early Eng¬
lish, contemporary with the front, but all that remains is
the square staircase-turret adjoining the nave, panelled
with a tall arch, with the curve in the masonry already
mentioned, a small portion of the west wall, contain¬
ing an elegant lancet window, and a tall fragment
-—whose preservation is remarkable—at the south-east
angle. It was somewhat lower than its fellow, the
heights being given at 89 and 105 feet respectively ;
its upper portion was unbuttressed, with single belfry-
windows, and a parapet and pinnacles, which seem to
have been a Perpendicular addition, though not at all
of the same richness as that of the north tower. To
judge from the drawing, neither in outline nor in detail
can it have been worthy of the front of which it formed
a part, rising but little above the apex of the gable, and
having altogether a bald and meagre appearance.
North Tower.—The north tower is a fair Perpen-
in France a special rule confined tlie use of a pair of equal towers to
archiépiscopal churches and royal abbeys.
LLANDAFF CATHEDRAL.
17
dicular structure. Some portions of the original tower
remain within, together with the lower part of the stair-
turret at the south-east angle, which, it will be seen, is not
quite identical with that on the other side. This tower
seems to be borrowed from the Somersetshire type, but
plainer, and, as is adapted^ to its position, of more massive
proportions. It may be in some respects compared with its
neighbour St. John's at Cardiff, and the two well exem¬
plify the difference in treatment between a tower forming
a portion of a front, and one at the end of the building.
But the comparison is hardly fair, as the Llandaff ex¬
ample is much better in its own kind than that at Cardiff.
The latter is well-proportioned, but belongs to the in¬
ferior^ type of the Somersetshire towers, which consist of
a mere aggregation of stages, not, like Wrington and
Glastonbury, of one harmonious design throughout.
But, while this is very conspicuous at Cardiff, it is not
so at Llandaff, on account of its more massive propor¬
tions ; indeed the gorgeous mass of panelling presented
by the other two examples could hardly have been toler¬
able except in a tower which had a much greater portion
of its height free than this has. The low belfry-stage
could only have presented a travestie of the soaring ones
so magnificent in the other cases. As it is, it is quite
satisfactory, and the three-light belfry-window fills up
the space, while the long single two-light one at Cardiff
1 That is considered simply as a member of a façade. For the
effect of the whole church, next to a central tower^ a pair of lofty
western ones, with tall spires, would have been most desirable.
2 See the author's History of Architecture," p. 386.
D
18
LLANDAFF CATHEDRAL.
has a meagre effect. These belfry-windows possess
some of the best instances of that beautiful eniichment
usual in the west of England, the ornamental stone¬
work between the mullions instead of the common
luffer-boards.
But both towers, Llandaff and Cardiff, share, or rather
have shared, one great fault ; a want of connexion be¬
tween the parapet and the lower part of the tower. The
original one at Llandaff has been destroyed ; but it is
easy to supply it from its neighbour. In the best Somer¬
setshire towers there is a gradual increase of ornament
from the bottom ; each stage is richer than that below it,
and the gorgeous open parapet crowns the whole. The
pinnacles at the angles are the continuation of the but¬
tresses, and if there are smaller ones between, they rise
in a similar manner from between the belfry-windows.
But at Cardiff a parapet of extreme gorgeousness is
added, without connexion or preparation, to a tower
whose whole height is equally plain. And, except at the
north-east angle, where the octagonal staircase-turret is
very skilfully treated, the parapet is actually top-heavy ;
and, though the pinnacles are very large, and assume the
form of open turrets, yet they have no connexion what¬
ever with the buttresses, which finish, with small deco¬
rative pinnacles of their own, just below the cornice.
This last fault must have been still more conspicuous at
Llandaff than at Cardiff; in the latter, the buttresses
being diagonal, the eye may possibly be carried, though
not by a very uninterrupted course, up their pinnacles
to the small ones (connected with the open turrets by
LLANDAFF CATHEDRAL.
19
diminutive flying buttresses^) which stand, external to
the parapet, on the extreme edge of the cornice. But at
Llandaflf the buttresses, though crowned with the same
small pinnacles as at Cardiff, are double, and placed at
some little distance from the angles ; consequently there
is no attempt at connexion whatever. These buttresses
offend also, not so much perhaps by their great projec¬
tion, as by the extraordinary multiplicity of small stages
into which their height is divided. Yet this same ar¬
rangement has an admirable effect in the tower of St.
Sepulchre's, Northampton. But there the buttresses are
diagonal, which would not have suited this position.
In fact I think this elaborate parapet at Llandaff was
altogether a mistake ; more so even than at Cardiff, be¬
cause the low^ and massy character of the tower renders
it still more inappropriate. A good plain moulded
battlement, without pinnacles, such as is seen in many a
parish tower, would be far more in keeping ; but one
could hardly presume to recommend to the present re¬
storers of the cathedral so wide a departure from the
original design. I may venture to mention that, till I
saw the church, and perceived the modern and meagre
character of the existing battlements, I had always, in
contemplating the building from engravings, applauded
the supposed discretion of the architect in not introducing
1 These are found in several very fine towers, but I can never quite
reconcile myself to them. They are surely carrying the textilis aura
notion too far for external stone-work,
^ As regards the whole bulk of the church ; as a detached campa¬
nile, or western tower to a smaller church, it would not be remarkably
massy.
20
llandaff cathedral,
anything more elaborate. My facts were wrong, but I
have seen no reason to change my opinion.
Like Cardiff, this tower has an octagonal staircase-
turret, though at the opposite angle ; it is ingeniously
introduced on the base of the old square one, and appears
to have terminated in a rich open spire. There is no
doorway in either tower, the west windows of the aisles
being placed very low ; that in the north tower is exter¬
nally a Perpendicular opening, but it has lost its tracery.
Perhaps three western doorways would have been more
than so small a building could have required, either for
use or appearance.
§ III.—Nave and Choir.
Present Appearance.—The main body of the fabric,
as seen at present, appears one mass of hopeless con¬
fusion ; the hideous structure of the last century which
acts as the choir rising in strange contrast to the vene¬
rable ruins to the west of it. This latter portion of the
church is not simply unroofed, but both the external
walls and the arcades are very much mutilated. The
former, by dint of patching, present a continuous circuit,
though not to their full height; within the piers and
arches are perfect, except the arches nearest to the
present west front (!) which have been apparently muti¬
lated to make way for that precious monument of the
taste of a century back. Of the clerestory only a small
fragment remains in a single bay, but fortunately enough
to reconstruct its design.
llandaff cathedral
21
Along the modern choir the aisles have been preserved
untouched, except by the addition or substitution of an
unspeakably wretched battlement, far worse than that of
the tower. Over these rises the present clerestory, with
four round-headed windows of the meanest kind, and
meagre vases at the corners. A wall built across the
nave forms the west front alluded to in the last para¬
graph. The barbarism of this cannot possibly be sur¬
passed. The structure considered simply on its own
merits, is one of the most wretched of its own wretched
class, presenting nothing in the way of columns, rich
cornices, or any of the other ornamental features of the
style. And the incongruity of its position is quite un¬
paralleled ; the contrast between the stuccoed baldness of
the new front, and the grey and ivied ruin close to it ;
the grotesque absurdity of the clerestory rising above the
old aisles, so that the precious building loses whatever
chance it might have had from its natural proportions
and arrangements; the further contrast now produced
by the stately high roof recently added to the presbytery;
all unite to stifle every feeling of compunction ; even the
two or three architects who still continue to echo the
rapturous exclamation of Mr. Eustace—How superior
are pilasters to buttresses and colonnades to arcades!"
could not refuse to lend a helping hand in the removal of
a structure which does not meet even their requirements.
Yet between the Taylor buildings and the choir of Llan-
daif who may decide ?
Original Design.—Exterior.—Enough however re-
oo
.'ai At)
LLANDAFF CATHEDRAL.
mains to enable us fully to realize its appearance previous
to the commencement of this work of devastation. The
choir and nave formed architecturally one piece, and the
long range of the clerestory—seven bays, exclusive of
those concealed by the towers—must have been extremely
effective. The arrangement consisted of two distinct
lancets in each bay, the bays being divided by flat pilas¬
ters running into a corbel-table ; a treatment of Roman¬
esque origin, but often retained in Lancet work, and with
extreme propriety in instances like this, where the style
is exhibited in an early and somewhat severe form.
The character of the aisles is very inferior ; being en¬
tirely without buttresses, the effect is bare and meagre in
the extreme ; and the range of three-light ogee-headed
windows—their tracery being a bad specimen of the mo¬
notonous Reticulated form—presents no relief or satisfac¬
tion to the eye. There is no porch, but ample means of
entrance are provided by two doorways on each side:
The western and larger pair are late Norman, and of very
considerable richness. " The former of these" [the
southern] says the present Dean, " is by far the most rich
and remarkable in its decorations ; its outer moulding is
of a pattern closely resembling the ordinary Etruscan
scroll—a circumstance, I believe, without any other ex¬
ample in our Norman ornaments ; the other features
consist of three common Norman shafts, supporting
arches moulded in the usual style of that sera, the central
member bearing a double lozenge moulding, and the
inner and outer zone each ornamented with double lines
of the common zig-zag mouldings ; within this series of
LLANDAFF CATHEDRAL.
IMPOST AND PORTION OP ARCH OF SOÖTH DOORWAY IN NAVE.
NORTH DOORWAY IN NAVE
J. Pricbard del
O Jewitt SÜ.
llandaff cathedral.
23
receding shafts and their arches, the side-jambs of the
doorway and the arch above have a moulding of angles
united by straight lines. The corresponding doorway on
the north is much simpler, but it is surmounted by a
dog-tooth moulding, although this feature is generally
characteristic of a style later than the true Norman ;
beneath this occurs a series of receding mouldings, alter¬
nately lozenge and zig-zag ; these are supported only by
a single shaft." ^
The smaller pair of doorways'more to the east are less
remarkable, being Decorated, contemporary with the
windows, and having merely moulded jambs, without
shafts or other enrichments. Over the southern one is a
curious niche of the vesica form ; the bracket for the
image still remains, but the image itself is gone.
Interior.—The internal view of the nave and choir
must have been, beyond all comparison, the most beau¬
tiful and attractive feature of the cathedral. Its effect,
when perfect, when the eye could gaze uninterruptedly
down its whole length, must of course be, for the present,
left to the imagination, but it is one which may very
easily be imagined ; and, even now, the appearance of
the ruined nave is one of the most striking to be found
among the remains of ancient architecture. Its roofless
condition supplies the incidental advantage of a rare and
beautiful combination ; the arcades and the interior of
the west front being seen in close juxtaposition with the
tower as an external object. And in no ruined ecclesi-
^ Archceologia Canibrensis, Yol, I., (New Series,) pp. 29, 30.
24
LLANDAFF CATHEDRAL.
astical building is less subject of offence given ; the
roofless portion being still sedulously preserved as a
portion of the church, and guarded from all injury or
disrespect. And, viewed architecturally, the merits of
this part of the building are very great ; though neither
large nor richly adorned, it may claim a high place on
many grounds among buildings of its own date and style.
The character of the Early English work is singularly
good ; besides its excellent proportions, it combines, in
a most remarkable degree, a great lack of ornament,
with not only the utmost excellence of detail, but a
considerable effect of richness. This is probably owing
to the finish of execution, which is most conspicuous,
taking away all notion of rudeness, and to the presence
of floriated capitals, which certainly impart a much
greater character of enrichment than any other individual
member. The internal treatment of the west end is
especially excellent, and deserves the more attention, as
the mean internal appearance of a western portal is often
a marked blot upon churches of great magnificence.
With regard more immediately to the doorway, this
is chiefly the result of bringing down some of the
jamb-shafts of the windows to the ground on each side
of it, which at once has the effect of making the
latter a real portion of the design, and not a mere neces¬
sary evil. But, far beyond this, the western triplet,
viewed internally, is most admirable ; even the external
beauty of the west front in no way prepares the spectator
for the marvellous display of art which it presents within.
The fall of the ground allows a great increase of height
LLAFBAIFF €ATM FBRAI. .
THE HATE
llandaff cathedral
25
beyond that of the external façade—there being a descent
of several steps into the nave from the west door—of
course greatly to the improvement of the general effect.
This allows much greater height to be given to the
central stage containing the window, without encroach¬
ing on that below ; the space occupied without by the
tympanum of the doorway being taken into the former,
while the loss is made up to the latter by the space
gained below the external basement. The height thus
gained allows the triplet itself, with a rich array of arch-
mouldings and jamb-shafts, to occupy the whole width
of the church, (the narrower intermediate arches of the
exterior not appearing within,) without the width of each
lancet being made disproportionate. The internal jamb
comes, of course, considerably lower than the real cill of the
lancets. The skill with which the internal and external
arrangements, each the better suited for its own position,
are adapted to each other, deserves our best study and
admiration. Too often the west front, being a mere ex-
cresence without, contributes nothing to the internal
effect ; here the discretion of the architect, in his simple
and natural treatment of the exterior, has had its reward
also within. Instead of a mere mask without, and a bare
wall within, we have the same feature of consummate
beauty in both ; only that of the exterior is but the shell
of the far higher loveliness within.
Division of Nave and Choir.—In the interior a little
attention will readily discover, what a consideration of the
exterior will not reveal, namely the respective limits of
e
26
LLANDAFF CATHEDRAL,
the ritual nave and choir. The part of the church we are
now examining, namely the constructive nave of eight
bays, contains, as I mentioned before, both the true nave
and the ritual choir. The limits of the two may readily
be ascertained, especially as the old arrangements of the
choir appear to have been retained after the changes
of the last century. The stalls occupy the same position
now as they did in Browne Willis' time, namely the
two eastern bays of the constructive nave ; these are
distinguished by a solid screen between the pillars, which
is contemporary with the arcades, as is shown by the
stilting of the bases, and by the insertion of an Early
English sepulchral niche in the western bay on the south
side. The third arch from the east may well have been
filled by the rood-loft, and the remaining five have re¬
mained as the real nave. This is distinguished from the
ritual choir, not by any constructive feature, but by a
change in the architectural detail, precisely as is the case
in Westminster Abbey, where the same arrangement is
followed. The piers in the ritual choir are of a different
section from those in the nave, being composed of fewer
members, and having recessed instead of projecting roof-
shafts.^ The width of the ritual choir from pier to pier
is thus made somewhat greater than in the real nave,
probably to gain more room for the woodwork required
1 Similarly there are, in tlie nave, shafts towards the aisle supporting
no part of the arch, which are absent in the choir. We may perhaps
infer that the nave aisles were designed for vaulting, and the choir
aisles not, but this is not absolutely conclusive, as there are similar
shafts in the choir aisles at St. David's, where no contemporary vault¬
ing could have existed.
