Full text of "Lombardic architecture; its origin, development and derivatives"

of Toronto

Jack Ftyrie

LOMBARDIC ARCHITECTURE

VOL. II

THE ABBEY CHURQ

RDIC ARCHITECTURE

UN, DEVELOPMENT AND
TIVES <* ** By G. T. RIVOIRA

ED BY G. McN. RUSHFORTH, M.A.

KR EIGHT HUNDRED 1LLVSTRJTIONS

VOL. II

WILLIAM HEINEMANN

MCMX

THE ABBEY CHURCH OF METTLACH.

[Frontispiece

LOMBARDIC ARCHITECTURE

ITS ORIGIN, DEVELOPMENT AND
DERIVATIVES * * By G. T. RIVOIRA

TRANSLATED BY G. McN. RUSHFORTH, M.A.

WITH OPER EIGHT HUNDRED ILLUSTRATIONS

VOL. II

LONDON .

WILLIAM HEINEMANN

MCMX

957225

CONTENTS OF VOL. II

PAGE

INTRODUCTION ....". .... 3

CHAPTER I
THE BEGINNINGS OF THE LOMBAKDO-NORMAN STYLE IN BURGUNDY . 5

CHAPTER II
THE LOMBARDO- NORMAN STYLE IN NORMANDY . . . 45

CHAPTER III

ECCLESIASTICAL ARCHITECTURE IN ENGLAND FROM CONSTANTINE TO
THE NORMAN CONQUEST . . . . . . .125

CHAPTER IV

THE LOMBARDO-NORMAN STYLE IN ENGLAND . . . . .194

CHAPTER V
THE EARLY CHRISTIAN MONUMENTS OF IRELAND . . . .255

CHAPTER VI

THE ECCLESIASTICAL ARCHITECTURE OF GERMANY (FROM CONSTANTINE
TO THE ELEVENTH CENTURY) .... . . 267

CHAPTER VII

THE LOMBARDO-RlIENISH STYLE ....... 303

PART II

VOL. II

ones.

INTRODUCTION

AVING traced the origins of the Lombardic vaulted basilica, I now
proceed to reconstruct the history of its principal derivatives in the
countries north of the Alps. In doing so I shall confine myself to giving
the main outlines, leaving to others the task of indicating the secondary
These outlines will be based exclusively on original monuments which have
come down to us in a state of complete or partial preservation, or of which we
have descriptions or drawings ; so that students of mediaeval architecture may have
some clue to guide them in the labyrinth where at present they have lost their way.

The views of writers on the origins and development of the great styles of
ecclesiastical architecture practised in the Transalpine lands between the epoch of
1000 and the first appearance of the Pointed style show immense divergence. Round
these origins and round this development there has grown up such a dense growth
of " influences "Syrian, Roman, Gallo-Roman, Byzantine, Barbarian, Arabic-
having their source, when evidence fails, in the imagination, that it is no easy matter to
unwind the tangle. And the difficulty has been intensified by the erroneous belief
that the East was the chief inspirer and authoritative guide of all the mediaeval
Christian systems, from the Romano-Ravennate down to the Pointed style.
To all this must be added :

(1) The imperfect acquaintance with history, philology, construction, statics, art,
possessed by most of those who approach the subject of the Middle Ages without
understanding them, and, moreover, have neither the knowledge nor the ability to
deal with the main question at issue in its entirety, and therefore lose their way in the
by-paths of secondary problems.

(2) The unwillingness which some people have to open their eyes to the clear,
tangible truth of facts, preferring vague probabilities to simple reality.

(3) The conservative spirit, to which everything that is new is abhorrent.

(4) The confusion generally prevalent with regard to the dates and essentially
distinctive characteristics of the buildings belonging to the styles which form the
subject of this work.

B 2

4 INTRODUCTION

Hence, it is easy to see how difficult will be the road which I am about to
traverse, and how arduous the ascent of that height whose still virgin summit it is
my ambition to attain. However, with the aid of a long period of preparation devoted
to the acquisition of knowledge which is not to be learned in schools or books, and of
twenty years of research and study extending over some thousand buildings
scattered over the ancient Roman world, I shall make the effort to reach that summit
in the cause of Monumental Archaeology. I am well aware that my work will arouse
criticism ; but I am supported by a lively confidence that, though in the course of
time it may require correction and completion, that work nevertheless marks a sure
advance along the road of truth. And truth, sooner or later, will always win the
day.

The leading idea of this part of my book is as follows. With the appearance of
the Lombardic basilica, a spring-tide of new styles of building burst forth north of
the Alps, the chief being the Lombardo-Norman and the Lombardo-Rhenish, as I
prefer to call them. I do so because the terms " Romanic " and " Romanesque,"
applied to this great art of the Middle Ages, do not seem to me to be justified ; for
we might equally well call the Byzantine style " Romanesque," seeing that the
Byzantine vaulted basilica in its supreme expression, St. Sophia at Constantinople,
had its origin in the great Baths of Rome, as I have explained.

It was the monks of St. Benedict who acted as the sponsors of these styles ; it
was under their protection that they were nurtured ; it was to the Benedictine services
that they were adapted. Brought into conformity with the tendencies and peculiar
character of the peoples among whom they were introduced, and even modified to
suit their climates, clothed in new forms, and treated with variety of conception, these
two styles made remarkable progress, and were diffused far and wide by the agency
of the Benedictine monks.

In the course of my argument I shall again have to shatter more than one legend
about the real origins of the chief characteristics which go to make up Lombardic
architecture ; and I shall have to apply the same treatment to other legends con-
cerning the birth and growth of the principal styles which derived from it their life
and sustenance.

The illustrations provided to explain and confirm the text will, as before, be, in
the main, reproductions of photographs. Mere drawings, in which the artistic element
is emphazised at the expense of truth, have in the past lent themselves too readily to
illusion ; and though they may thereby satisfy the taste of the dilettantes of archi-
tecture, they only lead real students astray.

CHAPTER I

THE BEGINNINGS OF THE LOMBARDO-NORMAN STYLE IN

BURGUNDY

first solid foundations of the Lombardo- Norman style were not laid,
to judge by the buildings which have come down to us, in Burgundy, as
some think, but in Normandy. Burgundy, however, can claim the credit
of having provided the field for the free exercise and consolidation of the
forces which later helped to lay those foundations, and of having produced the first
flowers of the new style.

The soul of this work of preparation was the Benedictine monk, William of
Volpiano (961-1031), born on the island of San Giulio in the Lake of Orta, brought
as a youth to Cluny by Majolus abbot of that monastery (948-994), and afterwards
appointed abbot of Saint B^nigne at Dijon (990) by Bruno bishop of Langres
(981-1016). With regard to this man of great learning, of iron will, a great reformer
of the monastic orders, architect and builder of churches and convents, the diffuser of
Italian culture in Burgundy and Normandy, we know, among other things, that he
took in hand the erection of the new abbey church of St. Benignus at Dijon
(1002-1018), and also built the abbey of Fruttuaria in Piedmont founded in 1003 and
consecrated in ioo6. 1234

Of this preparatory work the famous church at Dijon, in which some of the
characteristics of the Lombardic basilica appeared for the first time, was the highest,
most solemn expression. Let us turn to it.

THE CHURCH OF SAINT BENIGNE AT DIJON, built by St. Gregorius bishop
of Langres (507-539) above a crypt which he raised in height, 5 was renewed in 871
by Isaac bishop of the same diocese (859-880). Having fallen into decay, and part
of it being in ruins, it was rebuilt by William of Volpiano, the extreme eastern part
of the previous structure being preserved. The rebuilding was begun in 1002, and
the dedication took place in 1018.

The church, however, was unfortunate. In 1096 the central tower fell, and was
rebuilt. Following this operation and the repairs of the damage caused by the
catastrophe came a second dedication performed by Pope Paschal II in 1107. In
1 136 the whole building suffered considerably from the terrible conflagration which

1 Mabillon, Ada Sauct. Ord. S. Benedicli. Vila s. Guillelmi abi>atis Divionensis. Auetore Glabro
Rodulfo nionacho.

3 Id., Vila s. Guillelmi abbatis. Exccrpta ex libra de rcvtlatione, aedificalione, el auctoritale mottasterii
Fiscamnensis.

3 D'Achery, Sfitileguim. Chronica S. Benigni Diviontmis. ' Savio, op. fit.

Man. Germ. Hist. Gregorii ef. 7'iironensis liber in gloria marlyruin.

6 LOMBARDIC ARCHITECTURE

devastated Dijon, and the restoration necessitated thereby must have been far-
reaching inasmuch as it involved a second reconsecration at the hands of Pope
Eugenius III in 1146. In 1271 the rebuilt tower over the crossing again collapsed,
injuring not only its immediate surroundings but other parts of the church as well. 1
After this disaster the portions that had suffered most were rebuilt in the new or Pointed
style, thus satisfying the passion for innovation which followed the appearance of
the new architecture. Under these circumstances, all that was left of William's church
was the rotunda, with the ancient chapel of St. John Baptist at one end of it, and a
considerable part of the termination of the basilica at the other. Finally, in 1792 the
pickaxe demolished everything of his church that still remained above ground.
The part below ground, with its vaulting broken in, and degraded into a receptacle

acpdcation. de to, couD ef-dcce (jut resk ctej anci^^nnesEr/(iscs e/u. siaxeme Stc/e c^j

Irots R-otundtj ctu, ^^^^^

irorrtta ,/n f-a

A SaJ ftaff ttctj prmtirrt Egliie IStl't furff!rtfO"*f.i-tgm Jf Lujyrfj
eeHtunfnctrtlfrit Jjt .' i.t..,iti- Jif, /, . *f-,:i/,.: ..*!' J'otfjtctit it la K#lvtt,{r if,
'f S^Gryev-f . afoa&JJU'lt </l

KeU-itJf J., .nit..-*

J,i "i.l:~i C. lr a ,...te

Jf Li setpnijt F.jlur Ijflff'ar S

iFifrk ft orila* oart-tfrbe C-*iif*ui
Cfrrtsrlt. att OnJ

Je -f

F'g- 367. Dijon. Saint Benigne. Section of the eastern end (1002-1018, and Vlth and IXth Centuries).
(From Plancher, " Histoire gintrale el particulitre de Bourgogne")

for rubbish, has been restored of recent years to the condition which it now
presents.

The chronicle of Saint Benigne 2 contains a description of the interior of the
church, though it is not without omissions and inaccuracies. Plancher 3 has furnished
interesting particulars about the portions still surviving in the first half of the
XVIIIth century, and of these he gives two ground plans, a section of the elevation,
and two views of the exterior.

The structure consisted of a basilica, terminated towards the east by a rotunda.
Both basilica and rotunda were of three stories, one being underground. The latter
or crypt was composed of a circular vaulted structure consisting of two concentric

1 Man. Germ. Hist, Annales S. Benigni Divionensis.

* D'Achery, Spicilegium. Chronica S. Benigni Divionensis,

3 Histoire ghitrale et particiiliere de Botirgogne.

BURGUNDY

f .SI

circles of columns enclosed by an outer wall with half-columns, lighted by four
windows splayed on the inside, and with a chapel at the eastern end used as a
chancel. Over the central space rose a kind of open octagonal tower, having three
series of arcades one above the other, the two lower supported by columns, the
uppermost by piers. It was crowned by a cupola with a circular opening at the
top (Fig. 367).

To the west of the rotunda was another structure, shaped like a T (in the under-
ground part), consisting of a nave and a transept or cross nave, both of the same
length, and separated in either case from
aisles by two rows of columns. The nave
and aisles were flanked by closed vestibules
forming outer aisles, and terminated in an
arcaded apse which contained the tomb
and altar of St. Benignus. On the eastern
side of either arm of the transept was a
rectangular chapel with an apse, and a
semicircular recess taken out of the thick-
ness of the wall. This arrangement was
based on the lines of the previous church,
in the form which it had acquired in the
IXth century ; a fact which has come to
light in the recent restoration.

Of this underground church there sur-
vives the rotunda with its eastern adjunct,
the chapel containing the tomb of St.
Benignus, and the eastern end of the
basilica (Fig. 368).

In the rotunda (Fig. 369) the isolated
shafts are surmounted by Pre-Lombardic
cubical capitals, hollowed out at the angles,
each of which is filled by a smooth pointed
leaf, while the faces are left plain. There
are two exceptions in the middle row of
columns, with the faces bearing a human
figure, sometimes of very curious character,
in the act of supporting with its arms a
conventional abacus (Fig. 370). The crosses
roughly engraved on two of the angle leaves
of one of these capitals are a later addition.
The figures are worked in rather round

relief, and without undercutting. Design and execution alike are barbarous and
elementary.

In the outer circle of columns, opposite to the tomb of St. Benignus, are
two more capitals carved with animals of rude but robust design, boldly conceived,
treated with variety of action (and not a uniform one, like the human figures on
the two capitals just described), strongly, and sometimes completely undercut
(Figs. 371, 372). These belong to a period considerably later than the rebuilding
of the early Xlth century. They may probably be connected with the restoration
necessitated by the fall of the central tower in 1096, and this is the date to which an

D&cripttanoueaptua&eti flu. Plan jtom&tralele la b
J'Jicruaru et da mortxouac. tjui

~ltjt I*

j'urft

ft **4 Li trtii m*irtJ J* /nJIvfeW*. Jr fort Ointfr ft
<fuilu>iH rtatfHt tin an '

MMM*tt6ii

Fig. 368. Dijon. Saint Benigne. Underground
plan of the eastern end (1002-1018, and Vlth and
IXth Centuries), (from Plancher, " Histoire
gMralt et particuliUrt de Bourgogne")

LOMBARDIC ARCHITECTURE

experienced eye would be inclined to assign them. From the capitals spring
round arches.

The vaulting, which has been rebuilt on the pattern of the few old portions
which survived, is of barrel form over the inner annular aisle, and alternately of

Fig. 369. Dijon. Saint Benigne. Crypt of the Rotunda (1002-1018).

barrel and intersecting form over the outer one. The barrel vaults are constructed of
radiating oblong blocks of stone, roughly cut. The central space, originally open
at the top, is now covered by a cupola.

In the arcaded apse round the tomb of St.
Benignus still stand six original columns with Pre-
Lombardic cubical capitals hollowed out at the
angles, and with plain faces. The southern area of
the east end of the basilica, to the side of the saint's
tomb, is divided into aisles, the vaulting of which has
recently been replaced ; and here are seven capitals of
the same date as the two carved ones facing the
tomb. One of them shows foliage and monsters.
Another has grooved leaves. Another has similar
foliage and interlacings, with four heads at the angles
each holding in its mouth a leaf which takes the place
of a volute. These heads are by the same hand that
executed those at the angles of one of the capitals
about the tomb. A fourth is ornamented with large
volutes at the angles, roses, palm leaves, and animals
(Fig. 373). The fifth is carved, partly with plain
leaves the tips of which are divided and curl over in volutes (i.e. crockets), and partly
with curiously wrought foliage and interlacing wicker work. The two remaining
ones show a double row of stiff plain leaves. Those of the bases which are old

Fig. 370. Dijon. Saint Benigne.
Crypt of the Rotunda. Capital
(1002-1018).

BURGUNDY

are stepped, and differ in form from those in the rotunda belonging to the time of
William of Volpiano. They are clearly of another period.

In the northern arm, only a portion of which survives, two original capitals may
be observed, one cubical with the faces left plain, the other the fellow of the two
which we pointed out in the central circle of
the rotunda.

The presence in this arm and in the
arcade round the tomb of St. Benignus, i.e.
in the most important part of the rotunda,
of rudely executed Pre-Lombardic capitals of
the type which prevails all through the crypt,
is a clear proof that the elaborately worked
capitals which we assigned to the end of the
Xlth century are certainly not of the same date
as the crypt itself.

The chapel of St. John Baptist consists
of a rectangular chamber, regularly orientated,
roughly built of stones with very thick joints.
The cross vaulting has been rebuilt. Beyond
it is another rectangular chamber, even more
rudely constructed, showing a mortar which
differs from the other. Here again the barrel
vault has been replaced.

The difference between the masonry of the walls of the two chambers, and also
of the outer wall of the rotunda, makes it certain that the three structures are not of
the same date. And seeing that this adjunct to the termination of Saint Benigne is
mentioned about the year 938, when the dedication to St. Mary took place, 1 we must
ascribe it, for the present, to the restoration of the church in 871. Then, when we

Fig. 371. Dijon,
the Rotunda.

Saint Benigne. Crypt of
Capital (1096-1107).

Fig. 372. Dijon. Saint Benigne. Crypt of
the Rotunda. Capital (1096-1107).

Fig. 373. Dijon. Saint Benigne. Capital at east
end of the underground church (1096-1107).

take into consideration the barrel vault of the chamber at the eastern extremity of
the church, the method of covering regularly adopted for the earliest Christian

1 Chomton, Histoire fie Ftglist Saint- Binignt de Dijon.

io LOMBARDIC ARCHITECTURE

oratories in France, and in presence of the greater rudeness of the masonry of this
chamber, and also of the persistence with which it was preserved, evidently as being
a spot of peculiar sanctity, through the two rebuildings of the original church of
St. Gregory, bishop of Langres, in the IXth and Xlth centuries, we may fix the date
of this oldest chamber at a period earlier than the IXth century, and perhaps in the
time of the said bishop, while the other chamber may be ascribed to the IXth
century.

On the ground floor, or church proper, the rotunda had the same plan as in the
crypt ; but in the outer wall were engaged whole, and not half columns.

The basilica was in the form of a cross, with aisles and galleries, and was some
210 feet long and 85 feet wide. The height was about 50 feet in every part of the
interior except the nave, which was some 65 feet. It was supported by massive
quadrangular piers and columns, with shafts at the angles of some of the piers in the
galleries. The nave, terminated by an arcaded apse, was flanked by vaulted
double aisles ending in two chapels and two recesses like those of the underground
church. Over the crossing rose a broad tower of stone.

On the first floor the rotunda was in the form of a crown, with a smaller circle of
isolated columns, and a larger one of columns engaged in the outer wall. So far as
one can judge from Plancher's sectional view, the capitals belonging to these columns,
and to those of the upper story of the rotunda, were of the cubical shape to be seen
in the underground part.

Interesting details of the two stories above ground are furnished by two
paintings of the XVIIIth century, preserved in the sacristy of the present Saint
BeViigne, which represent the rotunda in process of demolition.

In the upper story of the rotunda a chapel projected towards the east, above
the square sanctuary of which rose a low tower of the same form, heightened in the
Xllth century. This story communicated with the two below it by means of two
spiral staircases, rising as high as the roof, contained within two round towers, also
increased in height in the Xllth century, placed to the north and south of the rotunda,
and forming part of its structure. The elevation of these towers was little more than
that of the rotunda, as may be seen in the two pictures referred to, and in another
which goes with them, giving a view of the interior of Saint Benigne before 1792. At
the foot of each spiral stair two passages started, connecting the rotunda with two
arcaded galleries, and also with two spiral staircases which gave access to the roof and
a way protected by a parapet all round the outside. Two other staircases, formed
symmetrically in the front wall of the church, connected the galleries with the aisles
below them. The building was provided with eight towers, in which other columns
occurred, and had three entrances.

The exterior of the original portions of William's church still existing when
Plancher wrote, and when the pictures mentioned in the sacristy of the present Saint
Benigne were executed, was decorated with blank arcading and arched corbel courses,
in some cases continuous, in others broken by lesenas.

Such are the main features of the church as it has been described to us. About
the rotunda we know enough to form a clear idea of what it was like. Not so of the
basilica, where too much essential information as to its statical, constructive, and
decorative features is wanting for us to be able to reproduce its real character even
approximately. Nor are the inferences which we may draw from the various data,
and from a comparison with other relevant buildings, sufficient to throw much light
on the subject. A few considerations will show this.

BURGUNDY

1 1

It is known that, in the portion which formed the junction between the rotunda
and the basilica, the walls of the central body rose very little above the horizontal line
of the aisle roofs (Fig. 374) ; whence it might be inferred that the walls of the nave of
the basilica were constructed in the same way. It is equally clear from Plancher's
evidence that the nave was vaulted. We are also informed by a statement in the
Dijon Chronicle that the aisles were vaulted as well. So that we might infer the same
for the galleries, and consequently that the entire building was vaulted. This would
account for the remarkable thickness (over 5^ feet) of the outer walls, and also for the
adoption of the plan of keeping the walls of the nave low, with the object of resting
the vaulting on them in such a way that the thrust should be resisted by the gallery
vaulting and by the outer walls. This idea would be suggested by a fear in the minds
of the architect and builders of compromising the stability of the nave if its walls were

t/tt vV<i- du. Jfftetib'itju. ct ttfj rtjtcj Jfj ajieteiu &a<vnt'U ftu v jtnit

JQilttJ t^yttft/l'K t/ll inclttf

Fig. 374. Dijon. Saint Benigne. Elevation of the eastern end (1002-1018, and Vlth and IXth Centuries).
(From Plancher, " Histoii e gtfnJrale et particultire de Bourgogne.")

raised high enough to admit of windows : a fear which is not difficult to understand
when we consider the condition of statical science as applied to vaulting at the
beginning of the Xlth century.

On the other hand, the clear reference in the Chronicle to the parts of the basilica
which had vaulting (very probably rude cross vaults like those in Ste. Marie at Bernay,
founded in 1013, and also designed and built by William of Volpiano), viz. the aisles
(". . . geminas porticus dupliciter transvolutas "), the very serious damage suffered
by the church in the fire of 1 136, and, thirdly, the fact that William had provided
mainly wooden ceilings for the nearly contemporary church at Bernay, are all things
which lead one to believe that originally the basilica had its nave, transept, and
gallery roofed with timber in the same fashion as the church of Bernay, and also that
of Cerisy la Foret, rebuilt in 1030 either from the plans or under the immediate
influence of William of Volpiano. In that case, the wooden ceilings of Saint Benigne
must have been substituted for a solid roof of masonry in the course of the restora-
tions occasioned by the fire referred to, involving a reduction in height of the
previously lofty walls of the nave so as to bring them into relation with the new

12 LOMBARDIC ARCHITECTURE

conditions of equilibrium established for the structure. It is impossible to suppose
that these walls were at first blank and windowless, as Plancher's views would suggest,
since the maximum height of the church (some 65 ft.) mentioned in the Dijon
Chronicle (where the height of the nave and not of the central tower, as has been
suggested, must be referred to), presupposes a lofty wall space for the introduction of
light above the galleries, just as at Bernay and Cerisy la Foret.

The doubt which arises as to the manner in which the basilica was roofed may
be extended to the piers which separated the nave from the aisles. For while, on the
one hand, one might imagine that they were alternately larger and smaller, on the
other, the existence of supports of uniform size in the church at Bernay, known to have
been built by the same architect, makes one suspect with good grounds that the same
design was followed in both buildings.

It has been a common view that the plan of Saint Benigne/so far as relates to the
idea of joining a basilica on to a rotunda, was suggested by the example of the Holy
Sepulchre at Jerusalem. This, however, is an erroneous idea, based on plans of the
original church of Constantine which are partly imaginary, like De Vogu'6's, 1 though
the world has shut its eyes and accepted them, and exhibit a basilica ending in a
semicircle with an aisle round it and three apses. In other cases the theory is based
on statements which have no facts to support them. This is not surprising when we
consider that the history, at any rate up to the time of the Crusades, of the complex
of buildings which make up what is commonly called the Church of the Holy
Sepulchre at Jerusalem is still involved in obscurity. And as I feel that it is a very
desirable thing to try to remove that obscurity, I therefore undertake the task, so far as
I may be allowed to accomplish it by the incompleteness of the historical documents
and the absence of much essential information as to its construction and decoration
which might throw light on the changes through which this famous sanctuary has
passed.

THE CHURCH OF THE HOLY SEPULCHRE was only just founded by
St. Helena in the year before her death, 327, when she went on pilgrimage to
the Holy Places in the hope of forgetting the tragedies which had taken place
in her family. The building was erected by Constantine the Great, and dedicated

in 335-

It consisted of the " Anastasis," or Church of the Resurrection, at the west,
containing the Holy Sepulchre ; and the great Constantinian basilica, or " Martyrion,"
at the east. Between them came the sanctuary of Calvary (in Cyril of Jerusalem
[IVth century] "Golgotha" means two things: the entire hill, and the summit
or mound on which the cross was planted -) which was a space open to the sky,
enclosed by a silver railing. 34567 On this spot, brought into prominence by
Constantine's levelling operations, 8 a chapel was erected by the younger Melania
(t 439)- 9

] Les tglises de la Terre-Sainte.

- Migne, Pair. Gr. vol. 33. Descriptio et historia basilicae Resurrectionis.

3 Biblioteca dell' Accademia storico-giitridica, 1887. Gamurrini, S. Silviae Aquitanae peregrinatio ad Loca
Sancta (about 385-388). 4 Eusebius, Vita Comtantini.

5 Tobler et Molinier, Itinera Hierosolymitana et descriptiones Terrae Sanctae. 5. Eucherii epitome de Loci's
aliijiubits Sanctis (about 440).

6 Id., op, cit. Brtviariits de Hierosolyma.

7 Id., op. cit. Antonini martyris perambiilalio Locontm Sanctorum (about 570).

8 Wilson, Golgotha and the Holy Sepulchre.

9 Analecta Bollandiana. Vita s. Alelaniae lunioris. Auctore coevo et Sanctae familiari.

BURGUNDY 13

About the Church of the Resurrection all that we know is : on the authority
of the "Jerusalem Breviary" (of about 530 according to one view, or about
420 according to another) that it was in the form of a rotunda ; from the evidence
of Eusebius, that it contained magnificent columns ; and, thirdly, from the statement
of Antiochus, 1 that it must have had a wooden roof, as it was set on fire by the
Persians (614).

Constantine's basilica is described by Eusebius as follows. It excited wonder
by its dimensions. The nave was supported by columns, and flanked by double
aisles with galleries over them. The interior was lined with marbles of various
colours, while the exterior was faced with stone so finely wrought and adjusted
that it did not yield in beauty to the marble. It terminated towards the west
in a hemispherical apse, the interior of which was encircled by twelve columns
symbolising the Apostles, each supporting a silver vase an arrangement like that
of some exedras in the Thermae of ancient Rome. 2 The ceilings of nave and aisles
glistened with gold. The front of the church was turned towards the east, and
contained three doors. Before it extended a spacious atrium.

This orientation is a confirmation of my statement that the Basilica Ursiana
at Ravenna (370-384) was the first to have the apse placed at the east. Another
piece of evidence is the basilica erected by Constantine at Baalbeck in the middle
of the great court of the temple of Jupiter (138-249), with nave and aisles separated
by rectangular piers ; where the three semicircular apses (the lateral ones being
terminated by little sacristies) are at the west, while in the transformation which the
basilica subsequently underwent, the principal apse (semicircular internally but
pentagonal externally after the Ravennate fashion) was placed at the east after the
Ravennate model. This has been made clear by the excavations recently carried out
in the great buildings of Heliopolis, which I have had an opportunity of examining
personally. 3

The external appearance of the basilica, with its gabled roof, the fa$ade
with its three doors, and the rotunda, is represented in the important mosaic from
the church of Madaba, which gives a map of Egypt and the Holy Places, and is
considered to belong to the early years of the Vlth century, 4 or, more probably,
to. the time of Justinian (527-565). The church of Madaba, with its apse
semicircular both internally and externally, shows that, at the time when it was
built, the Ravennate plan of apses polygonal on the outside had not yet penetrated
to the region east of the Jordan. In other parts of Syria, too, long after Christianity
had become the official religion with Constantine, churches had apses in the form
of round half-towers, as for instance the basilicas of Tafkha (ascribed to the
IVth or Vth century) and Sueida (believed to be of the Vth). In the Syrian lands
it seems that apses, curvilinear internally and polygonal externally, did not appear
before the Vlth century. Two of the earliest examples are St. George at Ezra
(515-516) and the cathedral of Bosra (511-512).

In the mosaic referred to the artist represented in flattened perspective (" proie-
2ione ribaltata") the essential features of the individual buildings which made up the
vast complex of Constantine's sanctuary of the Holy Sepulchre, as they would appear
to a spectator standing facing the propylaea. He has omitted the cloistered fore-

1 Migne, Pair. Gr., vol. 89. Epistola Antiochi monachi Laurat Sabae abbatis ad Eustachium pracposilwn
monasterii Attalinae civitatis Ancyrac Galaliae.

3 Palladio, op. fit. l Puchstein, Schulz, Krencker, op. fit.

* PP. Cleophas et Lagrange, La m0sai</ut gJographiqtu de M&dabS.

LOMBARDIC ARCHITECTURE

court of the basilica, the enclosing wall, and Golgotha, which he was unable to include
in his view (Fig. 375). He shows the propylaea which precede the fore-court, with the
great flight of steps ; the pedimented facade of the basilica with its three doors ; the
gabled roof of the basilica; and, last of all, the dome of the Anastasis which is
represented in the same cadmium-yellow and orange-yellow tint used for the surface of
all the pediments of basilicas figured on the mosaic. The perspective of the main
body of the church hides the apse and a portion of the rotunda.

The fate suffered by the buildings connected with the Holy Sepulchre when
Jerusalem in 614 fell into the hands of the Persian king, Chosroes II (591-628) (who
was joined by the Jews of Ptolemais and the Galilaean mountains 1 ), is well known :
they were destroyed by fire. Afterwards, the patriarch Modestus set about rebuilding

them (616-626): "he raised up again the venerable
churches of our Saviour Jesus Christ which had been
burned ; viz. Calvary, and the Resurrection, and the
venerable sanctuary of the precious Cross, the mother of
churches, &c." 2

Some light, though incomplete, is shed on the
results achieved, by the evidence of an eye-witness,
Arculf. From the description which he dictated to
Adamnan, 3 and from the ground plan of the buildings
which he has preserved (Fig. 376), we learn that
between the Anastasis and the Martyrion was inter-
posed the church of Golgotha, having on its right
another church of the same form, dedicated to St.
Mary ; and we gather that the principal church was no
longer, as in the time of Constantine, the basilica but
the rotunda. Lastly, we learn that the rotunda (" mira
rotunditate ") was annular, and that its dome was
sustained by twelve columns " mirae magnitudinis."
At the east were two entrances, each with four
openings. In the centre of the building stood a
circular isolated tugurium or chamber containing the
actual Sepulchre.

Adamnan describes it as being " tota lapidea," from
which one would infer that the church was entirely
constructed of masonry, and, consequently, that the main walls survived the disaster
of 614, and were only restored by Modestus. But this was not the case, for
we learn from Eutychius 4 that the dome seen by Arculf was of wood, and
that it was restored by Thomas, patriarch of Jerusalem, who imported the materials
from Cyprus. And so Constantine's church of the Ascension, also rebuilt by the
patriarch Modestus, which was in the form of a rotunda with concentric aisles,
had a wooden roof except over the central space, which was open to the sky.
Another VI Ith century building at Jerusalem with a wooden roof was the octagon
known as the Mosque of Omar, erected in the precinct of the ancient Jewish Temple
between 684 and 687 by the Caliph Abd-el-Melek. For though the existing dome

1 Rampolla del Tindaro, Santa Melania Gitniiore senatrice Kotnana.
- Migne, Pair, Gr. vol. 89. Epistola Antiochi monachi.