LLANDAFF CATHEDRAL.
í^/vv
INTERIOE ELEVATION OF ONE BAT OF THE NAVE
Scale, one-eighth of an inch to a foot
J. Pricbaid, del O Jewitt, sc.
llandaff cathedral.
27
for the former, without trenching more than was abso¬
lutely necessary upon the open central space.
Arcades.—The lateral elevations of this church well
deserve attentive study. They consist simply of a pier-
range and clerestory, there being no architectural trifo-
rium, but merely a passage in the clerestory. This again
displays the judgment of the architect; the positive
height is srnall, not too small for the complete triple
division, had the style been Romanesque ; but the inser¬
tion of a triforium worth inserting, one of any dignity,
and really forming a distinct feature in the composition,^
would have allowed less space to the pier-range than was
required by the spirit of the Gothic style and pointed
arch. The result is a pier and arch of very excellent
proportion, combining sufficient grace and lightness with
great solidity.
We may not inappropriately contrast the arcades of
Llandaff with those of Southwell Minster, of much the
same scale and style. Here also the height did not
admit of an architectural triforium ; but the arrangement
is very different from that of Llandaff. At Southwell no
height is gained to the pier-range, but the triforium is
taken into the clerestory. That is, there is not, as at
Llandaff, a mere passage in front of the window, but a
considerable blank space below the window-cill belongs
to what is architecturally treated as the clerestory-range.
The consequence is a much more massive pier-range, of
a proportion more resembling those of churches with a
^ See " History of Architêcture," pp. 364-66.
28
LLANDAFF CATHEDRAL.
large triforium. as Romsey and Ely ; the piers are lower^
the arches less bold and soaring. Hence also a diíFerent
treatment of the pier-range in point of detail ; that at
Llandaif, being light and Gothic in itself, requires but
little aid from ornament, and is left comparatively plain ;
at Southwell, the range, having a proportion rather
Romanesque than Gothic, seeks, and most effectually
finds, relief from its massiveness in a much greater
elaboration of mouldings in the architrave, and of clus¬
tered shafts in the pier. Both arrangements are excel¬
lently managed ; and both, I venture to say, are greatly
superior to the—in this style—extravagant lightness of
the arcades in the nave of Lincoln, which seem suited
only to churches like Dorchester or Stafford, where there
is little superincumbent mass.
The difference between the piers in the nave and choir
will be best studied in a section ; it will be observed that
in neither is there any great freedom of clustering. The
pier is neither an aggregate of small shafts, nor yet
an assemblage of such around a larger central one, but
consists of an angular mass with a cluster of three shafts
attached to the principal faces. The shafts have the same
keel form as those at St. David's, and one or two share
the peculiarity of the absence of the neck-moulding.^
The arch is without any moulding, strictly so called, but
besides an outer label of the keel form terminating in
foliage, another similar one is inserted in an angle of the
1 I have been struck by the general resemblance of the Early
English work at St. David's, Llandaif, Cheriton in Gower, and
Slymbridge in Gloucestershire.
LLANDAFF CATHEDRAL.
INTERIOR ELEVATION OF ONE BAY OF THE CHOIR
Scale, one-eighth of an inch to a foot.
J Prichard. dol
O Jewitt. 8C
LLANDAFF CATHEDRAL.
29
pier ; an arrangement of which I do not know another
instance. The eastern arch of the choir^ on each side, is
blank, a smaller one being open under it, a singularity
whose meaning we shall have hereafter to explain. This
pair of arches, and that next to it, forming the extent of
the ritual choir, have been already mentioned as being
filled up between the piers with a solid wall. This is in
fact a screen to form a backing for stalls, and is not un¬
usual in choirs, as at Romsey, St. Cross, and Rothwell in
N orthamptonshire.
The clerestory appears to have consisted of the two
windows in each bay, each with a smaller arch on each
side. The bays are divided by shafts running up the
whole height of the wall; in the choir, as was before
said, these are let into the wall ; the arrangements of
the clerestory here are utterly destroyed. In the nave,
a smaller cluster of three with a base rests on the
capital of the inner cluster of the pier. This is not
a very common arrangement, being in fact something
intermediate between the common Early vaulting-
shaft corbelled off above the pier, as at Ely and other
instances, and the Continuous vaulting-shaft rising
uninterruptedly from the ground. It is an attempt
to combine the general effect and magnificent vertical
expression of the latter arrangement, with the distinct
existence of the pier characteristic of the Early Gothic.
It is something intermediate between the merely hori¬
zontal division of the Early English portions of St.
Alban's, and the almost premature verticality of the
nave of Lichfield.
30 llandaff cathedral.
Roof.—But these shafts are most important as reveal¬
ing to us the original covering of the nave and choir.
There can be little doubt but that this part of the
cathedral was originally^ intended to be covered by a
flat ceiling, a feature more common in Romanesque
and Early Gothic churches than is usually supposed.
This is shown by the roof-shafts, which are continued
up to the summit of the masonry, instead of being
terminated much lower down, as they must have done,
had vaulting of the ordinary kind been intended.^ And
that these shafts were designed to carry a flat ceiling,
and not an open or canted timber roof, or a barrel
vault, appears from the internal view of the west end,
where the ledge for the ceiling to rest upon is distinctly
visible, and while the masonry below is of ashlar, that
above, which would have been concealed by the ceiling
is of rubble. This arrangement we cannot conceive co¬
existing with any other form of internal covering.
I shall probably stand almost alone in thinking not
only that such was the historical fact, but that the archi¬
tect judged rightly in adopting this form, provided the
natural and more appropriate finish was, for any reason,
out of the question, and in regretting that, supposing any
1 It may have been only contemplated, and never actually erected,
just like so much vaulting at St. Davides and other places, but I see
no reason why it should not.
2 This proof is of course conclusive, but we may remark that the
arrangement of the clerestory is adapted only to a horizontal finish of
the wall. The two distant lancets are evidently designed to fill up
the whole parallelogram ; vaulting, unless sexpartite, would interfere
with the whole design. Compare the clerestory at St. Alban's, which
is also fiat-roofed.
llandaff cathedral.
31
impediment precluded there also the addition of a vault,
the precedent has not been followed in the new roof of
the presbytery. I have so largely argued out this
question elsewhere,^ that I will now only repeat my
unchanged opinion that the carrying out of verticality
by contrast so conspicuous in this mode of roofing pro¬
duces a much greater effect of height, and is a better
development of the Gothic principle than the high-
pitched timber roof added to the walls without any
œsthetical connexion with them.
§ IV.—Presbytery, Chapter-house, and Lady Chapel.
Exterior of Presbytery.—This part of the building
is the constructive choir, if we may so speak when there
is only one degree more architectural difference made
between the ritual choir and the presbytery than exists
between the nave and the former. Within there is
indeed an arch,^ but externally the whole fabric seems to
have been continued with the same width and elevation ;
the only distinction being in the arrangements of the
clerestory. The present clerestory of the presbytery is
entirely new, its predecessor having been entirely de-
1 " History of Arcliitecture/^ p. 347. The general argument tells
equally, whether the flat internal covering be a ceiling beneath an
external roof, (which may be of any pitch,) as in all the Early and
many of the later examples, or whether the roof itself be made flat or
nearly so, as in most Perpendicular examples. Yet the effect of the
two arrangements is by no means identical.
2 At present a new one, but bases were found, proving that one
anciently existed.
32
LLANDAFF CATHEDRAL.
stroyed. The aisles, also, though untouched in their
mass, had received greater alterations than those of the
choir, as the windows were substitutions in the style of
the choir clerestory. Happily however the jambs were
left sufficiently perfect to admit of a certain restoration,
the tracery alone being left to conjecture ; and we shall
see hereafter that these jambs are not among our least
important guides in tracing out the history of the build¬
ing. They are Decorated, of an earlier character than
those of the nave and choir. The presbytery itself con¬
sists only of two bays, but the aisles consist of three,
being continued a little way along the Lady chapel. On
the north side the aisle presents a continuous range from
the west to the east end, with a regular succession of
windows, broken only by several interruptions in the
masonry, and by the very irregular direction of the wall,
consequent probably on the numerous patchings and
alterations which it has sustained. And it must be
confessed, that the prospect of this long extent of
unbuttressed wall, at once monotonous and irregular, is
one of the least pleasing to be found among our greater
churches. On the south side the projecting and quasi-
transeptal chapter-house provides the only break in the
whole church. This occupies the western bay of the
south aisle of the presbytery. The remainder of the aisle
also is less uniform than its fellow. It is lighted by two
windows, but irregularly placed, and of altogether diffe¬
rent designs. The western one is a large flat-headed
Decorated window of five lights ; the mere occurrence of
a flat head in this style is of course nothing remarkable.
LLANDAFF CATHEDRAL.
33
being common enough in small churches, especially in
Northamptonshire, but it is not very usual to find it
either on so large a scale, or in so great a church.^ The
reason is obvious ; in the low wall of the aisle or cleres¬
tory of a parish church it is often desirable to make a
window of greater width than the space would allow, if
a pointed arch were employed.^ Hence the square-
headed and segmental-arched windows of the Decorated
style, and the four-centered arch of the Perpendicular ;
forms meeting the practical requirements of the position,
and not inconsistent with the horizontal cornice of the
timber roof above. But as the form can but seldom be
necessary in the prominent positions of a small church—
the large window at Northborough is clearly an indi¬
vidual freak—so it is still less in harmony with the
regular arrangements, the tall narrow bays, and pointed
vaulting cells of a cathedral aisle. It is only the vault-
less aisles and semi-parochial character of Llandaif that
could have admitted of such a whim ; even here no ne¬
cessity or advantage called for it.
The other window is an ordinary pointed one ; in both
the present tracery is conjectural, but the jambs original,
or at least completed from fragments remaining of the
original work. An ugly modern door between them is
a great eyesore. The windows at the east end of the
aisles have Perpendicular tracery inserted in the Early
Decorated jambs. At this point we arrive for the first
1 Those in the aisles at Peterborough, Early Decorated, are a re¬
markable exception.
2 See History of Architecture,^' p. 350.
F
34
llandaff cathedral.
time at the consolatory presence of a buttress. Those at
the angles of the aisles are pedimented, but low, not
rising above the cill of the windows. A rich open para¬
pet has been added during the present restoration to the
part of the south aisle east of the chapter-house, the effect
of which is extremely good.
Interior of the Presbytery.—The internal view of
this part of the church presents a remarkable contrast to
that of its western portions. There the prevailing style
was not only beautiful in itself, but was perfectly and
harmoniously carried out, the Decorated rebuilding of
the aisle walls in no way interfering with the general
effect of the Early English arcades. But in the presby¬
tery the greater proportion of the work is at once of a
very inferior character in itself, and is strangely and
awkwardly introduced into an earlier design. Decorated
arches are cut through Norman walls, and on the south
side in a very singular manner. Here we have, as
every visitor to the cathedral must have observed, the
remains of two Norman windows cut through by the
present Decorated pier-arches. On the north there are
two complete pier-arches, entirely obliterating all such
traces. Fragments of a Norman string were, however,
discovered during the restoration. On the south we
have only one complete arch, with the head of a Nor¬
man window appearing above it, and the beginning
of another, which cuts into another Norman window,
and stops suddenly, leaving the western jamb of the
latter quite perfect. It is clear, then, from this and
LLANDAFF CATHEDRAL.
35
from other reasons, that this Decorated arch never could
have been intended to be completed, and it is diíEcult to
understand why it was ever commenced. To the west of
this is a solid wall, perforated only by a modern door¬
way, leading into the south aisle.
The comparison between this Decorated presbytery
and the Early English parts is, as is so often the case
when those two styles are brought into close^ juxta¬
position, extremely painful. I am not clear that the
section of the piers is not an imitation of the older one,
but the beauty of proportion is lost, and the poor
moulded capitals at once strike the eye by their inferi¬
ority to the beautiful foliage of the earlier portions;
one wonders that some of the approximations—distant
indeed—which the Decorated style could supply to the
consummate loveliness of its predecessor, such as we see
in Bishop Gower's work at St. David's, were not called
in to avoid so humiliating a contrast. Still the general
effect of the presbytery, though spoiled by its contiguity
to such a rival, is by no means to be despised. Not a
little of its merit indeed is derived from the new cleres¬
tory, the appearance of which is highly satisfactory, and
harmonizes very well with the arches below. The idea of
the latter has been caught, namely a Decorated version
of the nave, there being no distinct triforium, but a mere
passage in the clerestory range. The open roof too,
though a feature quite out of place in a cathedral, has
perhaps as good an effect as one of the kind can have.
The presbytery is connected with the Lady chapel by
a large and wide Norman arch, like a chancel arch, and.
36
LLANDAFF CATHEDRAL,
as is usual in arches of that kind, with the western face
much the more enriched of the two. Unfortunately
being now filled with a partition of boards separating the
part of the church at present used for divine service from
that still undergoing repair, a good view either of its
general effect, or of many of its details, cannot be ob¬
tained. But, besides the rich studded moulding, figured
in the "Glossary," and commented on by the Dean,^ there
is much in this arch worth notice. It is needlessly and
awkwardly stilted ; and we may especially remark the
two shafts side by side continued in heavy rolls along
the soffit; a feature characteristic of Early Norman
work, and in fact most probably the retention of an
Anglo-Saxon usage.
In the aisles of the presbytery we have chiefly to
remark the western bay on the south side, which
forms a vestibule to the chapter-house, and is made
quite distinct from the rest of the church, opening into
the aisles by not very lofty arches, and into the
chapter-house and the presbytery itself by mere small
doorways. This bay is vaulted—being the only portion
of the cathedral west of the Lady chapel that is so—
in a Transitional or incipient Early English style, less
advanced in' character than that of the choir, though,
as we shall see, not impossibly of the same date.
The vaulting is pointed, but the ribs, and two of the
capitals of the vaulting-shafts must rather be called
Romanesque, while the other two capitals are good
Early English.
1 Archœologia Camhrensis^ iit supra, p. 28.
LLANDAFF CATHEDRAL,
IMPOST A.ND PORTION OP ARCB. BffiTWBliíN PRaSBíTERY AND LaDY UHAPEL
J iTicbard, del
O.Jewiit sc
llandaff cathedral,
37
The Lady-chapel.—The beautiful chapel which com¬
pletes the design to the extreme east is one of the most
pleasing features of the church. Viewed externally, there
is nothing to be at all compared to it, except the west
front, and within, its beauty, though of a very different
kind, is fully equal to that of the nave and choir. It is
moreover remarkably contrasted with the rest of the
building ; without, by its tall buttresses and long narrow
windows ; within, by its beautiful stone vaulting. I have
already remarked that, standing as it does, it suggests
the notion of a parochial chancel, magnificent in itself,
though somewhat small for the general proportions of
the building. Taken alone, it would form an excellent
model for a collegiate or palatial chapel.