3 Tohler et Molinier, op. cit. Arculji relatio de Loci's Saudis, scripta ab Adamnano (about 670).

4 Migne, Pair. Gr. vol. Ill Etitychii patriarchal Alexandrini annales (Xlh cent. ).

Fig. 375. Portion of Mosaic in
the Church of Madaba (Vlth
Century).

BURGUNDY

is the result of a rebuilding (1022) of the original one destroyed by an earthquake
(1016), still it must have been copied from it. It has also been observed before now 1
that there were other ancient buildings in Syria with wooden domes, as the thinness
of the supporting drums indicates. Among them was the cathedral of Bosra ; and
the same is said to have been the case in the church of St. George at Ezra, where
the present conical cupola of light concrete construction resting on raccords (the
rest of the church being built of dressed stone set without mortar) is clearly of later
origin.

A short description of Holy Places in Palestine, thought to be of the VI Ith century,
incorporated in the " History of Agvan " compiled by Moses Kagankavatsi probably
at the end of the Xth century, 2 throws additional light to that provided by Arculf
on the round church of the Resurrec-
tion, and some quite new light on the
basilica of the Invention of the Cross
or " Martyrion." The domed rotunda
was one hundred cubits both in
diameter and height. It contained
two concentric circles of columns, with
above them two ranges of twelve
columns each. The Martyrion, sepa-
rated from the Anastasis by a space
of at least twenty cubits, also had
ranges of columns one above the other,
the number of shafts being sixty-five
or seventy-five. This limited number
of supports, in the case of a basilica
with ranges of columns in two stories,
suggests the small proportions of the
Martyrion as restored by Modestus on
a diminished scale ; and this explains
why Adamnan in his plan gave more
space to the rotunda as compared with
the basilica.*

These structures are mentioned

about 720 by the Venerable Bede. 3 It is known that in 812 they were sacked by
the Arabs from Egypt, and that the patriarch Thomas restored them between
813 and 833 in the days of the Caliph Mamun. It was after his restoration that
they were visited by the monk Bernard. 4 We know, too, that they were damaged
by a fire in 936. But the most serious disaster happened in 1010. Hakem, the
Fatimite Caliph of Egypt (996-1021), commanded the destruction of the rotunda;
and the order was so scrupulously carried out by the governor of Ramleh that, as we
learn from William of Tyre, 5 Radulphus Glaber," and Ademar, 7 it was levelled with
the ground. A new rotunda was erected in 1048 by direction of the Emperor

1 Texier and Pullan, op. (it.

2 Palestine Exploration l-'una Quarterly Statement, October 1896: Nisbet Bain, Armenian description of
the Holy Places in the seventh century.

3 Tobler et Molinier, of. eft. De Locis Sanctis (about 720).

* Id., op. fit. Itinerarium (about 870). * Op. (it.

" Delisle, Historiae sui temforis. De eversione templi Hicrosolymorum et caedejudaeortim.
~ Id. , Ex chronico Ademari Cabanensis.

A. Tegurium rocundunu.

B. Scpulchrum Domini.

C. Altiria dualia^

D. Alcariijt
. Ecclcfu..

F. Golgothina Ecclefiru.

G. Inloco AlraruAbrahinu.

H. In quo loco Crux Dominica cam bmii I.atronum_
I. MenfiligncL. (crucibuffiibcerrarcpcrtaclt
K Ptatcolitin qua die ac noQe lampadcs ardent.
L. SanSr Manx Ecclcfia.
M. Conftinnnilna Bjfilici.hoccftMirtyrmmj.
W. Eiidra cum Calicc Domini.

Plan of the buildings connected
Century.

Fig. 376. Jerusalem. Plan ot the buildings
with the Holy Sepulchre in the VI Ith
(From Gretser, " Adamanni Scot 'o A i 'term' dc
Terrae Sanctae.")

LOMBARDIC ARCHITECTURE

Constantine X Monomachus (1042-1054). This was built "politis et quadris
lapidibus," according to the statement of Radulphus Glaber, and was supported

by twelve monolith
columns and six piers.
It had six doors, and
galleries with sixteen
columns. The roof
was of wood, with a
circular opening in
the centre. Next to
it on the east was an
atrium containing
^. various chapels, later

put under cover by
the Crusaders and

HP iiLliR^ : connected with the

rotunda, thus forming
the basilica of the

=^\ ^ Holy Sepulchre.

. siBfe!g ' ' '

Other chapels flanked

it on the south. 1234
The date at which

Fig. 377. Jerusalem. Rotunda of the Holy Sepulchre in 1586.
" // devotissinw viaggio di Gerusalemme.")

(From Zuallardo,

the buildings connected with the Holy Sepulchre began to undergo the alterations
made by the Crusaders is not precisely known. It is, however, certain that it was
later than the years
i 102- 1103 and
1 1 06- 1 1 07, the dates
of the pilgrimages
to Jerusalem of
Saewulf and of the
Russian Hegu me-
nus Daniel, respec-
tively. The latter
credits the Crusa-
ders only with the
construction of the
isolated chapel or
tiigurium over the
tomb of the Re-
deemer. And the
alterations must
have been effected

Fig- 379- Jerusalem. Facade of the Church of the Holy Sepulchre in 1586.
(From Zuallardo, " // devotissimo viaggio di Gerusalemme."}

shortly before the
decade 1155-1165,
for we gather from John of Wurzburg 5 that the construction of the new church, and

1 Bourgogne et Martinet, Kelatio de peregrinatione Saewulfi ad Hierosolymam et Terram Sanctam
A.D.I. MCII et MCIII.

- De Khitrowo, op. cit. Vie et pelerinage de Daniel, higoumene russe (1106-1107).

3 Tobler et Molinier, op. cit.Qnalitersita esl Jerusalemt(\xfoie 1096). William of Tyre, of. cit.

5 Tobler, Descriptiones Terrae Sanctae Johannes Wirziburgensis, Descriptio Terrae Sanctae.

BURGUNDY I?

the alteration of the rotunda necessitated by putting the two buildings in direct com-
munication, were a recent event. As a matter of fact the dedication took place in I r43. 1
For the church thus produced by the conjunction of the rotunda of Constantine Mono-
machus (1048) with the new church (Xllth century), the principal entrance was made
on the south, approached through a " mout bel place" 2 alas, how different nowadays !
In course of time the sanctuary was remodelled, tampered with, partially
rebuilt, and for the most part concealed by structures of every description. In the

Fig- 378. Jerusalem. Fasade of the Church of the Holy Sepulchre (Xlllh Century).

rotunda, of which I give a view taken in 1586 (Fig. 377), the three lofty semicircular
recesses can only just be made out. The one to the south has been broken through
at the bottom, and each has three windows in the half-dome.

A conspicuous piece of evidence about the church, in spite of its injuries, is the
facade (Fig. 378). It was all built at one time, and of tufa, not limestone like the
Constantinian building, remains of which may be seen in the Russian Convent. The

1 Mariti, op. cil.

'* Mirlu-lant ct Kaynauil, ftim'raires A Jerusalem tt description de la Tcrrc-Saintc.rnou\, L'tstat de la ct'U
de Jhrrusaiem (about 1231).

Vi it.. II c

LOMBARDIC ARCHITECTURE

of the Xllth century a uniform patina, alike in the

The de exhibits four archivolts with fluted voussoirs, recalling the Futuh

the Crusades), 6 though in reality
it cannot be older than the taking
of Jerusalem by the Crusaders
(1099). This is proved by the
compound piers, designed, like the
corresponding half-piers and ex-
ternal buttresses, under the potent
influence of the Lombardic School ;
and also by the Lombardic door-
way in the fa9ade. More evidence
is provided by a capital (supposing
it to be old) in the transept, carved
with the motive of shallow trun-
cated and inverted half-cones,
with pointed leaves in the free
spaces, which was imported from
the West and may be seen in the
abbey church of St. George at
Boscherville(XIth and Xllth cen-
turies). And in any case it is not
credible that the architect of St.
Anne, who came, it seems, from
France 7 (it has been suggested
that the building was of local
execution, but that the plans were
brought from Europe), endowed
Palestine with a more developed
system of construction than ex-
isted at home. And in St. Anne (where the arches are pointed) the transept has
barrel vaults coming up to the central conical cupola, which rests on spherical
pendentives, while the nave and aisles have cross vaulting with visible transverse
arches So that we shall be within the truth if we fix its date (as De Vogue 8 has
already done) in the second half of the Xllth century, and, to be more exact,
after Judith, daughter of Baldwin II (1118-1131), had taken the veil there (1130).

1 Tobler, op. cit. La cites dejherusalem (about 1187).

2 Michelant et Raynaud, op. cit. Ernoul, op. cit.

3 Id le continiiateur aiwnyme de Guillaume de TyrLa sainte citi dejherusalem, etc. (1261).

* Libcllu: de Loch Sanctia. 5 Zuallardo, 11 devotissimo viaggio di Gerusalemme.
Mauss, La piscine de Bethesda a Jerusalem.

' De Luynes, Voyage oT exploration a la Mer Morte, a Petra, et sur la rive gauche du Jourdain.

* Les Aglises de la Terre-Sainte.

Fig. 380. Jerusalem. Church of St. Anne (Xllth Century)-

BURGUNDY

Fig. 381. Jerusalem. Church
of the Holy Sepulchre
Capital in the chapel of
St. Helena ( Vlth Century).

The building, partly constructed of materials which bear

masons' marks belonging to the age of the Crusades, was

not long after (1192) turned into a school by the great

Saladin, who had recaptured Jerusalem in 1187.

After this short digression we resume our subject.

The existing Church of the Holy Sepulchre, so far as one

can judge from what is visible, does not contain any decora-
tive fragment from the first foundation. Thus the four

capitals with foliage and cauliculi springing from a basket

of woven wickerwork, now used in the chapel of St. Helena

(Fig. 381) and mutilated in order to make them fit, are

detached specimens of the epoch of Justinian, analogous

to similar ones
in the mosque
" El Aksa,"

taken from the magnificent basilica of the
Virgin completed by Justinian l and de-
scribed by Procopius. 2 Again, the de-
tached Corinthian capital having sharply
cut acanthus leaves with the points turned
over and a carved torus (Fig. 382), now to
be seen on the ground floor of the northern
limb of the church, belongs to the same
period, as shown by its close relationship
to Justinian's Corinthian capitals in the
church of the Virgin just mentioned, and
also to the capitals of the same order
belonging to the so-called Golden Gate
near that church.

The same thing is true of the Byzan-
tine capital formed of a basket of wicker-

t ig. 382. Jerusalem. Church of the Holy Sepulchre.

Capital (Vlth Century). work,

C O m -

pletely undercut, from which spring cauliculi at

the sides of a wreath, with leaves below, to be

found close to the other one we have just described

in the Church of the Holy Sepulchre. This, too,

may be compared with the similar ones in the

mosque " El Aksa " (Fig. 383), and is to be ascribed

to the Vlth century, and in fact to the long reign
of Justinian I (527-565), when magnificent con-
structions made Jerusalem the most splendid city
of the East after Constantinople. This basket
capital of the Holy Sepulchre must be distin-
guished from another of the same kind (Fig. 384)
surmounting a half wall-column at the entrance
to the court in front of the church, because, in
spite of the similarities between them, they differ in composition, design, and
1 Conder, The City of Jerusalem. ' Corpus script, hist. byz.De aedificiis dn. Justiniani.

C 2

383- Jerusalem. Mosque " El Aksa."
Capital (Vlth Century).

20 LOMBARDIC ARCHITECTURE

execution. Moreover, the latter is carved in the same sort of stone as the facade,

and belongs to the same date.

In the Vth century at Jerusalem the art of carving would not have been equal

to the production of the capitals described above ; not to speak of the fact that it was

only in that century that the basket capital was
created by the School of Salonica. One of the
earliest examples at Jerusalem would be a capital
which is believed to have belonged to the church of
St. Stephen, erected by the Empress Eudoxia, wife
of Theodosius the Younger, and dedicated by her in
460, 1 supposing it were undoubtedly proved that the
basilica with nave and aisles recently discovered
really goes back to the days of that empress.

Moreover, in the IVth century, that is to say
in the reign of Constantine the Great, the design
and carving of capitals followed the type still to be
seen in the Church of the Nativity at Bethlehem
(Fig. 385), founded by the Empress Helena (327) but

Fig. 384. -Jerusalem, church of the erected by Constantine. For the existing basilica,

with its nave and double aisles, ending in a choir
with three apses (a Latin plan, previously employed

Holy Sepulchre. Capital in the
entrance court (Xllth Century).

in the basilica of St. Paulinus at Nola, which must have been built after 394, the year
when Paulinus settled in the " Cemetery " of Nola, for the work was finished by 402
roofed with timber in all parts except the three apses, lighted by large round-headed
windows some of which have been blocked up or altered, is the result of a remodelling

Fig- 385. Bethlehem. Nave of the Church of the Nativity (327).

carried out under Justinian I, and not of a complete rebuilding, which that emperor
had indeed in his mind but did not carry out. Hence his anger with his deputy,

1 P. Lagrange, Saintc-6tienne et son sanctuaire (I Jfriisalem.

BURGUNDY

whom he caused to be beheaded. 1 Th:it result is fortunate for us, as it has
preserved the nave colonnades of the original church up to the point where the choir
begins.

In them the shafts are formed of a yellowish stone from the local quarries,
and were made, with the bases and capitals, expressly for the church. The bases
are Attic ; the capitals Corinthian, the body slightly moulded into a bell shape, but
without a torus forming part of it. The acanthus leaves have not their tips arched
over, but still retain a classical character, while the flower on the abacus is marked
with a cross. The execution is so uniform that they seem to have come not merely
from one stone-yard but from
a single chisel, wielded by a
hand which was fairly skilful
though somewhat lacking in
power.

These colonnades cannot
be assigned to the age of
Justinian, for, like the nave
of the old Vatican basilica
(IVth century), they are sur-
mounted by architraves and
not by arches. Nor can the
capitals. Very different were
the capitals of the days of
Justinian in Palestine, as those
of the mosque " El Aksa " and
of the so-called Golden Gate
at Jerusalem (Fig. 386) tell us.
De Vogue 2 thought that this
structure might belong to the
Vth century, or at latest to
the Vlth; but it must be
assigned to the latter, and in
fact to the time of Justinian,
that is to say to the golden
age of Byzantine art. The
determining feature is the ob-
vious relationship between its Corinthian capitals and those of the same order and
style in the mosque referred to.

It has been suggested, 3 on the strength of the description of the Holy Places
referred to before as contained in the " History of Agvan," 4 that the existing church
of the Nativity retains hardly anything of the edifice of Helena and Constantine, and that
the latter was vaulted, like the Basilica Nova at Rome (the work of Maxentius and
Constantine), and afterwards destroyed by fire during the revolt of the Samaritans in
642. But the stone vaulting mentioned in that description must refer to the choir
apses. Besides, it is not even remotely conceivable that a colonnaded basilica could

1 Migne, Pair. Gr., vol. in. Eiitychii patriarchae Alexandrini annaies (Xth cent.).

a Lf Temple tie Jerusalem.

3 P. Barnabe (of Alsace), Le prttoire de Pi/atc et la forteresse Antonia.

1 Palestine Exploration Fund, loc. cil.

Fig. 386. Jerusalem. Golden Gale (Vlth Century).

LOMBARDIC ARCHITECTURE

ir rv the very heavy vaulting of that period ; and the same account speaks of ninety
a bl co uls One has only to think of the complex, ponderous system of supports
Ich th Trchitect of the Basilica Nova was obliged to adopt '- mg h,s
-omnlete roof of masonry, in order to see that my v.ew agrees with the facts,
also onfi med bv the circumstance that the architect of the nearly contemporary
Blsilca JuTia in'the Forum at Rome, having employed simple cruaform piers
vlt d his aisles, but did not venture to do so for the nave, and had to be conten
vi h a "ooden roof. When the bui.ders of the Constantinian age did construe
vaultin- over colonnades, it was only in the case of arcular bu.ld.np. And on the
other hand the Samaritans would never have been able to destroy by fire a structure
covered! like the Basilica Nova, with barrel and cross vaults of great thickness,
with the roof tiles resting directly on the extrados.

And now we will conclude our story of the Holy Sepulchre with a few brief

observations.

The plan of a tomb
standing close to a church
erected over a place of mar-
tyrdom is a Roman idea,
followed at Rome by the
Byzantines themselves, as is
shown by the Imperial Mauso-
leum which formerly stood to
the left of the old St. Peter's
in the Vatican. It is also
illustrated by the sepulchral
church or mausoleum of St.
Helena on the ancient Via
Labicana (now Casilina) (Fig.
387), erected in a region where,
among numerous other sacred
memorials, stood the tombs of
Saints Peter and Marcellinus,
in the cemetery "ad duas
lauros," over which Constan-
tine, after his official recognition of the Church (313), built a basilica in honour of
the two martyrs. 1234 This structure belongs to the same class as the magnificent
example in the Licinian Gardens at Rome (253-258), compared with which, though
the masonry of the walls is less finished, and even contains fragments of a dentilated
cornice, it shows a notable advance in the principles of construction. Thus the
organic structure of the cupola is different from that of the Licinian edifice. There
is no longer (as in the latter) a hemispherical vault, with its framework composed of
radiating ribs meeting in the crown, resting on a lofty polygonal drum lighted by
large windows, and strengthened on the outside by powerful buttresses corresponding
to the re-entrant angles inside, and raised somewhat by means of steps above the
impost of the cupola in order to be the better able to resist its thrust. Here, on the
contrary, we get a vaulted dome, lightened by concentric rings of amphorae. Owing
to the relief which this provided, the dome of the mausoleum of St. Helena was able

1 Aringhi, Roma subtcrranea ntmissima. * Caetani-Lovatelli, Varia.Una gita a Tor Pignallara.

3 Marucchi, La cripta storica dei santi Pietro e Marcellino.

4 Tomassetti, Delia campagna Romano. Archivio della R. Sociela romana di storia patria, 1902.

Fig. 387. Rome. Mausoleum of St. Helena (IVth Century)-

BURGUNDY

Fig. 388. Jerusalem. " Tomb of the Judges" (1st Century A.D.).

to be set up, without any
buttressing, upon a lofty
drum, lightened by niches
on its exterior, and pierced
by windows. And this
fixes its date to a period
later than the Licinian
Nymphaeum, in other
words to the early years
of the IVth century, and
after the death of Maxen-
tius (312), when Constan-
tine the Great was un-
disputed master of the
Empire. The edifice was
the regal monument des-
tined to receive the body
of his octogenarian mother; for it is to Rome, as Duchesne has proved, 1 that Eusebius
refers in his Life of Constantine when he mentions the burial, seeing that in those
days the Emperor had not yet turned his eyes to Byzantium. 2 The works necessary
to convert the latter into the seat of government were only begun in 328, and Con-
stantinople was not dedicated till 33O. 3

The aisled rotundas of the West were
not copied from the Holy Sepulchre at
Jerusalem, i.e. the earliest example of the
(fF form which the East can show, as is the

general idea. Rather they were adapted from
j* , the plans of Nymphaeums and tombs at

Rome, which was their place of birth, and
formed a natural centre for their development,
since no other city, or rather country, could
ever show so large, or so varied, or so imposing
a series of circular buildings, evolved from the
germ of the round Etruscan tomb. And to
those structures at Rome all others of the kind
in the Roman Empire have to give place.

We are acquainted with the form given to
sepulchral monuments in Palestine and the
other districts of Syria before the time of
Constantine. There is, for instance, among
many others, the so-called " Tomb of the
Kings " near Jerusalem, believed to be the
burial place of the queen of Adiabene (about
the middle of the 1st century A.D.). It is cut

g- 389- Jerusalem. " Tomb of Absalom '
(1st Century A.D.).

out of the rock, and the front was originally supported by two columns and decorated
with festoons of leaves and fruit and with palm branches. There are also the so-
called "Tomb of the Judges" (Fig. 388) and the "Tomb of Absalom" (Fig. 389), both

1 Le liber fontificalis.

2 Bosio, Roma sotterranea.

Van Millingen, Byzantine Constantinople.

24 ' LOMBARDIC ARCHITECTURE

regarded by De Vogue 1 as belonging to the three centuries before the destruction
of the city by Titus (70), while others think that they belong to about the first half of

the 1st century A.D. 2 Then there is the
tomb of Roman type at Kusren Nueijis
(Fig. 390) of the Ilnd century, and
that of Hamrath with square base and
stepped pyramidal roof at Sueida,
thought to belong to the end of the 1st
century B.C. Next comes the tomb of
Aemilius Reginus (195) at Katura, sur-
mounted by pairs of columns. Lastly,
there are the tower-tombs, the most
remarkable specimens of which are to
be seen at Palmyra (Fig. 391) ; and the
typical examples at Petra (the most
important being the Khasneh Firun or
Fig. 390. Tomb at Kusr en Nueijis (Ilnd Century). " Treasury of Pharaoh ") belonging to

(From a photograph furnished me by the " Palestine ^ Qf thg Emp j re Any Qne who

Exploration /-una. ) J

can find earlier example in this part

of Asia, or in Greece or any other Mediterranean country, of structures of the type
of the Licinian Nymphaeum, or of the Mausoleum of St. Helena, and of aisled
rotundas with solid cupolas resting on isolated supports, like Santa Costanza, will
be heartily welcomed. I was never
so fortunate.

In the Asiatic provinces, vaulted
and domed rotundas with aisles were
an importation from Rome, and
originally were roofed with timber.
This is shown by the celebrated
church erected by Constantine in the
middle of Antioch, which, according
to Eusebius, 3 was octagonal, with
galleries round the interior. It is not
certain whether these were used as a
matroneum, seeing that (as De Vogue
also noticed 4 ) it may be gathered
from a homily of John Chrysostom,
presbyter of Antioch, and afterwards
patriarch of Constantinople (398-404),
that in the churches of the old seat
of the patriarchate of the Eastern
Church the men were placed on one
side of the building and the women

on the other, in enclosures formed by Fig. 391.-

barriers of wood or more permanent
material. 5 This church at Antioch, built probably after the death of the Empress

1 Le Temple de Jerusalem. - Phene Spiers and Anderson, The Architecture of Greece and Rome.

3 Vita Constantini. < Syrie centrale.

5 Migne, Patr. Graeca, Vol. &.Homiliae in Malthaeum.

BURGUNDY

Helena, and even after the dedication of Constantinople (330), as it is mentioned by
Kusebius after those events, must have been entirely ceiled with wood, and devoid
of vaulting and a central cupola. As a matter of fact, Chrysostom, preaching in the
principal church at Antioch, rebukes his hearers for finding his sermons too long,
seeing that they are comfortably seated with a magnificent ceiling over their heads,
whereas they were content to stand for a whole day in the circus, exposed to sun
and rain, without ever complaining of the length of the performance.

Among the examples which still exist, the mausoleum at Rome known as Santa
Costanza, erected between 326 and 329 l (Fig. 392), is to be looked upon as the proto-
type of the vaulted aisled rotunda. This structure, which, as Duchesne says, 2 both
research and the best accredited opinions refer without a shadow of doubt to the time
of the Constantinian dynasty, is separated by only a short interval from the sepulchre
of St. Helena. The cupola
of the latter and its circu-
lar drum (both rising in
stepped outline from a
ground floor of similar plan,
with exedras) have only
to be set on an open arcade,
and you have the aisled
rotunda of Santa Costanza.

And any one who has
the wish and the ability to
investigate the difficult sub-
ject of Roman architec-
tonic science in what is re-
garded as the age of de-
cadence, but really marks
the culminating point of
that science, as though the
architecture of Pagan
Rome, before sinking into
the long slumber from

which it was to be roused at a later time by the efforts of the Popes and the
surrounding artistic influences, concentrated itself in one last ray of dazzling light-
any one, I say, who makes that investigation will readily perceive how, in the course
of the strivings of the " Urban " builders after the solution of the most difficult
problems of equilibrium, the architect of the Basilica Nova had already indicated to
his successor of Santa Costanza the conception of raising a great central dome on
isolated supports by the aid of barrel vaults.

The rotunda of Santa Costanza another instance of a circular mausoleum close
to a basilica erected by Constantine, in this case (326-329) over the tomb of
St. Agnes in which the masonry in the original parts of the facing, formed of bricks
of every quality and size, with thick joints, resembles that of St. Helena, must have
been built before the works at Constantinople were begun. To take part in those
works, destined to fit Byzantium for becoming the seat of government, craftsmen
were invited (328) who would represent the best from every part of the Empire. In

1 De Rossi, Mitsaici cristiani delle chiese tii Kama anterior! al seco/o XI'.
5 Le liter pontificalis.

Fig. 392. Rome. Santa Costanza (IVlh Century).

LOMBARDIC ARCHITECTURE

Lydus 1 we read of a portico said to have been built by Campanians and marble
workers from Naples and Puteoli. And it must have been finished before the Roman
School, already weakened when Maximian made Milan his capital, was, so to speak,
deprived of its vitality by the drain caused by the works of " New Rome," and before
the progress in the science of construction and equilibrium, which had reached its
climax in the days of Diocletian (284-305), Maximian (286-310), and Maxentius
(306-312), had received its death-blow. This fatal event readily explains the sudden
arrest of the splendid development which was leading to ever newer and bolder
systems of vaulting systems now recognized as worthy of the rulers of the world --
and the appearance of structures, of great dimensions indeed, but made up of old
materials, and only roofed with timber. Such were the first great Christian basilicas,
or occasionally a public market like that on the Celian at Rome (364-383), later
converted into the church of Santo Stefano Rotondo (468-483).

Any one who cares to give even a
passing glance at the illustrations of Mon-
tano s or Bramantino, 4 or the Vatican draw-
ings, 5 will find specimens of every kind of
Roman sepulchral annular rotundas. Some
have a central cupola buttressed by barrel
vaults, with, sometimes, in the basement of
the outer wall, a series of curved or rectan-
gular recesses ; while the inner concentric
circle presents in some cases twelve or sixteen
columns, either single or in pairs, on a
common plinth, in others eight cruciform
piers, in a third kind twelve square piers
with half-columns. And if he wishes, he
may at the same time verify, specially in
Palladio 6 and Serlio, 7 the fact that the build-
ings of ancient Rome offer in their amazing
variety (due to a large extent to the use of
concrete) every one of the plans which we
are told must have come from the East, but which, on the contrary, the East
borrowed from the West. For if the reverse had been the case, the East ought to
have been able to show all that variety of plan in examples of earlier date than
those of the Rome of the Empire ; and this is certainly not the case.

Before leaving this subject of the Holy Sepulchre, I must draw attention to two
sketches of ancient Roman circular buildings which I have noticed in the Uffizi, and
have had photographed for the first time (Figs. 393, 394). One of them is very
interesting, not only because the internal facing is entirely in brick, but also on
account of the form of the piers of the arcade which carries the cupola, and of
the construction of the cupola itself. The latter shows on its intrados bands which
intersect and form lozenge-shaped spaces filled in with horizontal brick courses.

And now to resume our subject. More than one feature of the church of Saint
Benigne is clearly derived from preceding buildings.

I. The rotunda in several stories is a direct descendant from the tombs with a

Fig. 393. Sketch of circular Roman building.

1 Corpus script, hist. by*.De Magistraiibus.
1 Op. dt. < Of. cit.

7 Op. eft.

Op. fit.

- Blomfield, The Mistress Art.
5 Vatican Library. Cod. Lat. 3439.

BURGUNDY

central tower of two or three stages, roofed with a circular vault (sometimes supported
by internal wall piers), of which Rome affords so many examples. Any one who
wants to verify this may go for examples to Montano l and Bramantino. 1

II. The termination of the basilica, with its lateral niches and the rotunda behind,
recalls the plan of the Pantheon at Rome (120-124), which must have been before the
mind of Abbot William when making his design, seeing that the rotunda was dedi-
cated to the Virgin (the basilica had been previously dedicated to St. Benignus)
precisely on the anniversary day of the consecration of Hadrian's famous structure to
"Sancta Maria ad Martyres," on its conversion

into a church by Pope Boniface IV (608-615).

I 1 1. The arrangement of towers incorporated
with the facade had been a feature of the Cluniac
.-ystem from the time when Abbot Majolus intro-
duced it in Saint Pierre le Vieux (982). The
two round staircase towers flanking the rotunda,
and its encircling aisles and gallery, are suggested
by San Vitale at Ravenna (526-547).

IV. The idea of a tower raised over the
crossing (applied shortly before in Saint Pierre
le Vieux at Cluny) is derived from structures
such as the mausoleum of Galla Placidia at
Ravenna (about 440), or, more probably, from
the ancient church of Santa Croce (about 449) in
the same place. In France it was used from
very early times, its presence there in the Vlth
century being disclosed by passages in Gregory
of Tours and Venantius Fortunatus.

I ought here to put on record the fact that
the oldest example that can be produced of a
church with a lantern tower rising ever the chan-
cel, and, moreover, with a solid roof, is that of San Salvatore at Spoleto.
convenient to give a short account of it at this point.

394- Sketch of circular Roman
building.

It will be

THE BASILICA OF SAN SALVATORE OR DEL CROCIFISSO, ix THE CAMPOSAXTO
AT SPOLETO, had been dedicated to the Saviour by 815, but was known as San
Concordio in 1158.* It is a basilica with nave and aisles, which were originally more
lofty than at present, the nave having been supported at first by columns carrying an
architrave, after the fashion still to be seen in the presbytery (Fig. 395). It was also
provided with low galleries, like Santa Maria in Cosmedin at Rome after Hadrian I's
(772-795) rebuilding, as may be inferred from the remains of arches in the north wall
of the nave The apse, raised above the presbytery, is flanked by two square
sacristies, each provided with a small apse, which is a later addition. Above the
presbytery rises a tower (Fig. 396) strengthened by buttresses at the corners. The
transition from the square to the octagon is managed by means of four clumsy
triangular raccords, almost like sloping pieces of wall. Everything suggests that
they were made to carry a spherical dome, for which the present octagonal cupola
has been substituted. The facade is finished off by a pediment and two half-pediments.

1 Of. dt. * Of. cii.

* Sansi, Degti tdifti < dei frammentt slorifi delle antifHe eti <ti -S/Wcrf*.

LOMBARDIC ARCHITECTURE

Its upper part is decorated with four pilasters, reaching nearly to the cornice. The
capital of one of them has been found, and is now kept in the church.

There has been great variety of opinion about the date and origin of this very
ancient church. Thus, for instance, Hiibsch 1 thinks it belongs to the beginning of
the Constantinian age, and that it was a Christian church from its inception, though
the columns were taken from some Pagan building. But he is surprised to find the
square bay between the apse and the triumphal arch, with a cupola rising above it.
De Rossi, 2 on the other hand, regards it as the result of the conversion of a Pagan

temple into a church,
preserving the part
which forms the sanc-
tuary of the Christian
building, with the addi-
tion of the nave and
aisles and their facade.
This transformation
will have taken place
in the time of Theo-
dosius I (378-395) and
his sons.