The style is transitional from Early English to Deco¬
rated, having Early Geometrical tracery in the windows,
and some of the foliage being rather of a Decorated
character, but in other respects it hardly departs from
the purity of the Early English style. The effect of the
external range of buttresses and windows along three
regularly designed bays is very characteristic. The but¬
tresses are of three stages, the lowest being pedimented at
the window-cill, ranging with the buttresses at the east
end of the aisles. The upper stage is carried up unusu¬
ally high for buttresses not terminating in pinnacles, the
set-off being in the parapet, which is remarkably deep.
The buttresses are doubled at the angles, and there is
a small pedimented one beneath the cill of the east
window.
The tall windows are of two trefoil lights with a plain
38
llándaff cathedral.
circle in the head. The jambs are plain without, but
within the rear-arch is deeply moulded, and the jamb
adorned with that exquisite ornament, the detached shaft
of Purbeck marble. The eastern window, like so many
others, having received a substitute in the style of Mr.
Wood, has authentic jambs, but conjectural tracery.
There are five bays of good simple vaulting, rising from
Purbeck shafts, and with bosses at the intersections. The
whole effect of the internal view is probably surpassed by
no building of its own class—though, as it differs so
widely from most Lady chapels, that class is a small one ;
it would be hard to find an arrangement, or even a
detail, open to reasonable censure. Externally it cannot
so well be considered separately, and can only be re¬
garded as part of an ill-proportioned whole ; yet, so far
as it can be taken distinctly, it is singularly beautiful.
It cries however for a high roof, and will do so still more
loudly, when that over the rest of the church is com¬
pleted.
The aisles being prolonged along two bays^ of the
Lady chapel, there are necessarily arches between them ;
their character is somewhat singular, as the bases are
Early English, the floriated capitals certainly so, if clear
of all Norman traces, while the abaci are Decorated.
The Chapter-house.—The Cathedral is completed
by the quasi-transeptal chapter-house, projecting, as we
have seen, from the south aisle of the presbytery. Its
1 Two bays of the vaulting of the Lady chapel, though lighted by
only one window in each aisle.
LLANDAFF CATHEDRAL.
INTKHIOR ELEVATION OF ÖIDE
WINDOW OF LADY CHAPEL
Scale, bait" an inch lo a foot
CAPITAL. BASE. AND PLAN OF THE ARCH BEIWEBN
LADY CHAPEL AND SOUTH AISLE.
Scale, one inch to a foot
J. Pcichard. del O. JewiLt »c
CAPITAL. NORTH AISLE
Scale one inch to a foot
LLANDAFF CATHEDRAL
39
description will form a good winding up of our survey
of the present state of Llandaif; as we started with a
front unrivalled among those of its own date and scale,
we conclude with a portion of the fabric, which, if less
remarkable for beauty, claims a place second to none
on the score of singularity. For this chapter-house
is, among English cathedrals at least, absolutely unique.
There are two normal forms, the earliest and latest
being simple oblong rooms, while the intermediate
period produced the polygonal form, which, with
the two exceptions of York and Southwell, is vaulted
from a central pillar. But ut Llandaff we have a square
building with a central pillar ; the effect is not pleasing,
being that of a square playing at a polygon, just as the
sexpartite vaulting over the eastern bay of the choir of
St. Cross, and of the south aisle of Dorchester, give their
flat east ends the appearance of playing at apses ; but,
viewed historically, there can be little doubt but that we
have there not a confusion of the two types, but a genuine
example of transition between them. The architect evi¬
dently preferred a vault of a greater number of bays to
the heaviness of one vast square bay over the whole
apartment, or even to two oblong bays. He designed
his roof of four bays, which consequently required a
central pillar to support it ; it is exactly the same arrange¬
ment as in the great staircase at Christ Church, though
that, perhaps from its greater size and different use, does
not in the same way suggest the polygonal form. The
central pillar is a plain round one, and there is no great
amount of detail in the building.
40
LLANDAFF CATHEDRAL.
Externally, the chapter-house, though of better execu¬
tion than the Decorated aisles, is not altogether satisfac¬
tory. The fault seems to be that it in nowise proclaims
its own use ; it breaks the monotony of the outline, but
breaks it as an excresence, not as an integral part of
the fabric. It is sufficiently like a transept to make
the lack of the cruciform shape still more apparent ;
sufficiently unlike to do nothing towards supplying its
place. And it is not by any architectural propriety,
but simply by a sort of process of exhaustion, that its
real nature could be conjectured. Yet regarded with¬
out reference to its own end, or to the general grouping
of the church, it offers much that is attractive. It has
what at Llandaff must be considered the honourable
distinction of buttresses, and the windows, trefoiled
lancets, with a deeper external splay than usual, are
extremely graceful.
The upper part of this chapter-house has been much
mutilated, and it is at present covered with a very mean
roof. It is not easy to say what would be the appropiate
finish ; a flat roof and parapet would hardly be in cha¬
racter either with the style of the chapter-house itself, or
with the general appearance of the cathedral ; and the
lofty cone so admirably suited to a round or polygonal
structure, is perhaps not so well adapted to its square form.
There would appear to have been an upper story, more
than a mere space above the vaulting, as the jambs of
two mutilated lancets remain, an ugly modern window
having been inserted between them, and a piece of a
Decorated shaft—I know not from what part of the
LLANDAFF CATHEDRAL.
41
building—worked into the wall. There is no trace of
similar lancets ever having existed on the south side,
where however their presence might have been an im¬
provement, as the extent of blank wall over the lower
pair, which are placed low, has a somewhat meagre
effect.
Gr
CHAPTER IL
HISTORY OF THE FABRIC.
Having in the previous Chapter endeavoured to describe
the original design and present appearance of this extra¬
ordinary building, and fairly to weigh in the balance
of architectural criticism the merits and defects of its
singular position, its unique outline, the remarkable
beauty of several individual portions, it now follows to
trace out, as far as may be, the history of the fabric ; to
assign the several portions to their respective dates, and
to ascertain the plan on which the authors of each suc¬
cessive change conducted their proceedings, how much of
each scheme of restoration was accomplished, how much
existed only in the designs of its projectors, and how
much of the former still exists, how much has been sacri¬
ficed to still later changes. For an architectural history
can hardly be considered complete, unless it strives not
only to assign its date and author to each existing por¬
tion, but to reproduce, whenever practicable, the appear¬
ance of the building after each successive epoch of
change ; and it frequently happens that some small frag¬
ment of detail, some apparently insignificant change of
masonry, may reveal the fact of important and extensive
LLANDAFF CATHEDRAL.
43
alterations, of which little^ or no trace is to be found
in the appearance of the fabric at the present. This is
conspicuously the case with the subject of our present
inquiry ; we shall presently see that there is good reason
to believe that Llandaff Cathedral once presented, or was
at least designed to present, an appearance, not only in
detail, but in the grandest features of its external out¬
line, as far removed from the conceptions of its original
builders, as it is from any notion that could be derived
from the condition to which the chui'ch has finally been
brought. Outline, ground-plan, ritual arrangement,
have been completely altered more than once ; a small
building has been gradually developed into a large
one, and nearly all of this has occurred, apparently
without any chronicler to record some of the strangest
transmutations in architectural history. As the present
work of chronicling the destiny of Llandaff has not
fallen upon a Willis, neither has any Gervase been
found to clear the way where his assistance was so much
more needed.
And, besides this, the church has been subjected to
such a number of alterations following so closely upon
one another that it is often almost impossible to ascertain
their exact extent ; and, moreover, additional difiiculties
are produced by the most important reparations having
been so gradually carried on, as to allow of consider¬
able changes of style during their continuance ; some
of them, too, have produced such strange and unac¬
countable patching ; in a word, the whole character of
the building, and of the changes which it has undergone,
44 llandaff cathedral.
is SO thoroughly anomalous, that to unravel its history is
one of the hardest tasks that the architectural inquirer
could have undertaken.
In arranging the history of the cathedral, we shall
find, observing the caution given in the last paragraph,
that its existing features may be referred to three main
heads. I.—The original Romanesque fabric of Bishop
Urban, the earliest building on the site, of which any
portion remains. II.—Larger Early English additions,
which prolonged the church to its present extent west¬
ward. III.—A systematic Decorated repair, remodelling
the Norman portions left under No. II., and rebuilding
nearly the whole of the external walls. Under this head
I reckon the Lady chapel, though rather Early English
than Decorated, because it has no connexion with the
earlier Lancet work, while it can hardly be separated from
Decorated repairs apparently carried on uninterruptedly
from its completion. Finally, we have Jasper Tudor's
Perpendicular tower ; but this, as an incidental rebuilding
of an individual feature, does not affect the general his¬
tory of the building.
§ 1.—The Romanesque Church.
British Period.—Llandaff is usually regarded as the
most ancient episcopal see now existing in Great Britain ;
that is, I imagine, that no other of equal antiquity re¬
mains in its original site. And it so happens that the
early history of the establishment is better known than
the later, the celebrated Liber Landavensis unfortunately
LLANDAFP CATHEDRAL.
4S
terminating just at the point where for the purposes of
the present inquiry it should have commenced. I am
not acquainted, either from personal observation or au¬
thentic report, with any ecclesiastical structures in Wales
which can be certainly referred to a date earlier than the
Norman invasion ; certainly nothing of this kind is to be
found in the tAvo cathedrals which are the greatest archi¬
tectural ornaments of the Principality ; we know how
complete was the destruction even in England of the
greater ecclesiastical buildings ; much more would it
naturally be looked for among the smaller and ruder
structures of Wales. And how little the primaeval
Cathedral of Llandaif deserved to be excepted from this
class is clear from the minute description which is fortu¬
nately preserved of its most important dimensions.
" However," says the present Dean, " critical consi¬
derations may shake our full confidence in our monastic
historian, [the compiler of the Liber\ still his simple
description of the humble scale of the architectural relics
of that earlier age which his own eyes had inspected
bears the full stamp of exact accuracy. He informs us
that when Urban, the earliest bishop after the Norman
conquest of this district, was preparing to translate to his
own cathedral the relics of its canonised Prot-Episcopus
from the Isle of Bardsey, where he had been first buried,
having retired thither to religious solitude on the resig¬
nation of his ecclestical dignities, he found the primi¬
tive cathedral, founded by the saint, far too humble to
afford a suitable receptacle for his remains, for it con¬
sisted rather of a small chapel than a church, its length
46 LLANDAFF CATHEDRAL,
being only twenty-eight feet, its breath fifteen, and
height twenty. Two small aisles, however, are also
mentioned, (which, I suppose, should be added to these
dimensions,) as also a circular porch, (by which a semi¬
circular apse is probably meant,) having a radius of
twelve feet ; this would, therefore extend the entire
length to forty feet. This account will present an exact
analogy to those ecclesiological antiquaries who are
acquainted with the small dimensions of the earliest
churches of Ireland, ascribed to the age of St. Patrick,
and his immediate successors, who also flourished in the
fifth century, and who are said, like our Dubritius, to
have enjoyed the patronage of St. Germanus. We may
therefore conclude that these earliest founders of Chris¬
tian churches throughout the British Isles were contented
with edifices of the most humble pretensions."^
Such were the small beginnings of Llandaff ; but this
humble fabric may in one respect have equalled the most
gorgeous structures of succeeding ages ; it was doubtless
the best offering its founders could make. Equal in the
eye of piety, however inferior in that of art, to national
feeling it must have spoken in a manner which its
successor could never rival. It is a strange and not
altogether pleasant thought, and one in which Celt and
Saxon may singularly unite, that the greatest advances
1 Archœologia Cambrensis, No. I., (New Series,) p. 25. In one re¬
spect tMs little cathedral differed from its Irish fellows. The latter do
not employ the apse, though their arrangements are ritually identical
with the apsidal one. This, as Dr. Petrie has shown, points to their
derivation from ante-basilican models, from the earliest Christian
Churches of the days of persecution.
LLANDAFF CATHEDRAL.
47
in architectural skill and splendour in both countries alike
were the result of a foreign occupation, that not only the
castle, but the minster itself, were in fact, badges of
national misery and humiliation. And to no structure
of Wales does this remark more fully apply than to the
Cathedral of Llandaif.
But our only concern at present with the ante-
Norman days of Llandaif is to discover whether any
traces were left, or whether the subsequent design was
at all influenced by the arrangements or proportions
of the small British church which Bishop Urban, the
original founder of the present structure, found stand¬
ing in 1120. The Dean of Llandaif has expressed his
opinion that he built his cathedral westward of it,
leaving it to serve as a Lady chapel, and that the
great Norman arch at the east end of the presby¬
tery was made to open into it. If so, we have a fair
cha,nce of some part of its masonry still existing around
and above that arch. But in the first place, it seems
much inore in accordance with the general proceedings
of the Norman builders, even when they came into con¬
tact with structures of much greater pretensions than
this primaeval Cathedral of Llandaif, to suppose that they
would entirely destroy so small a building, and carry
out their own designs without reference to it. Secondly,
the existing arch appears far too large to have ever
opened into so small a structure as the British Cathedral
is said to have been ; even if it could have been contained
within its limits, it would most assuredly have violated
all proportion. If, then, it did not lead into the British
48
llandaff cathedral.
church retained as a Lady chapel, did it lead into an
original Norman Lady chapel ? or may not Bishop
Urban's choir have occupied the site of the present Lady
chapel, and consequently this arch have really been the
chancel-arch of his edifice ? This view was originally^
suggested to me by the eminent authority of Dr. Petrie,
and, though involving some difficulties, seems probable
on two grounds. First, a Lady chapel of the size and
prominence which such an arch seems to imply, is by no
means a common feature even in Norman churches of
very great size, much less in such small fabrics as Bishop
Urban's cathedral must, on any view, have been—one
far too small to have required a Lady chapel at all as a
distinct architectural feature. Secondly, one great arch
of this size, embracing the whole width of the church,
by no means resembles the ordinary approach to a Lady
chapel of any date ; we generally find the entrance made
by a greater number of smaller arches. On the other
hand, it has altogether the character and appearance of a
chancel arch. On the whole, then, the probability seems
to be in favour of the belief that Bishop Urban's choir,
of which no trace remains, occupied the site of the pre¬
sent Lady chapel, and that the fragments of Early Nor¬
man work, retained in the present presbytery, are portions
of his nave.
Extent of the Norman Church.—The original Nor-
1 Nothing however beyond this general notion was communicated
to me by the learned Doctor ; I may claim entire originality in work¬
ing out the details of the theory.
llandaff cathedral,
49
man cathedral, then, must have been a structure of com¬
paratively small size, though, as its remains attest, of a
very considerable degree of ornament. Its extent, east
and west, could probably not be ascertained without an
examination of the foundations. It is almost unneces¬
sary to state that the late Norman doorways in the pre¬
sent nave, which will come in for consideration hereafter,
throw no light on the extent of the early Norman church
of Urban. No one can imagine that his building ex¬
tended so far west ; while it is almost as difficult to sup¬
pose that his nave was confined within the limits of the
present presbytery.