Grisar's 3 idea is
this. Originally a
Pagan building adapted
to Christian uses, it
was given its present
form in the Xllth cen-
tury. It contains no
traces of work belong-
ing to the Constan-
tinian, Theodosian, or
Gothic periods. The
cupola is neither Pagan
nor Early Christian.
The only possible re-
mains of the original
Pagan structure which
occupied the site are
the plain jambs of the
main door in the facade;
and these were either set there in the Xllth century, or else kept in their original
position.

In my view, this extremely important building is the work of one period, as is
shown by the original masonry, not excluding the fa9ade, which is intimately con-
nected with the nave, and forms an integral part of the basilica. It also possessed a
central tower with a solid roof, a fact revealed by the masonry, and also by the pains
taken by the architect to ensure its stability.

1 Op. cil.

2 Bull, di arch, crisliaiia, \fy\.Spicilegio d'archeologia cristiana nelf Uinbria Delia basilica del
Salvatore presso Spoleto.

3 Nnovo bull, di arch, cristiana, 1895. // lempio del Clitunno e la chiesa spoletina di San Salvatore.

g- 395- Spoleto. Church of San Salvatore or the Crocifisso.
(IVth Century).

Chancel

BURGUNDY

The period of erection must be rather early, in view, not only of the two sacris-
ties flanking the apse, but also of the architraves which carry the nave walls.
Additional reasons are, the rudimentary form of the pendentives by means of
which the square of the tower passes into an octagon, and the large round-headed
unsplayed windows. The period is, perhaps, that subsequent to the age of Con-
stantine. I say this, not because it was only then that Pagan buildings began to be
robbed of materials to build churches, for that practice had begun as soon as
Constantine had conquered Maxentius, as we see from, among other instances, Santa
Costanza outside the walls of Rome. My reason rather is the character of the
carvings executed expressly for the fagade. They certainly do not exhibit the
power of the chisels of the time
of Constantine, but, on the other
hand, there is not as yet the
poverty, hardness, want of clear-
ness both in design and execu-
tion, which characterise Italian
work of the Vth century. Nor
do they exhibit the typical fea-
tures (and this applies equally to
the capitals in the three windows
of the facade) of Vlth century
carving. It is enough to com-
pare the way in which the bead
and reel mouldings and the
ovolos are treated in either case.

The decoration of the door-
way in the front of the church
inspired in after times, among
others, Melioranzio, the artist of
the well-known central entrance
to the cathedral of Spoleto
(Xllth century), who, though he
may have been superior in force
to the carver who worked at
San Salvatore, was inferior to
him in delicacy of execution.

The adoption in this singular
building of the Ravennate plan

of an apse at the eastern end, of which the Basilica Ursiana at Ravenna (370-384)
is the prototype, was made necessary by the nature of the site.

To return to Saint Benigne :

V. The arcaded choir appears to be derived from the very early open apse of
San Sebastiano outside the walls of Rome (366-384), which is the prototype of this
arrangement. Or it may have been suggested by the apse with three arches
opening into an ambulatory, in the basilica of Severus at Naples (367, and about
387). Other open apses that may have formed the model are those of the basilica
of St. Paulinus at Cimitile near Nola (394-402) ; the basilicas of Santa Maria
Maggiore (IVth or Vth century) and SS. Cosma e Damiano (526-530) at Rome ;
the Basilica Vincentiana, now San Giovanni Maggiore, at Naples (554-577) !

Fig. 396. Spoleto. Church of San Salvatore or the Crocifisso
(IVth Century).

30 LOMBARDIC ARCHITECTURE

Santo Stefano at Verona (Xth century) ; and, lastly, the cathedral of Ivrea (973-1001
or 1002).

VI. The introduction of galleries in the basilica merely indicates the adoption of
an arrangement which had become the fashion in Italy at the time, as applied in a
building of almost the same date as Saint Benigne, and certainly familiar to William
ofVolpiano I mean the cathedral of Ivrea. It was an arrangement which did not
originate then, any more than it originated in the days when San Salvatore at
Spoleto (IVth century) and the suburban basilicas of Sant' Agnese and the
SS. Quattro Coronati erected by Honorius I (625-638), or San Lorenzo by
Pelagius II (579-590), were provided with galleries. Nor did it originate under
Byzantine influence, as is commonly supposed.

It is indeed inconceivable that the Italian builders should have felt the need of
such influence, when their forefathers had, as long ago as the year 179 B.C., pro-
vided the Basilica Fulvia Aemilia with a gallery, as we are informed by the reverse
of a coin published by Babelon 1 and Cohen, 2 and by another reproduced by Hulsen. 3
Or, again, when they had before their eyes examples of civil basilicas, even divided
by piers into nave and aisles, and with a gallery above the vaulting of the latter, as
was the case with the Basilica Julia in the Roman Forum, which, even before its
rebuilding by Augustus (12 A.D.), possessed a gallery occupied, on the occasion of
important trials, by the two sexes separately. 4 At a later date, too, after its
reconstruction by Diocletian (284-305) and restoration by the City Prefect
Probianus in 416, it still exhibited its gallery carried by the cruciform piers and
the vaulting.

The fact about galleried basilicas is this. When the Easterns began to erect
Christian basilicas, and introduced women's galleries into them, as in the churches
at Tafkha (IVth and Vth centuries) and Kanawat (IVth century), they derived the
suggestion from the civil galleried basilica which had been developed (e.g. the
Ilnd or Illrd century example at Chaqqa) under the influence of Pagan Rome, which
was its creator.

The Chronicle of the Abbey tells us who was the architect and director
in chief of the works of Saint Benigne, so far at least as his many and various
duties and long absences allowed : " Et reverendus Abbas, magistros conducendo,
et ipsum opus dictando. . . ." 5 It also informs us that the prior Arnulf, from the
diocese of Toul, took part in the work of supervision. Many, indeed, from an
incorrect interpretation of another passage, would give this supervision, but restricted
to matters of decoration, to the monk Unald, of whom they make an accomplished
carver. But the Chronicle, though mentioning the assistance which he gave
to Abbot William ("ad omnia quippe quae sibi erant necessaria, praedicti fratris
iuvabatur solertia "), states that he was entrusted with the care of the church,
and performed his duties with such zeal that almost all the ritual ornaments were
gathered together by his efforts : in other words he was the Sacrist or Apocrisarius
of the church: " Denique iniunxit illi curam huius sacri periboli, quam tanta
prosecutus est cura, ut paene totum quicquid fuit ornamentorum in hac basilica, eius
studio sit aggregatum." .

As to the workmen, there are different opinions. If we are to believe Cordero,'*

* Description historique et chronolo s ique des monnaies de la Republique Romaine.

1 Description ginirale des monnaies de la Republique Romaine. 3 7/ F0ro Romano

* Lanciani, The Ruins and Excavations of Ancient Rome.

5 D'Achery, Spicilegium.Chronica S. Benigni Divionensis. e Of. cit

BURGUNDY 31

they were Italians. His opinion is based on the Chronicle. But although it
describes a remarkable emigration of Italians to Burgundy about the epoch of 1000
(" Coeperunt denique ex sua patria, hoc est Italia, multi ad eum convenire ; aliqui
litteris bene eruditi, alii diversorum operum magisterio docti, alii agriculturae scientia
praediti, quorum ars et ingenium huic loco profuit plurimum "), it makes it equally
clear what was the result, namely, to increase the number of monks under the rule
of William (" Crescebat ergo quotidie multitudo monachorum sub eius magisterio
degentium "). But there is no suggestion of Cordero's statement that William laid
the foundations of the new church of the monastery of St. Benignus " with the
aid of a band of Italian craftsmen."

Again, if we were to listen to Merzario, 1 the workmen in question were neither
more nor less than Comacine masters. But his view is based on nothing more than
the use in the Dijon Chronicle of the expressions " magistros conducendo " and
"opus dictando," which he thinks are taken from the CXLVth section of the Edict
of the Lombard king Rotharis (636-652), and on the assertion that, before the epoch
of about 1000, the Lombards, " with their methods, their formulas, their exclusive
predominance in the region of art," resorted to France to act as builders, and to
teach others to build ; and of that there is no proof.

On the other hand, our limited knowledge of the constructive, static, and
decorative elements of the original building does not give much help in answering
the question. However, I feel that we shall not go far wrong if we ascribe
the erection of the church to Italian master builders, associated with Burgundian
masons and workmen. Burgundy cannot have been entirely without such, for the
tradition of the art of building had never been interrupted there. Without going
back to still earlier times, as long ago as the partition (768) of the kingdom of
Pippin III (752-768) between his sons Charles (768-814) and Carloman (768-771),
opportunities of one sort or another had not been wanting to the craftsmen of those
countries for practising the art of building, either in the form of the erection of new
ecclesiastical edifices, or of the renewal and restoration of old ones. And this,
notwithstanding the frequent family and civil wars which followed the division
of the Empire made at Aachen in 817 by Louis the Pious (814-840), and culminated
in the period between the death of Lothair (840-855) and the deposition and death
of Charles the Fat (887). That partition marked in France (and equally so
in Germany) the extinction of the family of Charles the Great. For Charles
the Simple was held to be the bastard son of Louis the Stammerer (877-879), and
the last sovereign of this illegitimate line was Louis the Faineant (986-987).

Nor was building prevented by the raids of the Saracens or, what were more
serious, those of the Danes or Normans who, after the battle of Fontenay (841) had
opened the way for the destruction of the Frankish Empire, and the treaty of Verdun
(843) had brought it about, made themselves masters of most of the French rivers,
and spread terror, desolation, and death, in every direction.

Further, the builders of Burgundy had not been without opportunities of going
to Italy for training (if they felt the need of it) in the art of construction, in the days
of Louis III King of Provence (887-928), of Rudolf II King of Transjuran Burgundy
(91 1-937), f Hugo Duke of Provence (91 1-947) an< ^ his son Lothair (946-950), who
were elected Kings of Italy respectively in the years 900, 922, 926, and 946. And so,
the character of the barrel vaulting in the crypt of the rotunda of Saint Be"nigne
(rebuilt, as we saw, on the original lines) suggests the school of builders who con-

1 Op. dt.

32 LOMBARDIC ARCHITECTURE

structed the similar vaulting in the staircase of the campanile of San Benigno at
Fruttuaria (1003-1006). The arched corbel courses divided into groups, by lesenas,
which decorated the towers of the rotunda, point to the gilds of Upper Italy who
were the first to apply (in the tower of San Satiro at Milan, of 876) this form of
architectural decoration to towers, and had used it with good effect, not long before,
on the towers of the cathedral of Ivrea. I have searched in vain through France for
a tower of certain date with this decorative treatment, older than the rebuilding of
Saint Benigne.

The Pre-Lombardic cubical capitals are clearly the work of the school whence
came the chisels which wrought the nearly contemporary original capitals in the
crypt and ambulatory of the cathedral of Ivrea (973-1001 or 1002), and the one of the
same date in the crypt of the cathedral of Aosta (Xlth century).

Then there are the Pre-Lombardic figure capitals in the Lombardic style, the
earliest specimens of certain date to be found north of the Alps. In those countries,
from Merovingian times onwards, the only previous example I can point to is a capital
in the crypt of Saint Pierre de la Couture at Le Mans (997), with water leaves, those at
the angles being of crocket form, and on each face a human head, infantile in execu-
tion and design, taking the place of the flower. But these capitals, with their
representation of a man supporting the abacus, reveal a chisel from Cisalpine Gaul.
This motive was dear to the Lombard gilds, and was borrowed by them from the
Romans, who used to represent living figures supporting with head or hand the abacus
of a capital ; or they may have taken it from the Etruscans, who sometimes in their
designs of squares in carving figured a man supporting the frame above him with his
hands, as shown in Fig. 152, and again on a sculptured stone of the archaic Etruscan
period in the Archaeological Museum at Florence, in which one of the figures is a
telamon holding up the interlaced top of a square compartment.

All these details prove that Piedmontese craftsmen trained in the Lombardic
School took part in the works of Saint Benigne at Dijon, or possibly some
Lombard gild, but not one of the best, considering the exceptional poverty and
rudeness of the capitals in the rotunda compared with the markedly superior art of
capitals produced by those gilds about the same time, e.g. those in the crypt of the
parish church of San Vincenzo at Galliano (1007), and others in the church of San
Babila at Milan (Xlth century).

But, granting this, we cannot believe that Burgundian builders were not
given some share in the work, either restricted to duties of secondary importance, or
even entrusted with those of a higher order under the direction of experienced
master builders from the Italian side of the Alps. It is well known that the erection
of Saint Benigne was aided by the moral and material support of the Court of
Burgundy, of Bishop Bruno who was related to the reigning family, and of Majolus
the powerful abbot of Cluny. It is therefore quite reasonable to infer that local
craftsmen were invited to take part in the work, and the best, inasmuch as the subject
was a church regarded in those days, as Radulphus Glaber says, 1 as the most magnificent
in France, and demanding a degree of knowledge of the principles of construction
which was for that age remarkable. Still, the local builders could not have possessed
an experience equal to that of the craftsmen who came from the south of the Alps ;
otherwise there would have been no need to summon the latter to France.

The employment of Burgundian workmen was also made desirable by the
financial difficulties with which from time to time Abbot William had to contend in

1 Mabillon, Ada Sand. OrJ. S. Benedicti Vita S. Guilltlmi abbatis Divimensis.

BURGUNDY 33

the course of his operations ; difficulties which were certainly not of a character to
encourage him to rely exclusively on the more expensive services of builders of his own
nationality. The presence of skilled Italian workmen tells its tale as to the real
capacity of the Burgundian builders, and further informs us that, as they were not
capable of venturing upon operations which demanded a considerable knowledge of
the principles of construction, the religious buildings erected in their own country in
the days before they were reached by Italian artistic influence must have been of a
modest character so far as those principles were concerned.

In addition to builders, Burgundy must have contained carvers. Then there
arises this dilemma. Either the carving of the original capitals in William's crypt was
executed by Italian hands, and, as we said, certainly not hands of great experience ;
and in that case, barbarous as it is, it was beyond the powers of the local chisels, for
otherwise they could have been entrusted with the work ; or else it is the product
of French hands (not, of course, of the monk Unald, as we showed just now), and
then these capitals represent the best that French artists could do at the beginning
of the epoch of about 1000, and give us a standard of their ability in the treatment of
the human figure.

At the time of the erection of Saint Benigne, and indeed throughout the first
quarter of the Xlth century, the carvers both of Burgundy and of France generally
were at a very low level in the treatment of the figure in sculpture the capital in
Saint Pierre de la Couture at Le Mans (997) tells us how low. It is only by
unduly moving back the dates of buildings that so many writers have been able to
prove the contrary. Thus the imposing two-storied porch of the monastic church of
Saint Benoit sur Loire is dated by Gailhabaud l and others in the two years following
the fire of 1026 under Abbot Gauslinus, bishop of Bourges (1014-1020-1029), who
began the erection of a tower of squared stone which he was unable to finish. 2 But
the capitals carved with figures of realistic or fanciful character, and other figure
sculpture on the exterior (very different both in design and execution from, I will
not say the very unfavourable specimens in Saint Benigne, but from those in the
monastic church of Cerisy la Foret [1030-1066] which show some improvement),
point to a much more advanced period, certainly decidedly later than the first half of
the Xlth century. The fact is that, so far as the church is concerned, the work of
Gauslinus was confined to mere restoration, and the rebuilding had to be undertaken
by Abbot William who was cut off before he saw its completion : " Ecclesiam
multis incendiis devastatam et senio praegravatam novo iacto aedificare coepit
fundamento." The new structure was finished in 1 108 under Abbot Simon, and
King Philip I (1060-1 108) was present at the dedication. 34

There are writers who are surprised that Saint Benigne, being so conspicuous as
it was, did not serve as a model for many other churches. But this is easy to explain
when we remember that the form of the building, besides being too complicated,
was not that on which the Latin Church had set its seal. It was out of touch with
the artistic traditions of the Western peoples, nor was it the best fitted for the
requirements of Western ritual. And so it came about that even its own author
did not repeat it.