Urbanas church probably consisted only of a nave and
choir, as we can hardly suppose that a tower or transepts
existed eastward of the present presbytery ; they would
surely have left some signs. It had probably no aisles.
Our evidence on this subject is derived from the appea¬
rance of the south wall of the presbytery—Urban's nave.
The extraordinary manner in which the Norman win¬
dows have been cut through by the Decorated pier-arches
has already been mentioned in the description of the
fabric. But there are some further questions arising out
of this portion of the building, which require to be dis¬
cussed at some length.
Tower-Porch ?—If there had been nothing connec¬
ted with these windows be5^ond what is visible from
the present presbytery we should have only inferred
from it that Urban's nave had no aisles, and remarked
the strange and unaccountable freak of the Decorated
h
50
LLANDAFF CATHEDRAL.
architect in his treatment of his predecessor's window.
A singular phœnomenon remains behind. The vaulted
bay, forming the vestibule of the chapter-house, opens
into the part of the aisle eastward of it by a rather low
arch, of which more anon. Over this, on the east side,
I remarked a seam in the masonry with a chamfer of
ashlar, as of the southern jamb of some opening, which
was not readily intelligible, though I ought to mention
that Mr. Prichard at once suggested that it was a squint
to the window in question. On removing the masonry
with which it was blocked, a long splay presented itself,
which finally led to the outer shaft of the window im¬
bedded in the wall, and revealing the original section.
The first inference would be that the vaulted bay, whose
existing features are Early English, or Transitional, was
added to the Norman work, and the masonry splayed off
to prevent the necessity of interfering with the window.
But the ashlar of the splay is part of the same stones as
the Norman shaft ; consequently this building, however
much disguised, is an integral portion of Urban's work.
It has been vaulted, as I before said, in Early English,
but its walls, at this point at least, must be essentially
Norman. I conclude that, at the Early English repair,
the greater part of this bay was internally cased with
ashlar, as all the decorative features are evidently of a
piece with the ashlar surface. A small extent of rubble
in the north wall may be a bit of Urban's work peeping
through. In the ground plan I have not marked Nor¬
man work, except in the north-eastern mass, as the only
part where we can be quite certain of its existence. The
LLANDAFF CATHEDRAL.
LLANDAFF CATHEDRAL.
51
core of the other walls is very probably of the same date,
but we cannot be quite sure, and all the visible features
are later.
If, then, we have here a further portion of Urban's
building, what was this structure, so curiously, I may
say, so awkwardly, attached to his nave ? To decide this
we are left to conjecture; but the most probable conjec¬
ture is that it was a tower, whose lowest story served as
a porch.^ And, if so, considering the general position of
side doorways, we may make a good guess at the extent
of Urban's nave, namely, that rit extended one bay west¬
ward of this porch, i.e., of the present arch into the pres¬
bytery. Whether such a tower was ever completed
seems very doubtful. Yet a small circumstance occurs
which at least seems to show that we have not the full
height of the Norman masonry in the east wall of the
vaulted bay. The ashlar of the splayed jamb terminates
suddenly, as if the wall had been altered at that point.
Either the Norman work was left unfinished in this part,
or else whatever was above it has been rebuilt. In either
case, we have not the full height of the building as first
designed. But as the masonry above seemed part and
parcel of that with which the opening was blocked, we
must incline to the latter belief, and conclude that some
upper part was destroyed,^ and a small portion rebuilt,
1 The bases of tlie jambs of a Norman doorway were found wbere a
new Early Englisb one now opens into tbe presbytery. They may
have been original, or they may have been no earlier than the Early
English or Transitional casing. I have not seen the fragments.
2 This later masonry cuts across the upper part of the 'window, as
may be seen by ascending the staircase of the chapter-house.
52
LLANDAFF CATHEDRAL.
when the window was blocked, though of course it may
only have been a small piece of wall, and not a completed
tower. It is clear that this blocking took place when
the Decorated arches were made; it joins with their
masonry, and fragments of the shafts and moulded
stones of the other side of the window, which could
hardly have been available except when they were re¬
moved to make way for the pier-arch, were used up
among the rubble with which the opening is blocked.^
If this was a tower, it is not improbable that its upper
stages were, or were intended to be, of an irregular
octagonal form, and that the splay we have been exa¬
mining was part of one of the smaller sides. There is,
indeed, no such appearance at the south-east corner, but
at that point there was no reason for bringing the chamfer
down so low, as was supplied by the window at the north¬
east.
It is difficult to believe the aperture of whose opening
an account has just been given, and the other Norman
arch to the east of it, to have been anything else than
original external windows. Yet it must not be concealed
that there appears no way by which the former can ever
have been glazed ; not only is the usual groove not to be
found, but there is no space left for it; all is splay.
Still it is easier to believe that the windows were glazed,
1 It is, however, an extraordinary fact that on the piece of wall thus
brought to light were manifest signs of whitewash^ a state of things
certainly repugnant to our ordinary notions of medias val proceedings.
Perhaps as the use of whitewash in Wales is now more extensive than
elsewhere, it may also have been of earlier introduction. I have
somewhere read of a Spanish church whitewashed about 1480.
llandaff cathedral.
53
or otherwise filled up, in some unusual manner, than that
they served any other purpose. It is almost impossible
to imagine that Bishop Urban's nave had aisles divided
from it by a wall perforated in this manner. Choirs^
indeed, are occasionally surrounded by a wall, con¬
tinuous, or nearly so, but even they do not present such
appearances as these. At the same time, even this view
would only be the greater of two not inconsiderable
difficulties ; for we shall soon find reason to believe that,
if this was not its original condition, it is one to which it
was actually brought by a subsequent alteration-
Supposed Clerestory.—And there is another question,
to which I should not myself have attached much impor¬
tance, had it not been supported by the opinion of the
Dean^ and Mr. Prichard. Both of them infer from the
plate in Dugdale, supported by the circumstance of many
fragments of shafts and other Norman fragments having
been found imbedded in the walls, that there was an
upper—a quasi-clerestorial—range of windows, if not a
regular triforium and clerestory. The plate does certainly
represent an upper range of round-headed windows ; and
though no one would, under ordinary circumstances,
build anything on the authority of representations in
which to represent pointed arches—especially if at all
obtuse or four-centred^—under the garb of round ones,
1 The Dean seems to take the existence of some sort of clerestory
for granted, without remarking on the singularity of the double range
of windows.
^ Sometimes indeed when they are quite the reverse ; any one who
was left to form his notion of the central tower at Hereford from
64
LLANDAFF CATHEDRAL.
is almost the general rule, yet in the present case they
do derive something like trustworthiness from having
represented the lancet windows in the nave clerestory
with tolerable accuracy/ Still I must confess that I
should require some much more unsuspected witness to
convince me of any point—especially one of so minute a
character as this—against which there rested any im¬
portant à priori objection. Now this view requires us to
admit one of two things, both so unusual and anomalous
that I can hardly believe either, unless it were supported
by some direct and trustworthy evidence. If the open¬
ings in the south wall of Urban's nave were genuine
windows—however glazed or otherwise closed—we then
have two ranges of windows over each other in the sides
of the aisleless Norman nave—a thing certainly not im¬
possible, as it is common in transepts, but hardly to be
credited without some stronger testimony than this. The
other alternative—that is, if we conceive the apertures to
have been originally blind windows into a quasi aisle^—•
is that of the ordinary double or triple elevation, with
the pier-range assuming the form of such a series of
fenestriform perforations looking into the aisle. Surely
to accept either of these alternatives we require some
stronger evidence than an old and inaccurate engraving.
Dugdale^s engraving would conceive it to be a specimen of rather
plain Romanesque.
1 Yet the same plate adds a row of buttresses to the aisle of the
presbytery, which one can hardly conceive having been destroyed.
2 If this was the case, I should be half inclined to accept the upper
range of windows, as otherwise Urban's nave would have had no
direct lateral light at all.
LLANDAFF CATHEDRAL.
55
And one of these we must admit on the south side ; on
the north, as the traces of Norman windows do not ap¬
pear, there may have been an aisle, with an ordinary
pier-range. But one certainly would not expect Urban's
church to have been thus lopsided, and as there is some
Norman masonry in the north wall—which is proved by
the fragments of the string discovered there—the Deco¬
rated arches must have been substituted for Norman ones
under a Norman clerestory,^ a possible process, but one
not to be lightly imagined. Further, if there were such
an upper range, the masonry of the walls must have been
nearly or quite as high as at present, so that either the
difference in height between the nave and the choir must
have been much greater than was usual in Norman
churches, or if we suppose Urban's choir to have been
higher than the present Lady chapel—no very probable
supposition—we should have an unusually enormous
blank space over the chancel-arch.^ And as for the
fragments of Norman detail found in the walls, even on
1 Pier-arches cut through blank walls, as at Cuddesden, Oxen, and
Iver, Bucks, are common enough—though it is not common to find,
as in the latter case, Norman arches cut through a Saxon wall—but I
do not at this moment remember an instance of pier-arches thus
substituted for earlier ones. There may, however, be such, as piers
inserted under earlier arches certainly occur, as in the choir of St.
Cross, and in Burton Latimer Church, Northamptonshire.
2 I believe, if we came to examine, we should find that the very
great blank space over Norman chancel arches of any width—I
do not mean such apertures as those in Gower—is usually to be attri¬
buted to a subsequent increase of height, either actually, by raising
the walls, as where a later clerestory has been added, or as far as
internal effect is concerned, by the removal of the original flat Nor¬
man ceiling.
56
LLANDAFF CATHEDRAL.
our view, the original west front, the original choir, the
windows in the north wall, would furnish a very good
supply of such.
We have thus made a fair guess at the general plan
of the first Norman Cathedral of Llandaif, a building
which, small as it was, perhaps as much surpassed its
British predecessor, as it is itself surpassed by the stately
fabric into which successive ages have developed it. If
our view of its plan and dimensions be correct, it must
have been altogether different from every other English
Cathedral ; however appearances may be explained, it
can hardly fail to have been unique in some point. And
yet, according to the notion of it already drawn out, it
might not, except in its diminutive size, have deviated
wider from the ordinary type than does its successor.
And I am persuaded that it is to the small size of Bishop
Urban's church that much of the singularity of the
present building is owing. For since his time, the
church has never been entirely rebuilt from the ground ;
subsequent changes, exactly as in the case of Dorchester
already alluded to, have consisted entirely in additions
made in every direction to the original diminutive fabric ;
Urban's cathedral was lengthened, widened, heightened,
but it still remained the germ and nucleus, around which
all the creations of later ages gradually arose. Under
these circumstances it was hardly possible for the church
to acquire the true cathedral outline. That can hardly be
obtained except by a direct design, embracing it from the
first, and such a general design for the whole church has
never been drawn out at Llandaff since the days of Urban.
LLANDAFF CATFIEDRAL. 57
I remarked above that, while a great analogy exists
between LlandaiF and Dorchester, the former approached
much more nearly to the genuine type of a minster ; that
though the cathedral outline is w^anting to the general
effect of the whole pile, the several parts taken alone are
conceived on much the same notion as those of other
great churches, while at Dorchester this is far from being
the case. The key to this diversity will, if I mistake not
be found in an important difference in the manner in
which the enlargements were effected in the two cases.
At Dorchester there is no reason to suppose that the
original arrangements of the Norman church have ever
been disturbed ; the nave has received the addition of a
single aisle, the choir that of one on each side, as well as
a great accession of length ; yet the nave still remained
the nave, and the choir the choir. But at LlandaiF we
have reason to suppose the case to have been in this
respect altogether different ; the most reasonable way of
explaining the appearances presented by the cathedral is
to consider that Urban's church was indeed preserved,
but not retained for its original use ; that his nave and
choir remained as the presbytery and Lady chapel re¬
spectively of the new building, while the main body of
the fabric, the nave and choir, were built independently
of it to the west. This extension westward is precisely
similar to what is usually supposed to have taken place
at Exeter ; where the present nave is built westward of
the Romanesque^ building which occupied the site of the
1 The Homanesque church at Exeter can hardly fail to have been
something well nigh, if not quite, as unique as LlandaiF itself. I shall
I
58 LLANDAFF CATHEDRAL.
present choir. It may also perhaps account for the
singular position of the west front with regard to the
steep descent reaching almost to its portal. Striking as
the effect is, it is one which is more likely to have been
developed incidentally than to have been intentionally
sought for by an architect ; and if the church has been
lengthened, if the original west front was therefore much
farther from the bottom of the hill, the former would
have been the case.
If then the early English nave and choir were built
independently, preserving Urban's church nearly intact
to the east of them, it is clear that nothing would hinder
the architect from producing a nave and choir, taken
alone, of genuine cathedral character ; while the presence
of the original small design, even as a mere appendage,
could hardly fail to have somewhat hampered his design,
and produced a confusion of ideas hardly compatible
with the production of a structure recalling, in its general
ground-plan and outline, the fulness of the usual cathe¬
dral pattern. We thus understand the omission of the
transepts ; had they existed, they must have been placed
either between the nave and the choir, or between the
choir and presbytery, that is, either across the centre of
the Early English work, or between it and the original
church. Now the length of the nave and choir is not
sufficiently great to have been crossed by a transept,
have presently to remark the—on any explanation—^unusual position
of the towers forming the present quasi-transepts ^ it is hard to con¬
ceive that a complete cross church with a third central tower ever
existed to the east of them.
LLANDAFF CATHEDRAL.
59
even without a central tower; with the present height
it would have been almost ridiculous. Nor could the
transepts have occupied the more eastern position with
much better effect. In that case Urban's church would
have formed the eastern limb, and it is clear that so
diminutive a structure would have had a much worse
appearance when thrust into a position of such unlocked
for prominence, than when retained simply as an adjunct
at the one end of an oblong structure. In the former
case its retention could only have been temporary, until
a more proportionate eastern limb—whether choir or
presbytery—could have been provided; yet everything
argues at all times, much more at this period, an inten¬
tion of retaining it permanently, so far at least as not to
involve a complete rebuilding from the ground.
I venture then to propose this theory as one which,
though it may sound somewhat startling at first, will
I trust, be found on the whole to explain the difficulties
and singularities of the cathedral more completely and
consistently than any other. If any one will suppose
that a Romanesque church ever existed, or was designed
to exist, occupying the full extent of the present
nave and choir, he has still to account for a presbytery
and something beyond it—to the east namely of the
great Romanesque arch^—unique as anything
which is involved in the view which I have just drawn
out. For, besides the question as to the fenestriform
apertures in the south wall of the presbytery, which
is an equal difficulty upon any view of the subject,
this view would suppose a Romanesque church of a
60
LLANDAFF CATHEDRAL
very extraordinary kind. On my view of Bishop
Urban's church, the only difficulty is the erection of
a cathedral church on a parochial type and scale. But
if the present presbytery were the presbytery of a
Romanesque building—whether ever completed or not
is of little consequence—^that building must have been
of a considerable size ; one can only conceive it to have
been designed as the eastern limb of a cross church, fol¬
lowing the common Norman arrangement with the choir
under the central tower. If so we should naturally have
expected the presbytery to terminate in an apse, instead
of which we have a large arch, like a chancel arch, and
something beyond, an arrangement of which I am not
aware of any instance early in the twelfth century. The
apsidal arrangement, we all know, was the prevailing
one in large churches even a good half century after
Urban's time, and the exceptional cases—all, be it ob¬
served, much later than the present—as St. Cross, Ox¬
ford Cathedral, fee,, give us simply a presbytery with a
flat end instead of an apse ; or, as at Romsey, a retrochoir
of utterly different character, and far less importance
than that which, on this supposition, we must imagine at
Llandaff.
I believe then that Urban built only a small, though
highly enriched church, consisting only of a nave and
chancel, with possibly a polygonal tower on one side ;
and that his nave survives in the present presbytery,
while the Lady chapel occupies the site of his chancel.
llandaff cathedral.
61
§ IL—Early English Additions.
Bishop Urban died in 1133,^ leaving, as I conceive, a
small edifice completed ; as appears to the Dean, a large
one unfinished. "We may conclude then," he con¬
tinues, •" that the presbytery had been completed before
the middle of the twelfth century—but this alone ; for
no other trace of *so early a date appears throughout our
whole fabric. The remainder of the choir and nave
would naturally have been the next portions taken in
hand." We agree then in holding that no structure
earlier than the present nave and choir ever occupied
their site ; the difference is whether they were a continu¬
ation—allowing for change of style—of Urban's'church,
or were erected altogether independently of it, treating
the latter much as the Dean supposes Urban to have
treated the original British building.
Date of the Nave. — If then Bishop Urban's
Cathedral was the small structure which there is
every reason to suppose it to have been, the changes
by which the church was brought to its present state
could not have commenced very long after its com¬
pletion. The enlargement of the building began while
Romanesque architecture was still not quite extinct,
and was concluded (for a time) in the earliest days of
the pure Lancet style. The west front, in which this
style appears in its perfection, the Dean attributes to
"the episcopate of William Saltmarsh, from 1185 to
^ Godwin.
62
LLANDAFF CATFIEDRAL.
1193, and the interior of the nave and the chapter-house
to his successor, Henry, elevated to the see from the
priorate of Abergavenny, and holding it till 1219."
Mr. Rees, in a note to his history of Abbey Cwmhir,^
seems inclined to place it still earlier, if indeed he
does not conceive it to be the genuine work of Bishop
Urban. But leaving this last view, which can be
paralleled with nothing except the notion that Coutances
Cathedral was built in the eleventh century, even the
more reasonable date assigned by the Dean is hardly
consistent with the facts of architectural history. No
recorded building of so early a date is anything like
so advanced as the west front of Llandaif. The Dean
refers to Canterbury choir ; but that structure, com¬
menced by a foreigner, William of Sens, is, not unnatu¬
rally, far more advanced than any contemporary build¬
ing in England. And even this bears no resemblance
to Llandaif ; Canterbury is the most decided possible
case of Transition or intermixture ; its Corinthianizing
columns, and ornaments half Romanesque, half Gothic,
have not the slightest similarity to the clustered piers and
pure Early English details of Llandaif. No English
building of 1180, or even 1190, can rank higher than
the Transition ; many are still decidedly Romanesque,
late, indeed, in character as well as date, but yet not
Gothic, or even Transition, but still Romanesque. Our
finest naves in that style, Peterborough and Ely, were
actually in progress at the time that we are told that a
pure Lancet structure was being executed at Llandaif.
1 Archœologia Camhrensis, Vol. lY., p. 247.
LLANDAFF CilTHEDRAL.
63
The earliest genuine Lancet work known is certainly to
be found in the low eastern aisles at Winchester, com¬
menced about 1202 by Bishop de Lucy. And even these
are in advance of their age, as much later work is often
not clear of Romanesque. The nave of Wells, com¬
pleted in 1239,Ms Early English, but by no means free
from traces of the preceding style. And to come nearer
home, and to an example less advanced than Wells, the
choir and transepts of St. David's Cathedral, which appear
to have been rebuilt after the fall of the tower in 1220,^
though they agree in some points with the work at
Llandaif, and have shafts of the very same keel form to
which the Dean has called attention, are yet decidedly far
less advanced, and from the Romanesque details still
employed in their capitals and arch mouldings, can only
be considered as Transitional. And though nothing is
more common than to find architectural forms in use at
a period later than their ordinary date, as the work at
St. David's just mentioned, yet to assign them to one
earlier ought not to be done without most cogent de¬
monstration. To suppose the former is only to suppose
an old-fashioned taste in a particular architect, while the
latter involves a revolution in the received ideas of the
history of architecture. And though the fact that work
of any particular kind in Wales is usually, from the
length of time which innovations took to find their way
^ Mr. Rees, in the article on " Cwmhir Abbey," antedates this
nearly a century, assigning it to " Bishop Robert, who held the see
from 1139 to 1166." Cwmhir itself may have heen founded in 1143,
but the arches he describes must be nearly a hundred years later.
^ Some parts are as late as 1248.
64
llandaff cathedral.
into SO remote a country, considerably later than similar
work in England, would not apply in its full force to
cathedral churches, still Llandaff is not exactly the place
in which one would look for architectural developments
so far outstripping those of all contemporary buildings.
We may then most probably attribute the Lancet
work of the nave and choir, both the west front and
the arcades, to a date not earlier than 1220, which seems
to agree very well with its character, which exhibits
the style when it has just worked itself free from
Romanesque elements, and yet has not attained the same
fulness and freedom which we see at Ely, or in St.
Mary's at Haverfordwest.
External Walls.—Our next question, then, is as to
the two doorways, north and south of the nave, which
would appear to point to operations going on at some
period intermediate between the time of Bishop Urban
and the date just fixed. These are quite late Norman ;
the southern one, indeed^ contains no detail actually in¬
consistent with the purity of that style ; but even a
slight comparison of it with the early Norman work in
the eastern parts will show that architecture had made
no inconsiderable advance in the interval between the
two. On the north side the case is yet far stronger ;
there we have in the label the genuine tooth-moulding
of the Early English, though certainly not in at all an
advanced form. Yet a doorway containing such a feature
must be called at least Transitional, and there seems no
reason to doubt but that this doorway and its fellow are
LLANDAFF CATHEDRAL.
65
of the same date. Now we must remember how very
long the use of the Romanesque doorway, not only with
the round arch, but often with actual Norman detail,
prevailed, especially in some districts,^ even when in
other respects the Early English style was fully confirmed.
Now I must confess that, from my experience of not a
few similar examples, I should not think it altogether
monstrous, if the evidence looked at all that vmy, to re¬
gard these doorways as actually part and parcel of the
Lancet work. Still one would not suppose this without
some cogent reason, and in the present case the evidence
is at least doubtful. For the doorway, which is certainly
part of the Lancet work, namely the portal in the west
front, though it retains the round arch, has quite cast off
all Romanesque detail. We may therefore fairly sup¬
pose these doorways to be a little earlier than that. Not
that we need imagine any erection to have taken place
in this part of the church after Bishop Urban, which was
destroyed to make room for the Early English nave ; all
that we need suppose is, that the aisle walls were built
before the arcades, and the stoppages which often took
place, or even the mere slowness with which such great
works were carried on, will allow us ample time to ac¬
count for the slight advance of style between them.
In the eastern part of the church we have a small
portion of work which is probably contemporary with
1 Especially in Nortliamptonsliire, a county several of wliose
localisms, characteristic, though, of course, not absolutely distinctive,
have, somehow or other, found their way to Llandaif. I may refer
to a paper in the Ecclesiologist, No. LXXXVI., p. 289, and to my
History of Architecture,'' p. 324.
»
K
66
llandaff cathedral.
these doorways ; this is the vaulted bay leading to the
chapter-house, already mentioned as being probably
originally the base of a Norman tower. The work here,
both in the vaulting-ribs and two of the capitals, (two
being pure Early English,) retains a considerable Ro¬
manesque tinge, and is clearly earlier^ than the Early
English of the nave. One can hardly doubt but that
it assumed this form with intention to be made the
entrance to the chapter-house, though that building, as
it now stands, probably followed the erection of the other
Lancet work in the church.
The arrangements of the church were thus completed,
as to the interior of the fabric, much as they still remain,
the nave and choir being added to the west of Bishop
Urban's work, (except probably one bay of the latter
taken into the choir,) while his nave became a presbytery,
and his choir a Lady chapel.
Eastern Towers.—One very remarkable circumstance
connected with this extension remains to be noticed,
which, if my view be correct, shows that its designers
contemplated a very different external outline of the
1 On further inspection, I ana less clear of this. The western arch
of the vaulted bay is, in its masonry, clearly of a piece with the
vaulting, while, in its architecture, it does not differ from that of the
choir and nave. I must again remark that it is safer to attribute the
Transitional work to an unusually late period than the pure Lancet to
an unusually early one, and it is no great marvel to find contem¬
porary work, even in the same building, widely differing in character,
especially about this period. But this need not affect the doorways ;
that part of the aisle walls may well have been commenced before
this bay was vaulted and cased.
LLANDAFF CATHEDRAL.
67
cathedral from what at present exists. I allude to some
singular appearances in the eastern bay of the aisle on
each side of the ritual choir. This pair of arches are
continued of the same height as the other members of the
arcades, but a considerable portion of their height is
blocked, and a much smaller arch, but of the same date,
formed beneath. In the choir itself no reason appears
for this deviation from the general design, but on enter¬
ing the aisle the cause is conspicuous enough ; the in¬
tent was, by strengthening the piers, to find abutment for
a large arch thrown across each aisle from this point.
The arches themselves do not exist, but their spring may
be very clearly seen against the flat inner surface of the
piers at the points marked c in the ground plan. Un¬
fortunately no traces are visible against the aisle walls,
as these have been rebuilt from the window-cill at a later
period ; and, as the arch died into the wall, nothing can
be looked for in the way of bases of shafts. In fact there
can be little doubt but that the arches were destroyed
when the reconstruction of the walls took place.
But what was the end and object of these arches ? I
strongly incline to believe that they are the western
arches of a pair of towers, the original Romanesque
tower-porch, if such it really was, being destroyed or
left unfinished, and its base converted into a passage to
the chapter-house. On the south side the eastern arch
also is distinctly visible, a strong arch of construction
thrown over the lower one leading into the vaulted bay.
On the north side the springing of the eastern arch
cannot be traced in the same manner as in the other
68
LLANDAFF CATHEDRAL.
places ; but there is an extreme roughness of m^asonry at
the point whence it would have risen, so that it may be
merely that it has been more effectually destroyed than
its fellows. The Early English design then embraced
two steeples flanking the choir, while two larger ones
flanked the west front. Regarding the choir as, what
architecturally it is, a portion of the nave, and the pres¬
bytery as the architectural choir, the position of these
towers would have been exactly similar to those at
Exeter, except that the latter have the aisle between
them and the main body of the church, while these at
Llandaff must have risen out of the aisles. The diffe¬
rence, in fact, is just analogous to the different treatment
of v/est fronts ; the Llandaff arrangement answers to the
common west front with towers terminating the aisles,
while that of Exeter recalls the fronts of Rouen and
Wells, where the towers are built beyond the line of the
aisles.^
The towers thus placed must have served very much
to break up the flat and heavy outline of the church,
and must have imparted a good deal of German character
to it. But on any circumstances of shape, design, or
finish, it is in vain to hazard conjectures; it is indeed very
doubtful whether they were ever finished at all, and, if
so, they were doomed to destruction in the course of the
century after that in which they arose,^
1 Mot only recalls, but is actually tlie same, if it be true that the
Exeter towers are the western ones of a Romanesque church which
has been extended westward.
2 It has been suggested to me that these arches were intended to
llandaff cathedralc
69
South Aisle of Presbytery.—To return to the aisles,
there seems reason to believe that some alteration took
place in the eastern part at some stage of this extensive
repair, of which hardly any traces remain. It is by
no means impossible but that aisles were added to the
presbytery, as suggested above, but at present without
disturbing the original Romanesque walls. The evi¬
dence on which this supposition rests is the fact that
the eastern arch of the vaulted bay is clearly part of
the Early English work, and as it must have opened
into something, some building must have been added
to its eastern face at this, or an earlier period. But
no trace of it remains, and its outer walls must have
been completely rebuilt afterwards, as the rubble wall
of the aisle immediately adjoining has no connexion
with the excellent ashlar from which the arch springs,
lead, not into towers, but into short transepts. This is hardly possible ;
one might imagine transepts the full height of the choir and yet not
projecting beyond the aisles, as at St. John's Church, Coventry, or
again transepts no higher than the aisles, but projecting beyond them,
as in very many paiash churches. But here the arcade shows that
they were not the full height of the choir, and the chapter-house that
they did not project beyond the aisles ; for in that case the south
transept would have left some traces against its southern wall. But a
transept which exceeds the aisle neither in ground plan nor in eleva¬
tion is no transept at all 5 it is at best what is sometimes called a false
transept. But even in this case, one cannot account for the arches
between the choir and these bays being lower than the rest ; if any¬
thing, one would naturally have expected them to be higher 5 whereas
this means of providing a more massive pier by diminishing the span
of the arch is just what we continually find in the case of engaged
towers, as indeed we find in a smaller degree in the western tower of
this very church.
70
LLANDAFF CATHEDRAL.
and is divided from it by the widest fissure in the whole
building.^
If then we are to assign a conjectural date to this ex¬
tensive enlargement and remodelling of the building, I
should consider that Bishop Urban's church remained
undisturbed for about sixty years after his decease ; and
that the nave and choir were commenced by Bishop
Henry of Abergavenny. As he is stated by Godwin to
have founded fourteen prebends in his church, we may
reasonably assign to his personal liberality, and to the
zeaP with which we may hope his new foundation
entered upon their duties, a benefaction to the material
fabric whose architectural character corresponds very
well with their date. The outer walls, as fixed by the
doorways, would suit the date of 1193 very well; and
the remainder of the nave and choir may easily have
1 A point of some difficulty is to be found in the existence of a
'pointed arch, blocked, immediately above the western face of this arch,
looking at first sight exactly like that of a window destroyed when the
arch was made. A round arch would be intelligible enough, but it is
hard to conceive any Pointed work being destroyed to make way for
this, which itself is not clear of Romanesque. The arch does not go
through the wall, and has no splay ; from this latter circumstance Mr.