Then, there are others who would see in it the first or one of the first specimens
of Lombardo-Norman architecture in France. But this is not in accordance with the
facts. The result, as it was, of the Roman, Romano-Ravennate, Byzantino-Ravennate,

1 L 'architecture du V' au XVII' slide. * Delisle, Ex libello Hugonis FloHacensis moncuhi.

s Id. , Ex chronico coenobii 5. Petri vivi Senontnsis. * Id. , Ex libello Hugonis floriacensis moncuhi.

VOL. II D

34 LOMBARDIC ARCHITECTURE

and Pre-Lombardic monuments seen by its architect in the course of his long
journeyings in Italy, the church at Dijon, to judge by what we know of it, did not
present one of the original and distinctive characteristics of the Lombardo-Norman
style, which we shall specify by degrees in the course of our work.

Nevertheless it had the merit of providing its designer with the opportunity of
putting into practice on a large scale, and with a difficult problem to solve, his ideas in
the matter of construction, and the monks who were his assistants with experience in
directing the works of a great building. Thirdly, it gave the local workmen the
advantage of taking part in the erection of an edifice of exceptional importance for
that time and place, and of acquiring familiarity with the very difficult art of vault
construction, complex forms of which were here presented.

From its underground part there did originate one characteristic feature of the
Lombardo-Norman basilica ; I mean the chapels projecting from the arms of the
transept. It appears in the Norman and English churches erected from the plans of
William of Volpiano, or if not from his plans, certainly from those of his disciples and
under his advice. In other cases the plans came from his pupils, or from Lanfranc
of Pavia.

Here, too, was seen for the first time in France the Pre-Lombardic cubical figure
capital. Previously, and as far back as the time of Pippin III (752-768), the Pre-
Lombardic type had been seen, and even in a decorated form, but not with figures.
Instances are to be found in the crypt of the abbey church of Flavigny (755-768), the
churches of Germigny des Pres (801-806) and Saint Pierre at Jumieges (940), and the
chapel of Sainte Blandine in Saint Martin d'Ainay at Lyons (about 966).

Such facts, combined with other pieces of evidence, put into our hands the main
clues to the history of the buildings erected in Burgundy and the neighbouring
provinces before the epoch of the year 1000. Their rarity is due, not so much to
Saracen, Norman, or Hungarian ravages, as to the passion for innovation in the Xlth
century, referred to by Radulphus Glaber, 1 which spread over the whole of France,
and destroyed so many buildings in order to reconstruct them in a form more
consonant with the new fashions. It is true there are writers who will not admit
their almost universal disappearance. Thus, for instance, Revoil, 2 led astray by the
erroneous idea that the Lombardic style, which he regards as the source of the
Carolingian, had reached maturity centuries before it was born, has dated in the
centuries from the VHIth to the Xth a whole batch of religious structures in southern
France. But what must we think of attributions such as the following ?

The chapel of Saint Gabriel near Tarascon is regarded as belonging to the first
years of the IXth century, and identified with one mentioned about 858 in a charter of
Charles the Bald and his first wife Hirmentrudis. And this in the face of its pointed
barrel vaulting, and the rose window enclosed within a pointed arch in its front. Yet
the pointed arch did not appear in European churches before the second half of the
Xlth century. And so, not only in the first half of that century, and in the .same
district as Saint Gabriel, do we find the chapel of the Holy Cross at Montmajour, near
Aries (1018), still presenting only arches which are round and not pointed like the
transverse ones in the nave and the one in the gable front of Saint Gabriel, but as
late as 1063 the chapel of Saint Trophime near Aries was built with arches and vault-
ing of semicircular form. Rose windows, again, were not invented till the next century,
and that is the date indicated by the carving in the chapel.

1 Delisle, Historiarum sui temporis &c.De innovatione ecdesiarum in Mo orbe.
1 Architecture rornane dn midi de la France.

BURGUNDY 35

The abbey church of Saint Guilhem du Desert (originally the church of Gellone)
is still believed to be the one founded by William, Duke of Aquitaine, which must

F'g. 397- Saint Guilhem du Desert. Apse of church (Xllth Century).

have been finished in 806. Later it was restored or altered under Louis the Pious by
Juliofredus, its first abbot. 1 2 But a glance at the apse (Fig. 397) with its range of deep

1 Delisle, Ernioldus Nigellus ExulDe rebus geslis Ludovici PH.
* Id., Vita Hludawici Pit imperatoris.

D 2

LOMBARDIC ARCHITECTURE

arched niches, decorated in front with shafts and enclosing arches springing from
heads, or at its Lombardic portal, and the ribbed vaulting of the porch, is sufficient to
show that they are not earlier than the Xllth century. And as the side walls of the
nave (Fig. 398) have facing of different and less regular masonry than that at the end
of the building, and, moreover, as even the plan of the church suggests that there has
been some alteration, one may infer that the side walls are older than the Xllth
century. But their date cannot be that of the original foundation, for the church of
Duke William was a mere oratory paved with precious marbles, which formed a
beginning for his monastic foundation, and was hastily finished. 1 It was no basilica

with nave and aisles over 42 feet
wide. To settle the two periods
of construction we have the
assistance of two dates con-
nected with important facts.
The first is the year 1076, when
the altar of St. William was
solemnly dedicated in the pre-
sence of the Papal Legate
Amatus, and the consecration of
the church followed. 2 With this
date we may connect the re-
building of the church, which
took place after the confirmation
of the privileges (1066) and the
immunity (1068) granted to the
monastery. 3 The other date is
1138, when the second transla-
tion of the relics of the saint
took place.

The corbel courses grouped
by lesenas on the side walls of
the church suit well a period
not far removed from the first
date (1076). The ecclesiastical
buildings erected north of the
Alps in the days of Charles the
Great had external decoration of
that kind, as we may see from the abbey church of Saint Riquier or Centula (793-798)
according to a view of it preserved for us by Mabillon, 4 from the Palatine Chapel at
Aachen (796-804), and from Theodulfs church of Germigny des Pres (801-806).
The earliest certain example of the treatment in France is presented by the two
staircase towers of Saint Benigne at Dijon, and there only by virtue of Lombardic
influence. And it was not till the Xllth century that any very extensive use was
made of it, as on the magnificent bell-tower of Saint Theodorit at Uzes, and that of
Saint Trophime at Aries.

Nor shall we offend against logic or probability if we assign the alterations, for

2 Mabillon, Ada Sanct. Ord. S. Bencdictil'ila S. Willelmi ducis ac monachi Gelloncnsis in Gallia.

3 Sac. Arch, de Montfellier.Cartnlaires des abhayes cfAniane et de Gelhne.

4 Ada Sanct, Ord. S. Benedicti. Vita S. Angilberti abbatis Centulensis in Gallia.

Fig. 398. Saint Guilhem du Desert.
Century).

Side of the church (Xlth

BURGUNDY 37

instance, those in the apse and the porch, to the years immediately preceding
the second date, 1138. The decorative scheme of a range of deep arched niches,
applied to the apses of churches, rotundas, and baptisteries, created in the case of
Sant' Ambrogio at Milan between 789 and 824, continued, and that too in the land of
its birth, to lack the adjunct of shafts up to the time of its transformation into open
galleries. In the heart of the world that created them, there are only rectangular piers
for the ranges of niches in the basilica and baptistery of Agliate (824-860); the
basilicas of San Vincenzo in Prato (835-859), San Calimero (IXth or Xth century),
Sant' Eustorgio (Xth century), San Celso (996), and San Babila (Xlth century) at
Milan ; the Rotonda at Brescia (Xlth or Xllth century) ; the baptisteries of Biella,
(Xth century) and Novara (Xth century). For this reason it must have been long
after the erection of San Babila at Milan and of the Rotonda at Brescia, and only
when the Lombard gilds had transformed their ranges of arched niches into open
galleries with small shafts, exhibited for the first time in San Giacomo at Como
(1095-1117), that the niches of Saint Guilhem du Desert, with their compound
supports and enclosing arches resting on heads, can have been constructed.

As for the portal, we know that this kind of doorway had its beginnings about
the year 1032 in Sant' Andrea at Montefiascone. Shortly after the middle of the Xlth
century we find it in a still modest garb in Saint Etienne at Caen (1066-1086), and
at the beginning of the Xllth, when it had been discovered that it was more effective
in proportion as the orders were multiplied, in a fairly advanced shape and sumptuous
attire at Cluny (1089-1 130), Vezelay (1096-1132), and other churches.

Lastly, there is the ribbed vaulting in the porch. We know that cross vaulting
with diagonal arches or ribs makes its first unquestioned appearance in San Flaviano
at Montefiascone (1032), and that it was afterwards used in the cathedral of Aversa,
in Sant' Ambrogio at Milan, and in the church of Rivolta d'Adda, all structures of the
Xlth century. But north of the Alps it was not seen till about the end of that
century; and Durham Cathedral, the first stone of which was laid in 1093, affords the
earliest dated example.

The church of Saint Quenin near Vaison, in its oldest part the chancel is
supposed to have been built by order of Charles the Great or his successors ; but we
have only to notice the ribbed vault of the apse, and the crouching animal carved on
the keystone of the vaulting, to see that we have to deal with a building of the Xllth
century. It was towards the end of the Xlth century that the carving of keystones
in vaulting started in Italy. There is a rudely carved lamb at the intersection of the
diagonal arches in the nave of the church at Rivolta d'Adda. North of the Alps it
appears with the rise of Pointed architecture, and in some cases in the very first
churches in that style. Thus the cross vaulting with moulded ribs in the western towers
(i 134-1 144) of the cathedral of Chartres has the point of intersection left plain, while
the main vaulting of similar type in the cathedral of Sens (begun in 1140) is
ornamented by a rosette at the intersections.

It is appropriate to notice here that an archaic example of ornament applied
to a keystone of vaulting is afforded by the well-known Etruscan tomb of the
Volumnii near Perugia (believed to be of the Illrd century B.C.), where the head of
a Gorgon appears in the middle of the stepped squares which form the ceiling.

The cupola of the church of Notre Dame des Doms at Avignon, together with
the other parts of the structure, is thought to be a work of the time of Charles the
Great. Now this cupola rises from Campano-Lombardic hooded pendentives, a form
of Italian origin which did not migrate to other countries till about the middle of the

LOMBARDIC ARCHITECTURE

Xlth century, when it had reached its full development. It appears as a conchiform
squinch in the western towers of Jumieges (1040-1066), though not long after there are
good specimens in the transept towers at Cluny.

In the cathedral of Vaison, the apse with the bay in front, and the apsidal
chapels, are reputed to belong to the Merovingian age. The three bays before the
latter are supposed to be of the early Carolingian period ; and the whole was
probably restored by Bishop Humbert I (996), who will have added the bell-tower
which rises above one of the minor apses. And all this though we know from
Boyer, the historian of the church of Sainte Marie de Vaison, that it was built by the
said bishop, and though the existing structure is, from its vaulting, evidently of the
Xllth century, when the east end of the original church was preserved in an altered
form. As a matter of fact, Choisy x considers the body of the church to be the result
of a rebuilding following the destruction of the city in 1160.

Having made these observations we will now return to our subject, and pass in
review some of the very rare buildings still surviving in Burgundy and the neigh-
bouring districts, which really belong to the Merovingian (481-752) and Carolingian
(752-987) ages and the years up to the epoch of 1000, and retain sufficient constructive
and decorative elements to form the basis for profitable comparisons and reasonable
dating of other edifices which have been wrongly classified.

THE CHAPEL (NOW CRYPT)
OF SAINT LAURENT AT GREN-
OBLE. Below the presbytery of
the present church of Saint Laurent
(considered to belong to the Xlth
century) is situated a small church
which forms its crypt. It is a
three-lobed structure, with the ad-
dition of a fourth arm in the end
or western wall (Fig. 399). Over
the central part is a barrel vault.
The half-domes of the subordinate
apses in the transverse portion are
formed by concave sections carried
by small arches. The principal
apse at the eastern end has a similar
vault, and the angles of its face are
decorated with single shafts, above
which are doubled shafts. The large
apse at the west end was probably
added when the three-lobed cham-
ber was turned into a crypt : in
any case, it is the result of an
alteration in the structure. In the

*'g- 399- Grenoble. Crypt of Saint Laurent (Vlth Century). t^i, r ., , ,

body ol the chapel we notice the
stylobate on which stand the shafts

with capitals carrying high Ravennate pulvins (Fig. 400) supporting a plain architrave
arvmg of these capitals recalls that on two of the Visigothic period, with leaves,

1 Histoire de t architecture.

BURGUNDY

water-lilies, stars, and crosses, at
the principal entrance of the
Mosque at Cordova, founded by
Abderrahman in 785, to which
they were brought from the
church dedicated to the deacon-
martyr, St. Vincent (304), built
after King Reccared's conver-
sion to Catholicism (586-601).

The date of the chapel of
Saint Laurent, which I regard
as the oldest church in France,
is unknown. The general view
is that it may be ascribed to
the Vllth century. I believe,
on the other hand, that the
proper date is the second half
of the Vlth century, and per-
haps the period when the see
of Gratianopolis was held by
Bishop Isicius (573-601). A
date near to the Vlth century
has already been suggested. 1

The Ravennate figure pul-
vins must follow close on the

V\g. 401. Rome. So-called " Tempio di Sicpe" (Ilnd Century).
(from Giovannoli, " Vedute degli anluhi vestigj di Kama")

Fie- 400. Grenoble. Crypt of Saint Laurent. Capitals
(Vlth Century).

archetypes of the kind for western Europe, to be
seen in San Vitale at Ra-
venna (526-547). At the
same time they must be
earlier than the moulded
specimens in the Merovin-
gian crypt of Jouarre (653).
The capitals, too, cannot be
assigned to the time of King
Pippin (752-768) or later, be-
cause, as we shall see when
we come to the crypt of the
church of Flavigny (755-
768), a different type of capi-
tal was then in vogue in
France.

This chapel presents two
notable characteristics. The
first is that of the three
segmented half domes, the
earliest specimen of the form
that I have met with north

1 Bulletin arcAtfoiagj'fue, 1893
Reymond et Giraud, La chapelle
Saint- Laurent A Grenoble.

4 o LOMBARDTC ARCHITECTURE

of the Alps. The idea of breaking up the intrados of a cupola is said to be of
Eastern origin, though really it was a Latin invention, of the time of Hadrian

Fig. 402. Tivoli. Villa of Hadrian. Vestibule of the Palace (125-135).

(117-138). Thus, the intrados of the dome of the circular building known as the
" Tempio di Siepe" at Rome was composed of a succession of concave sections (Fig.
401). Another interesting feature of this dome was the round openings with which
it was pierced the earliest example of such treatment that I know.

Fig. 403. Baiae. Group of Thermal Buildings (Ilnd Century).