Prichard inclines to the belief—which, curious as it is, is by far the
lesser crux of the two—that it was merely an arch of construction.
If so, its insertion over the excellent piece of masonry below looks as
if it had some greater weight to support than at present, and as if its
designers had found, and perhaps even intended to preserve, Bishop
Urban's tower complete. If so, however, the design must have
changed very soon, as its retention would hardly be consistent with
the existence of the eastern pair of towers.
2 Why should not this new foundation in the twelfth or thirteenth
century have had results similar and as fortunate as those consequent
on the re-establishment of the Deanery in the nineteenth ?
llandaff cathedral.
71
been continued through the latest years of his own
episcopate and through that of his successor, William,
Prior of Gouldcliffe, who died in 1229.
Chapter-house.—Very soon after the completion of
the nave and choir, the chapter-house must have been
added ; its architecture may be considered a little more
advanced, as its Lancet windows have foliated heads.
We thus have the cathedral temporarily finished a second
time, about a century after its first completion by Bishop
Urban.
§ III.—Decorated Repairs.
The Lady Chapel.—The whole interior of the church,
from the west doorway to the chancel-arch, was thus
brought to its present condition, saving the alterations
effected by the Bath reformer of the last century. Of
the external walls of the aisles we cannot speak with
certainty, as hardly any trace of their original state
remains, but the whole internal elevation remains un¬
altered, with the important exception just made. But
all this time Bishop Urban's church existed to the east of
the splendid fabric which had supplanted it, its nave
serving as a presbytery, and its choir as a Lady chapel.
The next great work was the rebuilding of the latter part
of the church in a more stately form, to which we owe
the present very beautiful specimen of early Geometrical
architecture, whose character agrees very well with the
supposition that Bishop de Bruce was its founder. This
72
llandaff cathedral.
prelate, who is buried in the Lady chapel, filled the
throne of Llandaif from 1265 to 1287. His building
was a complete erection from the ground, and retains
no trace of Romanesque work, except the grand arch
opening into it from the presbytery, which shows that
the wall between Urban's nave and choir was preserved
intact, while the latter was completely destroyed to
make room for the present chapel. It has also been
subjected to no alteration in any subsequent style.
We thus have architectural works going on in this
cathedral almost uninterruptedly through the whole of
the thirteenth century ; and we consequently find an
excellent study of the Early English style, as traced, in
a series of pure though plain examples, from its first
development out of Romanesque, as seen in the passage
to the chapter-house, to its gradual sinking into Deco¬
rated, as exhibited in the chapel we are now considering.
Aisles.—Immediately on the completion of the Lady
chapel appears to have commenced that extensive repair
which has brought the presbytery to its present form,
and reconstructed the aisles throughout the church. This
appears to have been done from one uniform design, com¬
mencing eastward, but so slowly carried out as to present
great changes of detail in different parts. We shall see
that some parts of the aisles were now built from the
ground, while some contain portions of earlier work;
but throughout it practically amounts to an entire re¬
construction, as no architectural feature of the earlier
building has been allowed to remain, except the two
llandaff cathedral. 73
Romanesque doorways in the nave aisles. The general
appearance is that of complete Decorated aisles ; only, as
the lower part of the walls is in some parts original, we
may conclude that in the great flatness and want of
buttresses which disfigures all this part of the cathedral,
the Decorated architect did but reproduce the errors of
his predecessor. Internally, as I before said, we owe to
this reconstruction of the aisle walls their absence of any
testimony as to the appearances which I have supposed
to indicate the bases of towers. If these were ever com¬
pleted, they must have been destroyed at this time.
Aisles of Presbytery.—The reparation which we are
at present considering began at the extreme east end of
the aisles, and followed so immediately on the completion
of the Lady chapel that its earliest portions are actually
part and parcel of the same work. The east wall of the
aisles is continued from that of the chapel, the buttresses
at the south and north-east angles are of similar design,
and the same Early English string runs along the east
end of both aisles, and, on the south side at least, under
the most eastward of its windows.^ The extent of this
first portion of the Decorated work is probably marked
on the north side by a singular break in the wall, like
an enormously wide pilaster sloping backwards and
dying into the wall. As this is in a line with the piers
^ The string here is modern, but a restoration of the original one ;
to the north there is at present a Decorated string, but I believe con¬
jectural, the old one not having been preserved. I feel sure that an
Early English one would have been the true restoration.
l
74
llandaff cathedral.
of Bishop Urban's chancel arch, we may probably con¬
clude, though the wall there does not appear to afford
any evidence, that it extended to the same point of the
south aisle also, taking in one bay, namely that attached
to the two western bays of the Lady chapel.
The details of this first portion are by no means clear
of Early English ; besides the strings, we have Early
English corbels in the corners of the north aisle ; and
the window-jambs, though under Decorated labels, be¬
long rather to the earlier style. In the south aisle they
have a shaft with a broach above the abacus, in the north
a continuous roll with a shaft, but no capital. It will
be remembered that the arches into the Lady chapel on
each side partake of the same mixed character.
The next portion embraces the two bays forming the
north aisle of the presbytery ; here the architecture is
decidedly Decorated ; the two window-jambs are merely
moulded without shafts. In the south aisle we have no
work of this portion, probably because of the building
added to the vaulted bay being still preserved. The
only window here is later.
Presbytery.—But contemporary with these two bays,
we have a still more important change, no other than
the transformation of Urban's Romanesque nave, which
up to this point must have remained, internally at least,
without any alteration of importance, into the present
Decorated Presbytery. The existing arches were now
cut through the Norman walls. On the north side, as
we have already stated, this was effectually done, two
llandaff cathedral.
75
arches being completely formed, and no trace of the
older work, beyond the portion of a string already men¬
tioned, allowed to remain. On the south side we find
the extraordinary appearances which have been already
commented upon at length.
Again to attempt a conjectural date, which indeed is
little more than to write down the name of a Bishop, I
should conclude that the Lady chapel and these two
earlier portions of the aisles were carried on, slowly
probably, but without much actual interruption, during
the episcopates of William de Bruce already mentioned,
and of his successor John de Monmouth, who occupied
the see from 1296 to 1323 ; there being, if Godwin's
computation be correct, an interval of eleven years
between the death of Bruce and the consecration of
Monmouth.
Aisles of Choir and Nave.—The third period of the
Decorated repair brought with it the remodelling of the
aisles of the Early English choir and nave in harmony
with the recently erected aisles of the presbytery.
Throughout the greater portion of their extent the walls
were, as has been already hinted, rebuilt from the ground ;
but there are the following exceptions : First, The door¬
ways in the nave. Secondly, A small piece of masonry
continued from that of the vaulted bay in the south-east
angle of the south choir aisle ; we have here the basement
of the south-east tower ; the wall being naturally some¬
what thicker. But the wall, except a very small portion
immediately adjoining the arch into the vaulted bay, has
76
LLANDAFF CATHEDRAL
been rebuilt at least from the cill. Thirdly, On the north
side the two eastern bays were probably only rebuilt
from the cill, as there is a break in the masonry just east
of the small north door, and a marked difference in the
basement. I may remark that this north wall, thus
built at three or four different periods, presents much
irregularity in its direction ; at the point of junction
between the choir and presbytery it is especially remark¬
able.
Something was also effected about this time in the
south aisle of the presbytery, as is proved by the existing
window, which has beneath it a Decorated string. But
it is not clear whether the whole wall was built at the
first repair, and this window only inserted in it, or
whether the western part of the wall was built now. At
all events the present wall, as the great seam shows,
belongs to some period of the Decorated repair.
With the exception of this window, all the others of
this date are uniform ; the jambs have octagonal shafts
and capitals, and the tracery consists of the monotonous
Reticulated form under ogee heads. These are a localism
of Northamptonshire, and we find a third assimilation,
whether they be accidental or otherwise, to the architec¬
ture of the same distant county, in the single exception,
the broad window with a flat head.
"The reredos behind the high altar, consisting of a
double row of arched panels, flanked by two elegant
side arches of entrance to the space behind, of which the
cuspidated moulding is singularly light, being so much
undercut as slightly to detach it from the upper mould-
llandaff cathedral.
77
ings with which it is connected, appear, from the general
character of their execution, to belong rather to the Later
Decorated than to the Perpendicular style, though by
Browne Willis (in whose time it was surmounted by a
third tier of niches) referred to Bishop Marshall, in the
reign of Henry VII. His grounds for this opinion were
the occurrence of roses, the devices of the Tudor family,
emblazoned on the panels ; but it is easy to suppose
that these, and other decorations, may have been added
by that prelate to an earlier structure, as we have it on
record that he had been engaged in the general embel¬
lishment of the choir."^
Perpendicular.—T he N orth - west T o wer. —These
changes brought the whole church, within and without,
in all its most important particulars, to its present con¬
dition, or at least to its condition previous to the exploits
of Mr. Wood. All the peculiarities in plan and outline,
which render it so remarkable, had now been brought to
perfection. The Perpendicular period, though introduc¬
ing one of the most beautiful individual features of the
building, was not marked by any alteration at all affect¬
ing its general character and proportion. Besides the
insignificant insertion of tracery in the eastern windows
of the aisles, all that remains, all probably that ever
existed, of Perpendicular architecture in Llandaff Cathe¬
dral is the very fine north-west tower. This however
1 Dean Conybeare in Archœologia Camhrensis, ut supra, p. 35.
Bishop Marshall was consecrated in 1479, translated to London in
1489, and died 1493.
78
LLANDAFF CATHEDRAL.
was not built from the ground, as in the lower part, both
inside and out, considerable traces of its Early English
predecessor exist. It stiíl rests upon the original arches
both to the east and the south ; in the former Perpendi¬
cular corbels have been inserted in a singular manner.
Part of the west wall, as indeed we have already seen,
including a large portion of the internal jamb of the aisle
window, is a relic of the earlier tower.
This tower may be regarded as an exception to the
statement made above that the architectural splendours
of the two great Welsh cathedrals are to be regarded as
something exotic, and not calculated to excite any dis¬
tinctively Cambrian patriotism. For it has always been
attributed to one who was not only a native Welshman,
but in whom and whose family the blood of the ancient
British princes—if that lofty extraction be a real genea¬
logical reality—was brought into close alliance with the
royal houses of England and France, with the blood of
Alfred and St. Lewis. It was erected by Jasper Tudor,
Earl of Pembroke, and afterwards Duke of Bedford, a
son of the romantic marriage between Queen Catharine of
France, the widow of Henry V., and the simple Welsh
gentleman Owen Tudor, and consequently paternal uncle
to Henry VH.
This addition, or rather reconstruction, as being the
simple re-building, as far as difference of style allowed,
on the same plan, of a previously existing feature of the
church, must be regarded in itself rather as an incident
than as a prominent event in the architectural history of
Llandaff. Yet, as the last event till within the last few
llandaff cathedral.
79
years, which can be contemplated with any degree of
satisfaction, being the last change in the fabric made
during the prosperous days of English architecture, and
coming immediately before those storms, real and figu¬
rative, which were destined, in the three following
centuries, to beat with such violence upon the hapless
church of Llandaif, the historian may turn with regret
from this conclusion of the most interesting, as well as
the most difficult portion of his undertaking.
§ IV.—Decay and Restoration of the Cathedral.
It will not be at all required by my design to enter
into any minute history of the steps by which the
church which thus took between three and four cen¬
turies to bring to its final completion, was gradually
abandoned to a state of, in England at least and even
in Wales, unparalleled neglect and ruin. It forms no
necessary part of an architectural history, and is ren¬
dered unnecessary by the copious account given by
the Dean, to which I could add no further information.
Yet a few remarks on some of the more striking points
suggested by it, as well as on the great work on
which that venerable dignitary is himself so zealously
engaged, will form an appropriate conclusion to our
present inquiry.
Of the sixty-third prelate of Llandaff, his successor
Godwin gives the following account, his feelings, one
would think, being heightened by a prophetic perception
that he was himself one day to suffer by the corruption
80
LLANDAFF CATHEDRAL.
he records.^ " Anthony Kitchen, alias Dunstan, doctor
of Divinity, and sometimes Abbot of Eynesham, was
consecrated May 3. 1545. and enduring all the tempes¬
tuous changes that happened in the meane time, con¬
tinued till the 5. yeere of her Maiestie that now reigneth ;
viz. the yeere 1563. and then died, hauing first so im-
pouerished the Bishopricke by unreasonable demises, of
whatsoever was demisable, as there was no great cause he
should be so loth to leave it." He was the only prelate
who consented to take the oath of supremacy to Queen
Elizabeth, and one wonders that it was not he, but Owen
Oglethorpe of Carlisle, who had the honour of placing
the crown on the head of the " Maiden Queen."
But though Bishop Kitchen certainly exhibits the
famous Vicar of Bray magnified to episcopal proportions,
I am not sure whether somewhat hard measure has not
been dealt him when he is arraigned as " the great first
cause of these evils." ^ We must remember that from
his age dates the severance of bishops from their cathe¬
dral churches ; henceforth whatever is done or left un¬
done, whatever is preserved, destroyed, or repaired, it is
frequently the exclusive work of the chapter, and it is
very seldom that a post-Reformation Bishop has stepped
forward, like Hacket of Lichfield, in the character of
Poore or Wykeham, as the prime mover in a work of
1 At the time when he published the "Catalogue of English
Bishops/^ Godwin, afterwards Bishop of Llandaff and finally of
Hereford, describes himself on the title-page as " Siibdeane of Exceter,''
and it appears from many passages in his work that he was also a
member of the church of Wells.
2 Archœologia Camhrensis, ut supra, p. 36.
LLANDAFF CATHEDRAL
81
this kind. It does not appear that Kitchen directly con¬
tributed anything to the actual destruction of the cathe¬
dral, and the revenues which he alienated, rapaciously
and sacrilegiously indeed, would probably have no longer
been, in any case, regarded as a fund for its perpetual
reparation. Moreover in these alienations, he did no
more, though he probably carried the work to a greater
extent, than many bishops who have come down to
posterity with much better reputation. In adapting
himself to the times he only followed the course of many
both of the clergy and laity, including among the latter
some of the most munificent of Oxford's later benefactors.