Another example of a segmented cupola is the hemispherical vault of the
Serapeum in Hadrian's Villa at Tivoli, from which the architect of SS. Sergius and

BURGUNDY

Bacchus at Constantinople derived his inspira-
tion. We find it again in the dome of the
octagonal vestibule (Fig. 402) in the so-called
" Piazza d'Oro " of the Villa. Again, a group
of circular thermal buildings at Baiae (Ilnd
century), preserved in Sangallo's Vatican
sketch-book, had segmented domes (Fig. 403).
The form appears again in the half-dome at
the end of a hall close to the circular mauso-
leum in the Villa of the Gordians (Illrd cen-
tury) on the Via Praenestina near Rome. And
in one of Rossini's views 1 may be noticed a
round building called by him " Tempio di
Venere Sallustiana," with a similar cupola.

In the Byzantine world the earliest ex-
ample of a dome with continuous concave
segments (i.e. not alternating with flat bands
as in SS. Sergius and Bacchus) with which I
am acquainted, is the one in the convent church
of Myrelaion, also at Constantinople (919-945).

The second characteristic is that of the

Fig. 405. Klavigny. Crypt of the old abbey church (755-768).

Fig. 404. Part of elevation of wall in a
Roman Bath, (from a sketch by G. B.
Saiigallo in the Uffizi.)

two tiers of shafts which decorate
the apses. This motive was sug-
gested by the two tiers of
columns sometimes employed by
the Roman architects in the de-
corative treatment of walls in
their Thermae (Fig. 404). We
have not found it applied to
apses before its appearance in
this chapel at Grenoble.

CKYPT OF THE ABBEY
CHURCH OF FLAVIGNY. The
abbey of Flavigny in the Duchy
of Burgundy, which in 1626, as
Robert says, 2 lay "... ruinis et
ruderibus pene sepultum," was,
according to the same authority,
founded about the year 606.
Ansart, 3 on the other hand, states
that the date of its foundation is
not known. In the next place,
Courtpe 4 cautiously remarks
that the foundation with its dedi-

1 / selte colli di Roma antica e moderna.
Gallia Christiana.

3 De Sainte Keine cfAlise et de tabbaye
tie Flavigny.

4 Description gfnfrale et partictiltire au
duchif de Bourgogne.

LOMBARDIC ARCHITECTURE

cation to St. Praejectus, bishop of Clermont, cannot have taken place till after his
martyrdom at Volvie, which happened in 674* Mabillon 2 puts it exactly in the year
722 ; while Plancher 3 fixes it in the time of Theuderich IV (720-737).

Whatever may be the truth, the translation of the relics of St. Praejectus from
Volvie to Flavigny, and the erection of the church under his patronage, did not take place
till the time of Abbot Manasses (755-788) and King Pippin (753-768)." 5 According to
Hugh of Flavigny it was in 880 that the consecration of the church was performed by
Pope John VIII (872-882). This function must be connected with the translation

to Flavigny in 864 of the body
of Sainte Reine d'Alise, and
with the alterations or rebuild-
ing carried out in the church.

It does not appear that the
building was destroyed by the
Normans in their raid of 887
mentioned by the annalist
Hugh. In the course of the
XHIth century it was rebuilt
in the Pointed style, though the
old sanctuary and the parts
connected with it were pre-
served. In the XVI Ith and
XVIIIth centuries far reaching
restorations and alterations were
carried out. In the last century
it was deserted, and fell into
ruin. The materials were car-
ried away, and the only parts
preserved were the crypt with
a corridor flanking it, a portion
of the sanctuary, and a few
arches of the nave. 7

The only part of the origi-
nal church standing is the crypt
(Fig. 405), popularly known as
the prison of Sainte Reine. The
passage flanking it on the south,
and, above ground, the remains of the choir with open and blank arcades, one above
the other (Fig. 406), which were also preserved in the rebuilding of the XI I Ith cen-
tury, are work of a later date, as is shown by the masonry and the carving. This
date may very well be the second half of the Xlth century, as is suggested; 8 and
then the new works will have followed on the reform effected by King Robert
(996-1031) in 1025 or 1026.' Belonging to the isolated columns in the crypt there

1 Mabillon, Acta Sanct. Ord. S. fienedicti. Vita s. Praejecti episcopi Arvernensis et martyris.

2 Aiittales Ont. S. Benedicti. 3 Op. dt.

4 Migne, Pair. Lat., Vol. 154. Hugonis abbatis Flaviniacensis chronicon,

5 Robert, Gallia Christiana.

6 Migne, Pair. Lat., Vol. 154. Hugonis abbatis Flaviniacensis chronicon.

7 Bordet et Galimard, Kestes de I'ancienne basilique de Fabbaye binidictine de Flavigny.

8 Bordet et Galimard, op. cit. * Delisle, Roberti regis diploma/a.

Fig. 406. Flavigny. Abbey church. Remains of the choir
(Xlth Century).

BURGUNDY

still survive three of the original Pre-Lombarclic cubical capitals, of even ruder
character than those in the crypt of Saint Laurent at Grenoble.

Some would have it that the crypt was rebuilt or restored in the Xlth and Xllth
centuries. This view is quite untenable. The vaulting and the outer wall are mani-
festly of the same date, and the capitals, with their abaci, have been made on purpose
to fit the vaulting ; so that piers, vaulting capitals, and abaci form a single archi-
tectural whole of one date. Moreover, the remarkable rudeness of the vault construction
would ill agree with the quality of French masonry in the IXth and Xth centuries,
not to say that of the Xlth and Xllth. The capitals, again, whether on account
of their form or the rudeness of the execution, cannot by any means be ascribed to
the time of the successors of Charles the Great, and still less to a later period. Lastly,
the occurrence on two of the capitals of
the letter M, believed to be the initial of
Manasses the Great (755-788), the builder
of the crypt, 1 is another piece of evidence
which confirms the date of the building.

The crypt of Flavigny, or so much of it
as is left, is that erected on the occasion of
the translation of the relics of St. Praejectus
(755), and was the necessary consequence of
that event ; for crypts were constructed with
the special object of containing the bodies
of saints.

It is the oldest dated building in France
exhibiting Pre-Lombardic cubical capitals.
These are earlier than the examples in the
church of Germigny des Pres (801-806).
The importation of the type into these
regions was perhaps one of the results of
Pippin's descents on Italy. The conquest
of the Lombard kingdom by Charles the
Great did the rest, and the form spread all
over his Empire. It is a Comacine creation
of the second half of the Vllth century.
Before that time capitals of this type were

not produced. At the most, and then only very rarely, they were used in a decora-
tive way, like those for instance (to which I am the first to call attention) to be
seen in the representation of a temple on a Phoenician stone carving from Carthage
in the Semitic Room at the British Museum (Fig. 407).

In the next place, the intersecting vaulting of the crypt, with the vaulting arches
incorporated in it, shows that the device of visible arches, which had long been
practised by the builders of Rome, Ravenna, and Constantinople, was not yet
followed in France.

THE CHAPEL OF SAINTE BLANDINE IN SAINT MARTIN D'AINAY AT LYONS.
With regard to the monastery of Ainay at Lyons we are told by Mabillon 2 that Queen
Brunhildis (566-613) restored an earlier one known by the name of Interamnis,
founded in the Vth century by Romanus abbot of Condat (f 460). Choppin, 3 how-

1 Bordet et Galimard, op. at. 3 Annalts Ord. S. Bentdicti. * Monasticon.

Fig. 407. British Museum. Phoenician carving.

LOMBARDIC ARCHITECTURE

ever, says that it was erected by Brunhildis in 612, and afterwards destroyed by the
Hungarians ; while Robert l only states that the queen was the foundress of the
monastery at Ainay dedicated to St. Martin. Additional information is given by
Mabillon, 2 3 who relates that it was restored by the abbot Aurelian, afterwards
archbishop of Lyons (875-876-895). And Robert 4 states that Abbot Amblard, who
also became archbishop (957-978), rebuilt it in consequence of the damage done by

the Hungarians in 937. 5

It is to this last renewal,
which took place about 966, that
the existing chapel of Sainte
Blandine belongs. 6 Recent re-
storations have revealed the
existence of work of two dates.
In the oldest, that at the end
of the building, the mortar con-
tains pounded pottery, while in
other parts it is made with sand
from the bed of the Rhone. The
former will belong to the work
of Aurelian, while the large
blocks of stone in the front ot
the existing church of Saint
Martin may come from the
building of Brunhildis.

The existing church is the
result of rebuilding by Abbot
Gaucerannus in IIO2, 7 conse-
crated by Pope Paschal II in
i IO7. 8 Later it was altered by
the addition of outer aisles and
a dome ; by the substitution of
sham barrel vaulting for the
original wooden ceiling ; and by
the conversion of the three
round-headed doorways of the
west front into entrances with
pointed arches.

The chapel of Sainte Blandine is of rectangular plan, with a sanctuary of similar
form (Fig. 408). Two columns are inserted in the angles of the frontal arch of this
sanctuary, while its side walls are decorated with blank arcading springing from
columns. All these columns have Pre-Lombardic cubical capitals, which tell us that
this type, introduced into France in the time of Pippin, was still the fashion in the
southern parts of the country in the Xth century.

1 Gallia Christiana.

2 Acta Sanct. Ord. S. Benedifti. S. Aurehani episcopi Lugditnensis elogium historicum.

3 Annales Ord. S, Benedicti. Gallia Christiana.

6 Annales Ord. S. Benedicti. * Martin, Histoire des tglises et chapelles de Lyon.

1 Sammarthanus, &c., Gallia Christiana Ecdesia Lugdunensis Athanaeum.
s Jaffe, Regesta pontificum romanorum.

Fig. 408. Lyons

Saint Martin d'Ainay. Chapel of Sainte
Blandine (about 966).

CHAPTER II

THE LOMBARDO-NORMAN STYLE IN NORMANDY

foundations of the new church of Saint B^nigne at Dijon had scarcely
been laid when Richard II, Duke of Normandy (996-1026), invited William
of Volpiano to come to Fecamp and reform the abbey of the Trinity. The
latter, remembering the negative results of a previous attempt by Majolus
the abbot of Cluny (948-994), at first tried to avoid the duty, on the ground
of the barbarous and savage character of the Norman dukes, who were more
inclined to destroy than to build churches, and more likely to drive monks away
than gather them together in monasteries. 1 But at last, in consequence of a fresh
and urgent entreaty, he gave way, and started on his mission.

At the dawn of the Xlth century Normandy was no longer in the miserable
conditions which had prevailed during the period between the settlement, in the days
of Louis the Pious (814-840), of the sea-kings' hordes on the island of Noirmoutier,
the centre from which they started, plundering, burning, slaughtering, in every
direction, and the treaty of Saint Clair sur Epte. By that treaty Charles the Simple
(893-929) assured to Rollo (911-931) the possession of Rouen and the lower valley
of the Seine, from the Epte to the sea. And though we have no documentary
information about the early days of the Norman dukedom, founded in 911 or
perhaps not till 921, still, we know that, after the Normans had obtained a foothold
and given the country their name, an epoch of depredation and ruin was succeeded
by one of security and internal development.

It is also certain that, with the accession of Richard II, and after the peasant
revolt had been crushed by Rudolf, Count of Evreux (997), in the manner described
by William of Jumieges, 2 the country had become so strong internally that it felt
itself capable of engaging in a series of expeditions against other territories, which
culminated in the conquest of England. Moreover, what is known about Richard
is far from confirming the charges brought by William against the Norman princes ;
so that we must suppose, either that these were instigated by the desire of putting
Richard's intentions to the proof, and obtaining his support in all things necessary,
or else were based on the low opinion he had of the religious sincerity of the
dukes. The latter point was not devoid of some element of truth, at any rate in the
case of the improvised Christianity of Rollo, or the skin-deep faith of Richard I " the
Fearless " (943-996).

There is, indeed, some discrepancy in the accounts given by the chronicles of the

1 Mabillon, Ada Sand. Ord. S. Benedicti. Vita s. Gitillelmi abbatis. Excerpla ex libra de revelatione,
aedificatione, el atictorilale monasterii Fiscamnensis.

- Duchesne, Historiac Normannontm scriptores antiqui Hisloria Normannorum.

46 LOMBARDIC ARCHITECTURE

conduct of Rollo the Rover after his baptism (912) by Franco, archbishop of Rouen;
so that, while we read in Ademar l that, on the one hand, he sacrificed his Christian
prisoners to the Scandinavian idols, and, on the other, made donations to churches
(" Christianos captvos centum ante se decollari fecit in honore quae coluerat idolorum,
et demum centum auri libras per ecclesias distribuit Christianorum in honore veri
Dei "), Dudo of Saint Quentin, 2 on the contrary, tells us that he was a good and
pious ruler, and a protector of the Church. All the same, there is good reason for
thinking that at the bottom of his heart he remained the pagan he was before
baptism (" idolisque contemptis, quibus ante deservierat "). 3 And with regard to
sacred buildings, if Rollo gave largely to the churches of Rouen, Evreux, Bayeux,
Jumieges, Mont Saint Michel, and Saint Denis, at his baptism, as we are told by
Dudo and William of Jumieges, he does not seem to have undertaken any general
restoration of the churches destroyed during the wars, as these chroniclers would
have us believe (" templa frequentia paganorum destructa restauravit "), seeing that
one of them, Jumieges, was, with the adjoining monastery, still a heap of ruins in the
days of William I " Longsword " (931-943). Again, in the case of Richard I
(943-996), though the " Benedictine Annals " and the two historians just referred to
mention sundry religious buildings as restored or founded by him, on the other hand
we know that, on the occasion of the invasion of Normandy by King Sihtric (943),
when a large number of Normans headed by Thurmod returned to paganism, he
followed their example. This sudden relapse, which may find an excuse in the
duke's youth, is confirmed by the clear testimony of Flodoard. 4

However this may be, William of Volpiano, immediately on his arrival at
Fecamp and institution as abbot, took in hand the reformation of that and other
monasteries (e.g. Jumieges, Mont Saint Michel, and Saint Ouen at Rouen), founded
new ones, restored or rebuilt the old ones, and was in every case made their superior.
The virtues and enterprise of the new abbot of Fecamp and his assistants speedily
fanned into flame a real religious revival which made its influence felt in every
direction. This revival, the result of a movement which is natural to peoples in an
elementary stage of culture, was accompanied by an intellectual efflorescence
produced by the establishment, in the abbeys subject to the supreme control and
authority of William, of schools open to every class of society. The Benedictines
were well aware that letters, aided by the arts, are one of the most effective weapons
for fighting barbarism. In this way the abbeys became seats of public instruction ;
and this was especially the case at Fecamp, where William took a personal share in
the work of education.

Having given this brief sketch of the conditions under which the work of William
of Volpiano and his successors was begun, let us see what was the state of the arts of
architecture and carving in Normandy in the century preceding the epoch of IOOO.

Of the very rare ecclesiastical buildings erected by order of the Norman dukes of
that period, the only one of which there are sufficient remains to provide material for
study and observation is the old church of Saint Pierre at Jumieges (940).

The oldest portions of the church at Fecamp, viz. the chapels of St. Peter and
St. Nicholas, and the round arches springing from continuous Lombardic capitals
ornamented with scroll work and undercut foliage in the ambulatory of the existing

1 A/on. Germ. hist.AdemarusHistonarumlibrilVasaec. K-ioaS.

2 Migne, Pair. Lot., Vol. 141 De moribus et actis primontm Normanniae ducttm.

3 Le Prevost, Orderici Vitalis historic*, ecclesiaslica.

4 Man. Germ. hist. Floiloardus, Annales.

NORMANDY

church in the Pointed style, have nothing to do

with the church of the Trinity founded by Richard I

in 990. Of that structure, with its nave and aisles,

Dudo of Saint Quentin has left a brief but valuable

description, in which he mentions the master

builder who acted as its architect (" petrarum fabro

architectoria arte perito"), and also the material of

which it was constructed. The fragments just

referred to must be assigned to the rebuilding

carried out by William de Ros, third abbot of

Fecamp (1087-1107), and described by Ordericus

Vitalis. " Nam cancellum veteris Ecclesiae, quam

Richardus Dux construxerat, deiecit, et eximiae

pulchritudinis opere in melius renovavit, atque in

longitudine ac latitudine decenter augmentavit.