As to the oath of supremacy, our only wonder is that so
many prelates and others^ refused it under Elizabeth,
who had submitted to the more extensive claims of her
father. And it may have been owing to some vestiges
of conscience that Bishop Kitchen neither crowned
Elizabeth, nor took any part in the consecration of
Archbishop Parker, although he was twice named in a
commission for that purpose, and on the second occasion
was actually the only bishop in the kingdom in regular
possession of a see.^
In fact the desolation of Llandaff is simply the greatest
1 The explanation is doubtless to be found in a personal feeling.
It was known that Elizabeth was prepared to exercise her ecclesiastical
supremacy in a manner which Henry had not done ; while there can
be little doubt that the system represented by him and Bishop Gardiner,
a church namely strictly papal in doctrine and ceremony, yet entirely
independent of Rome, had, till the cruelties of Mary's reign disgusted
the nation with all approximation to the old religion, a decided majority
among the people.
2 See Lingard, " History of England," vi., 668.
M
82
LLANDAFF CATHEDRAL.
instance of what befell nearly all our cathedral and other
churches in a greater or less degree. Every cathedral
suffered more or less ; the Welsh more than the English ;
Llandaif more than the other Welsh ones. And, if we
take in churches not strictly cathedral, Llandaif at once
ceases to be the worst example, as the collegiate church
at Brecon still remains in a yet more desolate condition ;
there divine service has been discontinued altogether,
while Llandaif has, I believe, always retained the ordinary
portion of a parish church. Many circumstances, as we
all know, concurred to produce that general apathy with
regard to our ecclesiastical structures which contrasts so
strangely and painfully with the feelings of preceding
times. Most of them are obvious enough ; the change
in religious belief took away many of the motives which
had led to the erection and decoration of churches ; the
extraordinary revulsion of taste in matters of art greatly
diminished the admiration with which the existing build¬
ings were regarded ; people were less likely to preserve,
repair, or embellish structures which they were taught
to look upon as barbarous and void of artistic merit.
Proceedings, partly of legitimate policy, partly of wanton
sacrilege, had greatly diminished the revenues available
for such purposes ; while the permission of marriage to
the clergy suggested quite different ways of employing
what remained. And, in the case of the cathedrals, I
am inclined to think that the severance between them
and their bishops, already alluded to, gave the ßnishing
stroke to their neglect. A corporation cannot be so
easily nloved as an individual, its many members cannot
LLANDAFF CATHEDRAL.
83
feel the point of honour so acutely ; differences of opinion
may unavoidably lead to compromises and half-measures.
Hence, while the mere vis inertiœ makes the worst cor¬
poration better—as far as preservation at least—than thé
worst man, not only is the best man far better than the
best corporation, but I am inclined to think the average
is better than the average. A resident bishop, really
recognized as the head of the cathedral, and feeling his
honour at stake in its condition, must in the nature of
things have been a better guardian and preserver than a
body, ojDen to the general defects of corporations, arid be¬
sides consisting chiefly of persons only resident for short
intervals, and in most cases not recognizing the cathe¬
dral as their home, but having the interests most closely
touching them elsewhere. And to the peculiar circum¬
stances of Llandaff all this would apply with increased
force ; elsewhere there were generally many external in¬
fluences ; in most places the cathedral was at least prized
by the people, it was the pride and ornament of the city,
and the inhabitants of both the city and county felt their
honour concerned in its preservation. But in the remote
situation of Llandaff—as of St. David's also—its chapter
could hardly have been more amenable to influences of
this sort than the rector of any country village. And
possibly the peculiar constitution of the chapter may
have produced an additional unfavourable influence ; the
absence of a head must ever have been bad ; the dean of
a cathedral has a direct personal stake and responsibility
which he cannot so readily divide with others ; a really
resident dean fills to a great extent, as far as the fabric is
84
LLANDAFF CATHEDRAL
concerned, the position of a mediseval bishop. We may
surely venture to infer how much Llandaif lost from the
cessation of the decanal ofíice from what it has gained
by its restoration. Much doubtless is always owing to
personal qualifications, and I need not say how fortunate
this church has been in both the deans who have hitherto
presided over it ; I only mean that a resident personal
head, if a man of any honour and conscience at all, could
not but feel a sense of responsibility, which would have
been much fainter, if scattered over the consciences of
eleven colleagues. Numbers may be a useful check, but
it is personal energy which must supply the motive
power. Then the constitution of the body of prebendaries
was singular, and, as I should apprehend, not conducive
to united action. There does not appear to have been the
distinction usual in all the elder foundations, between resi¬
dentiary canons and simple prebendaries. Twelve canons,
with equal obligations to residence, would, according to the
ordinary system of only one resident at a time, give only a
month's annual residence to each, a period not sufficient
to create any real interest in the place in a mind the least
disposed to neglect its duties, What wonder then, if,
with no public opinion to bear upon them, and with
bishops even more completely severed from the cathedral
than elsewhere, so short a term of residence shrunk into
no residence at all, if all care and interest in the place
ceased, and the cathedral sank into a worse condition
than an ordinary parish church, inasmuch as the only
resident persons connected with it, the only ones who
could be reasonably expected to feel any love for the
LLANDAFF CATHEDRAL.
85
place or the building, were invested with no sort of
authority to do aught to arrest its destruction ? ^
Under all these influences, residence ceased, the choral
service was intermitted, the fabric was suffered to fall to
ruin, piece by piece, without any attempt to arrest or to
remedy the progress of destruction. Then came the œra
of Mr. Wood of Bath, the œra of plaster, whitewash,
urns, and conventicle windows. By his agency, about
the middle of the last century, the cathedral was finally
reduced to that state from which it is now just beginning
to revive. The greater part of the nave—by that time
probably the discovery had been made that a nave was a
useless appendage to a great church—appears to have
been now finally unroofed and forsaken ; the choir was
" beautified" in the taste of the times, while a small
portion,^ a sort of ante-chapel, was reserved, formed by
the eastern bay of the nave, and the site of the rood-loft ;
this was separated from the part of the church abandoned
1 Much of what has been here said of Llandaif applies equally to
Brecon, with its nominal dean, and body of prebendaries not dis¬
similarly constituted to those of Llandaif. But it is clear that at
Llandaif its position as the only church in the parish must always have
prevented a total cessation of divine service ; while at Brecon the
collegiate church is not the only one in the town. Somewhat similarly
all residence and most other features of a collegiate body died out,
though very gradually, at Wolverhampton, while its position as the
parish church of a large town insured the preservation of the fabric
and service.
2 Somewhat similarly, in Bristol Cathedral, where the real nave
has been destroyed, and in Merton College Chapel, where it has
never been completed, a sort of ante-chapel or quasi-nave is formed
by the crossing, and one or two of the western bays of the choir, the
screen being thrust eastward of its natural position.
86
LLANDAFF CATHEDRAL.
to entire ruin by the new west front. The choir was
completely boxed in and plastered, the arches on each
side being filled up, while those in what may be called
the ante-chapel were spared that degradation ; though
at the same time the opposition between the arches them¬
selves and the clerestory they supported must have been
thereby rendered still more conspicuous. If one can
imagine any notion of arrangement or precedent entering
the minds of the perpetrators of such destruction, one
might almost fancy they intended to recall the appear¬
ance of such college chapels as New College, All Souls,
and Magdalen. The beautiful range of lancets in the
old clerestory gave way to the substitutes which still
remain, for no long time we may trust, and numerous
other windows of the same character were inserted in the
ancient jambs in other parts of the church.
In these changes we have the crowning instance of the
utter lack of architectural taste which was so character¬
istic of the last century. It is sufficiently strange that
Italian architecture should ever have been preferred to
Gothic ; a style utterly confused and inconsistent as a
mere work of art to one which is the very perfection of
mechanical and œsthetical science ; a style alien to our
national and religious associations to one which is the
chosen offspring of our own creed and our own race.
Yet the preference of one style to another is after all
intelligible, and we must allow much to that sound
maxim. De gustihus nil disputandum. One can under¬
stand well enough that architects who were capable of
designing such structures as St. Paul's Cathedral and
LLANDAFF CATHEDRAL
87
Queen's College Hall and Chapel should have preferred
following out their own ideas to imitating the buildings
previously occupying their site ; the new buildings had
so much real merit that a little allowance for fashion and
self-satisfaction might well make them appear actually
superior to the old. And these were no cases of patch¬
ing, there was no room for incongruity ; the previous
buildings were completely destroyed, so that the ground
was practically as new for them as for the grandest of all
English Italian designs, the Radcliffe Library. Even
when Inigo Jones attached his portico to old St. Paul's,
though he is open to the charge not only of preferring
the worse style of art to the better, but of neglecting the
just respect due to an existing building p still he pro¬
duced an erection not altogether devoid of architectural
merit in itself, however incongruous and out of place.
To understand the lowest depth, to see what art and
taste could come to, it is necessary to undertake a jour¬
ney to Llandaif. Mr. Wood's performance was not a
preference of one style to another, but the deliberate
substitution of ugliness for beauty, yet more, the osten¬
tatious rearing on high of ugliness in the midst of beauty.
Really the modern choir of Llandaif is in no style at all ;
to call it Italian is a compliment almost as undeserved as
to call it Grecian ; it is simply hideous and unmeaning,
without reference to any principles of art whatever. A
1 This charge depends on the question whether the portico was
only the first instalment of a general reconstruction of the whole
church, or was designed to remain attached to the old building. In
the former case the general presumption was greater, hut the particular
accusation of incongruity of course does not apply.
88
LLANDAFF CATHEDRAL.
well proportioned range, even of engaged columns, is
after all, no contemptible object, and how grand interiors
may be made in the Italian style the structures already
referred to may testify. But at Llandaff there is nothing
of the kind, there is no architecture at all ; it must rank
not with St. Paul's and Queen's College, not with White¬
hall or the Clarendon, but with the meanest forms of the
dwelling-house and the conventicle. The sides are utter
bareness, the west front has just enough pretence to ren¬
der its vulgar display still more glaring. How this
could be ; how such a structure could ever have been
preferred, actually and seriously as a matter of artistic
taste, to the glorious fabric lying in ruins at its side, is
something which I must leave to the moral philosopher
to explain ; the architectural historian must resign the
attempt as beyond his powers. And it is clear from the
statement of the Dean that such was the case, and that it
was really preferred as a matter of taste ; that its hideous-
ness was not the result of niggardliness or want of funds
is shown by the large sum—£7000, and that doubtless
representing a considerably larger nominal amount now
—expended on these alterations. And I cannot forbear
repeating from the Dean's paper the judgment of the
" Rev. A. Davis," whose name deserves handing down
to posterity as its author :—" The church, in the inside,
as far as it is ceiled and plastered, looks exceeding fine ;
and when finished, it will, in the judgment of most people
who have seen it, be a very neat and elegant cliurchJ'^
After this, let no one speak of " churchwardenism" as the
acme of bad taste and destructiveness.
LLANDAFF CATHEDRAL.
89
Aîid now, in conclusion, for some notice of the great
work of restoration now in progress. One would have
thought that so noble a design could have been hailed by-
no one with any other feelings than those of unmixed
satisfaction, and that it could have stood in need of no
apology on any ground whether religious or artistic.
Yet it does seem necessary for a historian of Llandaflf to
go about to rescue its present chapter from being re¬
manded to the same class of destroyers" with their
predecessors of the seventeenth and eighteenth centuries.
The architectural world has been recently told, in no
measured terms, by an authority which, in one of the
kindred arts, is said to be no slight one, that restoration"
is the most complete destruction which can overtake an
ancient building. The doctrine that it is the duty of the
guardians of an ecclesiastical edifice to leave the church
entrusted to them in a state of ruin would doubtless have
been acceptable enough to the chapter of LlandaflP two
hundred years back, but in the present day Mr. Ruskin
may be left to enjoy this opinion by himself, as well as
to look on the tower of St. Ouen's as being, in his own
peculiar phraseology, " one of the basest pieces of Gothic
in Europe." That the sober warning with which Mr.
Petit concludes his volume on Tewkesbury is needed to
the letter no one acquainted with the present state of
ecclesiastical architecture will deny ; no destruction is so
complete as that involved in a bad restoration ; and how
very bad a large proportion of the restorations still
perpetrated are, every antiquary and ecclesiologist knows
but too well. Even where a restoration is good and
N
90
LLANDAFF CATHEDRAL
accurate, and the building is faithfully brought back to
its ancient and legitimate condition, it is too much the
custom to repair and furbish up every stone that is in
the least decayed, even though the stability and beauty
of the building be not the least affected, and so to destroy
that air of antiquity which is ever a charm to the eye
and to the mind, and is by no means devoid of use in
producing devotional feeling. But to protest against
restoration in the abstract, to involve the noble works
carried on at Llandaff and Ely and Hereford and Stafford
and Dorchester, in the same sweeping condemnation with
the barbarisms which have made the name of restoration
a bye-word, betokens a forgetfulness of the fact that " an
old church is not merely to be looked upon as a record
of past ages, hut as a valuable bequest for the use of
the present ; and therefore such restorations as are neces¬
sary to ensure stability, or propriety of appearance, must
be fearlessly undertaken."^
The force of the argument lies mainly in the words I
have given in italics. We must always remember that
ancient buildings consist of two distinct classes, to be
treated on entirely different principles. There are, first,
those which are simply monuments^ merely relics of the
past, and incapable of being applied to modern uses ;
secondly, those which, while equally relics of the past,
are also a trust for the present, being employed, or
capable of being employed, for their original purposes
at the present day. The former class comprises Roman
and Druidical remains, and the military structures of
^ Petit's Tewkesbury, p. 49.
LLANDAFF CATHEDRAL.
91
the middle ages. The great majority of these last
are now actually in ruins, and the few which are
retained as dwellings only prove that the two uses are
inconsistent. A castle cannot be employed as a modern
house, without spoiling it as a castle, and making an
uncomfortable house into the bargain. The second class
includes churches, of whatever kind, and those secular
erections of the middle ages—chiefly those of a public
nature—which are capable of being applied for modern
uses. The former class is typified by Stonehenge and
Caerphilly, the latter by Westminster Abbey and Hall.
From this distinction it follows that the first class are
objects of mere preservation^ the latter of restoration. In
a structure belonging wholly to the past, no modern
alteration, or even reparation, should be allowed ; in a
mere relic of antiquity, simply valuable as such, any
modern work is out of place; its presence at once destroys
the only value of the monument ; all we have to do is to
preserve, to hinder further dilapidation, we may prop,
but we must never rebuild. But in a building belong¬
ing both to the past and the present, no rational objection
can be brought against the introduction of any extent of
modern work simply as being modern. It is only against
had modern work, or the unnecessary destruction, obliter¬
ation, or alteration, of ancient work, that we can fairly
object. When a structure is still in possession of the
existing generation, and is actually employed for present
uses, we have two claims upon us, the memory of the
past, and the use of the present ; in adjusting the exact
limits of which the difficulties of the restorer consist.
92
LLANDAFF CATHEDRAL.
He has not simply to preserve what he finds ; he must
continually alter from the present, occasionally even from
the original state of the building entrusted to him. He
must show the greatest possible respect to the memory
of the past that may be consistent with the use of the
present, and consult the use of the present as far as is
consistent with due respect to the past.
The original error then is looking at a church^ or a
court of justice, a hall or college gateway, as if it were a
castle or a cromlech. To this Mr. Ruskin adds the doctrine
that when a building is utterly beyond hope, y/hen it is
dangerous, or otherwise impossible to be allowed to stand
as it is, the best course is to pull it down, and, I presume,
if necessary, build a new one. Yet I must humbly con¬
ceive that a chapter, or other body entrusted with the
guardianship of an ancient building, especially an
ecclesiastical one, is not only bound to maintain some
1 I am not now taking tlie religious view of tlie subject ; tlie remarks
made above apply equally to all ancient buildings retained for modern
uses, whether those uses be sacred or profane ; to the Royal Halls of
Westminster and Winchester—may the former be preserved to us
untouched !—as well as to their respective Minsters. For the same
artistic rules apply in both cases ; in a church restoration we are
simply supplied with higher and stronger motives; to antiquarian,
local, or national feelings, to love of art and beauty, a principle of
religious duty is added. But religion simply teaches us that we
ought to repair and keep clean" our churches ; taking for granted
certain grand principles of arrangement, it leaves to art to tell us how
it is to be done. So that when we have reached this point, we may
consider the artistic restoration of sacred and civil buildings under one
head. I may mention that the body in whose researches the present
work originated, the Cambrian Archseological Association, has always
been most honourably distinguished by great zeal for legitimate resto¬
rations, and correct views on the subject.
LLANDAFF CATHEDRAL.
93
fabric for the purpose required, but, under all ordinary
circumstances, to maintain, and therefore, when occasion
occurs, to restore, that very fabric which thej received
as a trust. It is not enough that the Dean and Chapter
of Llandaif should keep up some church or other for the
enthronement of the bishop and the devotions of his flock ;
it is surely their duty, simply as a matter of trust, to say
nothing of associations of every kind, to maintain that
very church which Bishop Urban commenced, and his
successors enlarged ; even though at last the identity—
one amply sufficient for all religious, and most artistic
purposes—should be only of the nature of that which
existed between the ship whose departure delayed the
fate of Socrates and that whose return precipitated that
of iEgeus. And this duty they are discharging effec¬
tually and nobly.
The work of restoration, commenced under the
auspices of the late dean, and continued with no less
zeal and energy by his successor, is now gradually ad¬
vancing from the east end, under the superintendence of
Mr. Prichard of Llandaff, the work thus enjoying the
great advantage of an architect resident on the spot.
The Lady chapel is now completed, and, during the
continuance of the repairs, is necessarily the only part
of the church employed for divine service. At present
the works are actively proceeding in the presbytery. It
is the sound principle set forth by the Dean first to com¬
plete the whole of that part of the church which is at
present roofed, and, so to speak, habitable, leaving, if
necessary, to another generation the more difficult and
94
LLANDAFF CATHEDRAL,
doubtful task of "the further restoration, or rather
reconstruction, of the western ruins/'
It is very remarkable that in the restoration of a
chürch which has undergone greater and more barbarous
transmutations than perhaps any other on record, there
should be so little call for original design on the part of
the architect. Enough remains to point out nearly all
the most important features ; even the destroyed south¬
western tower may be rebuilt with tolerable accuracy as
far as its outline and proportion is concerned. The only
important exception is the part just now completed, the
clerestory of the presbytery; this was necessarily original:
of that of the nave it will be remembered that, by some
happy accident, a single fragment exists, just sufficient
to indicate its original character. In the rest of the
church, we have seen that, even where the windows have
given way to Mr. Wood's insertions, enough remained to
replace the jambs with perfect accuracy ; the tracery
alone required to be entirely new. The whole restoration
is being conducted, as, in this case there can be no doubt
that it ought to be, on the most strictly conservative
principle ; not the slightest intentional change from the
ancient form of any part is to be apprehended, and those
portions of the building which are unavoidably conjec¬
tural are designed in strict harmony with that part of
the ancient work with which they are most closely
connected. The execution throughout is extremely good,
and the design of the original portions is also everywhere
good of its own sort, though in some places I could wish
that a design altogether different had been adopted.
LLANDAFF CATHEDRAL,
95
The Lady chapel has been most thoroughly and effec¬
tually restored ; it is only to be regretted that its arrange¬
ments—necessary for the time at least—as a parish church
take off somewhat from the general effect of its internal
architecture. A building, essentially of the type of a
chancel or a college chapel, is closely packed with benches
like a parochial nave, with hardly any open space even
at its east end.^ In this part of the church all that was
to be done was strict and accurate repair, with the single
exception of the tracery of the east window ; the former
has been done most faithfully, and the latter supplied in an
appropriate manner from the examples of like style and
date in York chapter-house. Externally I must most
strongly regret the absence of a high roof ; and I am in¬
deed at a loss to conjecture on what ground a low^ one has
been adopted in a structure where, as a mere matter of
precedent, it is all but a solecism,^ and where, what is of
more importance, the general effect of the whole build-
^ I am not sure that all the existing arrangements of this chapel
would be approved, even under the pressure of its very peculiar cir¬
cumstances, by my friends of the Ecclesiological Society, but I pur¬
posely abstain from everything which can in the least degree tend to
anything like ritual controversy. The conventional proprieties of such
points are not to be despised, but it is not always desirable to mix
them up with purely architectural considerations.
^ I believe I may do more than hope that the high gable standing
free at the east end may shortly he removed. Anything is better
than unreality.
2 I apprehend that, in England, the high roof was quite universal,
except under peculiar circumstances, until quite the later days of the
Decorated style. Warmington and Barnwell St. Andrew's are early
instances of the low roof, but there the special reason is plain, having
not to interfere with the belfry windows.
96
LLANDAFF CATHEDRAL.
ing imperatively demands it. Even at present^ with the
high roof of the main fabric confined to the narrow limits
of the presbytery, the contrast begins to be painful ; and
,much more will it be so when the low-roofed Lady
chapel forms the termination to the long line of steep
roof stretching along the whole nave and choir. Its effect,
which may be judged of from the ideal representation
which I have given, will be a most thorough break-down,
A high roof would doubtless have blocked the new east
window of the presbytery ; but the latter I cannot con¬
sider as of anything like the same importance in the
general estimate of the church, and might still have
remained, if thought desirable, as an internal ornament.
And we must remember that the clerestory of the
prebystery is conjectural ; we do not know for certain
what the old one was ; to justify so important a deviation
alike from beauty and precedent we ought at least to
have it demonstratively proved that there was a Deco¬
rated east window to the presbytery and that the roof of
the Lady chapel was lowered to make room for it—I
shall never believe but that it had a high roof originally ;
and even in that case I am not quite sure that we should
imitate such a proceeding, and ruin the effect of the
whole chapel—I might almost add of the whole cathedral
—for the sake of a single not very prominent window.
This clerestory of the presbytery I have already men¬
tioned as the only part of the church , where any great
extent of absolutely original work was required ; and I
have also, in an earlier part of the volume alluded to
its excellent effect, and to the happy manner in which
LLANDAFF CATHEDRAL.
97
the general type of the earlier work to the west of it
has been translated into a later style. There can be no
doubt that the course pursued has been the right one in
selecting the Decorated style for this portion. Though
its predecessor, whatever it was, had been entirely
destroyed, it is certainly most natural to conclude that a
clerestory would be added when the Decorated arches were
cut through, if none previously existed ; and we have seen
the difficulties attending the belief that there ever was a
Romanesque clerestory ; nor is it probable that the
Early English one of the choir would be continued into
the presbytery, which seems not to have been at all
affected by that enlargement. At any rate, none now
existing, it was clearly advisable to adapt the clerestory
to the now mainly Decorated character of the presbytery ;
and the result internally is exceedingly good. Externally
it is less pleasing ; the contrast between its two indepen¬
dent windows and the range of couplets and pilasters in
the nave and choir—a contrast internally precluded by
the arch between the choir and presbytery—will be
found unsatisfactory, but was, of course, unavoidable.
Still there is a bareness in the parapet which need not
have been, and which contrasts unpleasantly with the
beautiful open-work which now fringes the aisle below.
But the greatest fault is the roof, an open timber roof,
instead of a vault. I do not see how this can be justified
even by the peculiar quasi-parochial appearance of this
cathedral.
The selection of the tracery is very judicious, being
just on the turning-point between the Geometrical of the
o
98
LLANDAFF CATHEDRAL
Lady chapel and the Flowing of the western aisles. That
in the large flat-headed window is, with equal propriety,
of a later type than any other in the church.
The arch between the choir and the presbytery—the
chancel arch as it would be with the ordinary ritual
arrangement—is also new; it has a very good effect, but
I greatly fear that it is carried so high as to interfere
with the vaulting of the presbytery, and the flat ceiling
of the nave and choir. So strongly do I feel on those
points that I should like to see it reconstructed at a lower
level.
In the choir itself the arches blocked by Mr. Wood
have been opened, but the clerestory and roof have yet
to be reconstructed, and new fittings are required
throughout, the existing screen and stalls, though evi¬
dently retaining the old arrangements untouched, being
of the most miserable character possible. The execution
of this repair would complete the restoration of the
habitable part of the church, leaving, as was above-men¬
tioned, that of the ruined nave to a future generation.
If funds are not available for the whole, this is clearly
the right course. Yet, if a sum suflScient for so great a
work could be forthcoming in these days, the nave and
choir are so strict an architectural unity, and the build¬
ing is brought to so lame and impotent a conclusion by
Mr. Wood's west front, that it would be far preferable
for the whole of the Early English work to be executed
at once. At present the cathedral will have nothing
worthy of the name of a nave, only a sort of dimi¬
nutive ante-chapel ; while could the whole be restored
LLANDAFF CATHEDRAL.
99
at once, the long vista from the arch into the pres¬
bytery to the unrivalled western triplet, would be
surpassed by no church whatever of its own scale and
style. And I would again put in a word in favour of
the flat ceiling, so clearly intended in the original design,
and which, in the full length of the restored nave, would
have far more justice done to its general effect—one, I
imagine, greatly depending upon length—than in the
confined space of the present choir and ante-chapel.
The restoration of the nave will be a great trial of
constructive skill. Except in the south-west tower, the
whole design, both in composition and detail, can be
recovered with absolute certainty. But how far the
shattered and broken fragments of the arcades will
admit of sustaining any weight, and of being again
employed in connexion with the necessary modern
work, is altogether another question. The course clearly
is to preserve whatever is sound, and to reconstruct the
rest, even should it involve, what I should hardly con¬
ceive would be requisite, a complete rebuilding from the
ground. Ordinarily to patch up a ruin, to do anything
more than preserve it from further dilapidation, is
a barbarous absurdity. This remark of course applies
most fully to all military structures, and is also equally
applicable to most ecclesiastical ruins. The cases in
which a positively ruined church, as opposed to one
merely disused or desecrated, can be a legitimate object
of modern restoration are quite exceptional ones. To
repair any church is a good work, but the great majority
of ruined churches are in such positions that their resto-
100 LLANDAFF CATHEDRAL.
ration to sacred uses had at least better be postponed till
we have erected all the new churches at present required,
and restored all the old ones still in use to a sound and
seemly condition. And even in the few cases where the
restoration of a ruined church would be desirable on the
ground of church accommodation, the ruin is often in
such a state that the erection of a distinct new church
would usually be the wiser course.^ But the ruined nave
of Llandaff is altogether an exceptional case ; it is no
forsaken abbey, for whose preservation no one is respon¬
sible, and on which no congregation depends for religious
worship ; it is an integral part of an existing temple, the
cathedral church of a diocese, the parish church of a
parish, which, while it remains in its present condition, is
at once architecturally and ecclesiastically incomplete.
If church restoration be in any case a good work ; if art
and religion alike require that Llandaff Cathedral should
again exist as a whole, perfect and beautiful, we must
not be afraid of meddling with the lovely, but still
melancholy—and to a past generation disgraceful—
appearance of the nave in its present state. Even should
entire rebuilding be necessary, it must be dared at all
hazards. No picturesque or antiquarian feeling must
1 I should imagine tlie Friary Cliurcli at Reading might be an
exception. It is not simply ruined, but more foully desecrated than
any church I ever saw, and probably has few parallels in England.
Yet the nave appears to want little more than to be cleared and new-
roofed to be again used as a church. I should suppose an additional
church at Reading would not be undesirable, and surely some more
convenient place—-to put it on the lowest ground—might be found for
the town gaol.
LLANDAFF CATHEDRAL.
101
be allowed to interfere with the thorough carrying out of
the great work so happily commenced. Happily there
is no fear that any such will be allowed to interfere, but
in the present unsettled state of the great question of
restoration in general, a vindication of the only right and
necessary course can hardly be deemed out of place.
THE END.
TENBY: R. MASON, PRINTER, HIGH STREET.
CORRIGENDA.
Page 44, line 9, for Larger/' read Large.''
„ 45, ,, 4 from hoitom,for "ecclestical," read ^^ecclesiastical."
„ 63, ,, 1, The expression here is too strong, as I believe the
Early English choir at Lincoln bears date a few
years earlier than 1200. But, as this stands quite
alone, it hardly affects the general argument.
„ 74, „ 17, dele " two."
J^y the same Author^
PRINCIPLES OP CHURCH RESTORATION.
London : Masters, 1846. 8vo, Is.
A HISTORY OP ARCHITECTURE.
London : Masters, 1849. 8vo, 14s.
REMARKS ON THE NOMENCLATURE OP GOTHIC
ARCHITECTURE.
Oxford : Parker, 1849. 8vo, 6d.
THOUGHTS ON THE STUDY OP HISTORY, &c.
Oxford : Parker, 1849. 8vo, Is. 6d.
THOUGHTS ON THE THIRD PORM OP THE NEW
EXAMINATION STATUTE.
Oxford : Parker, 1850. 8vo, 6d.
NOTES ON THE ARCHITECTURAL ANTIQUITIES OP
GOWER.
London : Pickering. Tenby : Mason, 1850. 8vo, 2s.
AN ESSAY ON THE ORIGIN AND DEVELOPMENT OP
WINDC V TRACERY IN ENGLAND.
PAET I., WI I NUMEROUS ILLUSTRATIONS.
Oxfor : Parker, 1850. 8vo, 5s.
To he completed in Four Quarterly Parts.
This book is a preservation facsimile
produced for the Northwestern University Library.
It is made in compliance with copyright law
and produced on acid-free archival
60# book weight paper
which meets the requirements of
ANSI/NISO Z39.48-1992 (permanence of paper)
Preservation facsimile printing and binding
by
Acme Bookbinding
Charlestown, Massachusetts
2010