Navem quoque Basilicae, ubi Oratorium sancti

Frodmundi habetur, eleganter auxit." 1

No trace remains of the rebuilding of the church

of the Mother of God at Rouen, carried out in the

days of Rollo, Duke Richard I, and Bishop Robert I

(989-1037). Nor is anything preserved above ground

of the reconstruction in the time of Archbishop

Maurilius (1055-1067), who consecrated the new

work in 1063. It is to a later date that we must

assign the remains of piers and shafts under the

pavement to the left of the presbytery of the present

cathedral, which was begun after the fire of 1200.

The same may be said of the church of Saint Ouen, also at Rouen, founded

under the invocation of SS. Peter and Paul by Archbishop Flavianus (533-542) in the

reign of Clotaire I, the name
being changed to Saint
Ouen in the Xlth century. 2
It was rebuilt, if not built,
by the said duke according
to Ordericus Vitalis and
William of Jumieges. For
the most ancient part of
the present church, viz. the
apse in two stories marked
off on the outside by a billet
course, each of which con-
tains a window with angle
shafts and Corinthianesque
capitals, is not older than
the rebuilding carried out
between 1046 and 1126
(Fig. 409).

1 Op. cit.
Fig. 410. Jumieges. Church of Saint Pierre (about 940). 2 Robert, Gallia Christiana.

Fig. 409. Rouen. Apse of the old
Saint Ouen (Xlth and Xllth Cen-
turies).

LOMBARDIC ARCHITECTURE

Fig. 411. Jumieges. Capital in church of
Saint Pierre (about 940).

CHURCH OF SAINT PIEKRE AT JUMIFGES. It has been thought by some 1
that the relics of this church which survive belong to the first foundation (654) by

St. Philibert with the aid of Chlodovech II
(638-656) and his queen Bathildis. Really what
we see is a rebuilding of the original structure
(burned with the adjoining convent by Hasting
in 851) carried out by Duke William I, it seems,
in 94A 2 while Gonthardus was archbishop of
Rouen (919-942). The greater part of it was
destroyed to make way for the new construc-
tion of the XlVth century; and what was
left was damaged in the XVIth, and finally
reduced in the early years of the last century
to the deplorable state in which it now
appears.

In the small surviving fragment of the
original building, which had a nave and aisles
(Fig. 410), two features are to be noticed. One
is the triforium gallery, with groups of arches

enclosed within a relieving arch after the Romano-Ravennate manner, thus confirming
our idea as to the date of the building. It was an arrangement which did not become
the fashion north of the Alps before the time of Charles the Great (768-814). The
other is the presence of Pre-Lombardic cubical capitals (Fig. 411), which are another
confirmation of that date. Very different were the capitals in vogue in the north of
France in the Vllth century, and precisely in the time of Chlodovech II and Queen
Bathildis, as the crypt of Jouarre
(653) tells us.

The remains of Saint Pierre in-
form us how low was the standard of
building in Normandy in the Xth
century. Nor was this confined to
the Duchy, but must also have pre-
vailed in the adjoining districts, to
judge by the remains of the old
cathedral of Beauvais.

THE OLD CATHEDRAL OF
BEAUVAIS is popularly known as
the " Basse CEuvre." There are those
who would take it back to Merovin-
gian times (481-752), while others
regard it as a work of the VI 1 1 th or
IXth century. A third view places
it in the time of Bishop Herve (987-
997). It is this last date which

1 Loth, Histoire de fabbaye royale de Saint-
Pierre de Jumieges.

_Le PrevosL-OrAnV/ Vitali, historia Fig. 4 , 2 .-Angers. Remains of church of Saint Martin

ecclesiastica. (1030).

NORMANDY

Fig. 413. Beauvais. Old Cathedral (987-997).

suits it best, as Robert 1 shows; and

it explains Viollet-le-Duc's 2 state-
ment that the church was in existence

in 990.

Of the original structure, mutilated

when the new cathedral was built in

the Pointed style(XIIIth century), there

is preserved the front, and the first three

bays of the nave and aisles. They have

wooden roofs, and are separated by piers

of octagonal or quadrangular section

without capitals or impost mouldings.

The octagonal ones are hollowed out on

four sides in order to fit the springers

of the arches. These piers are the

precursors of the more elaborate

ones still surviving in the ruins of

Saint Martin at Angers (Fig. 412),

which are fitted to their imposts by

being hollowed out at the angles, each

hollow being occupied by a coarse

leaf. Saint Martin was rebuilt, with

the assistance of Count Fulco III Nerra

(1012-1040) and his wife Hildegarde,

before the year IO2O, 3 and dedicated by Hubert, bishop of Angers (1010-1047), in

I030. 4

In the old cathedral of Beauvais the window archivolts, with alternate voussoirs

of stone and brick enclosed by a ring of bricks laid horizontally (Fig. 413),

recall those at
Agliate (824-860)
and in the palace
of Constantine VII
Porphyrogenitus
(912-958) at Con-
stantinople (Fig.
414). The large
window in the front,
richly decorated
with stars in low
relief, is the result
of an alteration (Fig.
415). Ornament of

1 Cal/ia Christiana.

2 Diclionnaire rat-
so/ins <k rarchitcclute
frattfaisc Ju X' au XVI'
siicle.

3 D'Espinay, Notices
archi'ologiqtus.

Kig. 414. Constantinople. Palace of Constantine Porphyrogenitus (912-958). ' Gams, op. cit.

VOL. II E

LOMBARDIC ARCHITECTURE

this kind was revived by the Lom-
bardo-Roman style, which borrowed it
from buildings of the Roman period.

* * *

F jg. 4I5 ._Beauvais. Old Cathedral. (987-99?).

We will now supplement our ac-
count of the two buildings just dis-
cussed by that of five others of an
earlier date : the crypt of Jouarre
(653), the baptistery of Poitiers (VII th
century), the church of Germigny des
Pres (801-806), and the crypts of Saint
Aignan (814-840) and Saint Avit
(IXth century) at Orleans. We are
thus provided with a group of build-
ings of the Merovingian (481-752) and
Carolingian (752987) a S es which '
with the addition of the three of the
same periods described in the first
chapter of the Second Part, and of
Angilbert's basilica at Saint Riquier
(Centula) (793-79$) as preserved in a
view in Mabillon, 1 may furnish us with
typical examples which will enable
us to form an idea, incomplete per-

haps, but certain so far as it goes, of the state of ecclesiastical architecture in France
at those periods and up to the epoch of 1000.

THF CRYPT OF THE CHURCH OF SAINT PAUL AT JOUARRE was built (653)
by St Ado, placed under the rule of St. Columban, and endowed by Queen Bath.ldis
(649-680) 23 It exhibits work of three distinct dates. The first is that of
original foundation, and to this belongs the northern part of the structure with its
Corinthianesque and Composite capitals (Fig. 416). To the second is to be ascribed
the enlargement of the primitive crypt, represented by a chapel built against it,
dedicated to St. Ebrigisilus, bishop of Meaux (Vllth or VHIth century),
enlargement is to be connected with the translation of the relics of St. Potentianus
to Jouarre (847)." As a matter of fact, a pulvin carved with foliage, recalling
specimens in the abbey church of Mettlach (987-1000), and others in the small
chapel under the south aisle of the Castle church of Quedlinburg (997-1021),
may be assigned, on account of its form and the quality of the carving, to the
Carolingian age, and in it to a period later than the reign of Charles the Great.
To the third date belongs a final enlargement to the south-west, forming a pendant
to the chapel referred to (Fig. 417). This event may be connected 5 with the
installation in the Xlth century of a chapter of secular canons which is mentioned

1 Ada Sanct. Ord. S. Benedict! Vita s. Angilberli abb. Cenlulensis, auctore Hariulfo.

2 Man. Germ. hist. Vita sanctae Balthildis.

3 Delisle, Aimoni monachi Floriacemis, De geslis regum Francontm.

4 Mabillon, Ada Sanct. Ord. S. BenedidiDe translation sancti Potentiani in coenobium Jot reuse.

5 Enlart, Manuel d 1 archeologie fran^aise defnis les temps merovingiens jusqrf a la Renaissance.

NORMANDY 5I

by Mabillon. 1 This date suits the Lombardic capitals, some of them having figures,
a result of the Lombardo-Norman influence in art which made itself felt after the
epoch of 1000. Earlier than this, Lombardic figure capitals are not to be found
in France. From the Merovingian age onwards the only dated French figure capital
that I can point to is the one to which I have already called attention in the crypt
of Saint Pierre de la Couture at Le Mans (997).

On the one hand, the carved heads on the capitals at Jouarre are, to judge
by the one which is well preserved, poor work, and earlier than the Xllth century ;
but, on the other hand, a date not before the Xlth century is disclosed by the Attic

Fig. 416. Jouarre. Original Crypt in the church of Saint Paul (653).

spurred bases in this part of the crypt, for this Lombardic motive did not cross
the Alps till about the middle of that century.

This third phase was accompanied by a general alteration of the building,
and by the construction of the unraised cross vaulting. For the vaults belonging
to the Xlth century building are evidently made to fit the supports, and those of the
Vllth century in the crypt are of just the same character. Moreover, the cross
vaulting in the crypt of Saint Pierre at Flavigny tells us that in France, during the
Dark Ages, cross vaulting was of the continuous type, and not supported by visible
arches, as at Jouarre. I should mention here that Rohault de Fleury 2 thinks that
the first church had only a flat roof, and that the vaulting was added in the Xth
or Xlth century.

The most important things in the crypt are the Merovingian capitals. They dis-
play an artistic quality which one would look for in vain in Italian Lombardic capitals
of the Vllth century, and can only be compared (making allowance for the difference
of type) with the contemporary productions of the School of Ravenna. Another
noteworthy object is the shrine of St. Theodcchildis, the original parts of which have

1 Annales Ord. S. Bcnedicli.

La Afesse, Mudes archtologiqius sur ses monuments.

E 2

LOMBARDIC ARCHITECTURE

carvings of scroll work with
grapes, roses, lilies, and scallop
shells, the whole finely exe-
cuted.

This revival of art in
France in the Vllth century is
not difficult to understand when
we remember the assistance it
received in the form of the
foundation and endowment of
ecclesiastical institutions by a
whole series of royal person-
ages, from the pious Radegund
(538-587) to the great but un-
fortunate Brunhildis (566-61 3).
It was the latter who pro-
tected the mission sent by
Pope Gregory I to convert
England, and she was so im-
portant a patroness of archi-
tecture that a large number of
buildings with which she had
no connection were ascribed to
her. 1 And the series goes on
from her to Sigibert III (638-
650), and the virtuous and
capable Bathildis.

Fig. 417 Jouarre. Additions to original Crypt of Saint Paul
(IXth and Xlth Centuries).

THE BAPTISTERY OF
SAINT JEAN AT POITIERS
was erected over a Gallo-Roman sepulchral edifice. It consists of two parts, the
baptistery proper, and the narthex. The plan is an oblong, with apses projecting
from the three free sides (Fig. 418). The roof is of wood, except in the case of
the two subordinate apses to the north and south. The principle apse, and the north,
south, and east walls of the body of the building, are decorated internally with
arcading. On the outside it is ornamented with small pilasters, round and triangular
pediments, &c. (Fig. 419).

The front and the narthex were probably rebuilt after the conflagration which
devastated Poitiers in ioi8. 2 Traces of fire may still be seen in the building.

With regard to this most interesting of the earliest French churches we are
entirely without authentic documentary evidence for fixing the date of its construction.
The most likely date appears to be the time of Bishop Ansoald (6S2-696), 3 one of the
most important holders of the see. Structural works are generally to be connected
with characters capable of large undertakings, and Ansoald was one of the most
notable bishops of Poitiers.

The presence of capitals brought from elsewhere, which, though more bevelled,
recall by their nearly equal rudeness those in the chapel of Saint Laurent at Grenoble,

1 Man. Germ. hist. Annales Wormatienses. 2 Man. Germ. hist. Ex chronico Ademari Cabanensis.

3 De la Croix, tiide sommaire du baptistere Saint-Jean de Poitiers.

Fig. 418. Poitiers. Baptistery of Saint Jean (Vllth, Xlth, and Xlllth Centuries.)

Kig. 419. 1'oitiers. Baptistery of Saint Jean (Vllth, Xlth, and Xlllth Centuries).

LOMBARDIC ARCHITECTURE

and may well date from the long and prosperous reign of Theodoric the Visigoth
(449-451), fixes the erection of the baptistery in Merovingian times. Now, in that
period, the most favourable moment for its construction was certainly the time of
Radegund ; yet there is no mention of the event, which, considering the purpose of
the building, would be an important one, either by Venantius Fortunatus l who was
intimate with her, and was afterwards bishop of Poitiers (599-600), or by the nun
Baudonivia, 2 or by their contemporary, Gregory of Tours (573-595). So that we are
obliged to ascribe it to a later date, viz. the Vllth century, in which the period that
suits it best is the episcopate of Ansoald.

The most notable features of the baptistery of Poitiers are its architectural decora-
tion, both internal and external ; the triple frontal arch of the principal apse ; the
blank arcading in the apse, which leads the way for that in the church of Germigny
des Pres (801-806) and the chapel of Sainte Blandine at Lyons (966) ; and, lastly, the
triangular-headed arches and pediments, like the apertures in a dove-cot.

THE CRYPT OF THE CHURCH OF SAINT AIGNAN AT ORLEANS. The crypt
beneath the choir of the present church of Saint Aignan exhibits in its oldest

parts the remains of a structure
which may be referred to the time
of the Emperor Louis the Pious
(814-840), and was the result of
a visit paid by him to the place, 3
and also to the rebuilding of the
church by King Robert, the dedi-

^?J^^K5- * \ ca tion taking place in 1029* (Fig.

420).

To the IXth century belong
the wall-arches on the north and
south, the cubical capitals of which,

f^|3 as being artistically inferior to the

lr~- ' . . ... _

original ones of the same form in

- M ESP II ^H The dulfs church at Germigny

des Pr^s, must be dated later
than the first years of that cen-
tury.

To the beginning of the Xlth
century belong the two capitals
(one with foliage, the other with
figures) on the half-columns in the
west wall. They are rudely exe-
cuted, especially the one with
figures.

Fig. 420. Orleans.

Saint Aignan.
Centuries).

Crypt (IXth and Xlth

THE CHURCH OF GER-
MIGNY DES PRS was erected
by Theodulf, abbot of Fleury and bishop of Orleans (788-821), between 801 and

1 Man. Germ. hist.De vita sanctae Kadegiotdis.

2 Man. Germ, hist. Vita Radegundis reginae Francorum.

Delisle, Ennoldus Nigellus Carmen elegiacum de rebus gestls Ludovici Pii.
4 Id., Helgaldus Epitome vitae Koberti rcgis.

NORMANDY

internal adornment of the church (I am speaking of those that are original, for
the modern ones are devoid of the characteristic stamp of the cubical capitals of
the time of Charles the Great), and also by the narrow double-splayed windows
in the eastern apse, and the still narrower slits in the other parts of the church,
except the dome where the windows are fairly large. Both of these features are
characteristic of the contemporary School of Lombardy.

Fig. 423. Germigny des Prs. Church (801-806).

More evidence is to be found in the capitals of the shafts belonging to the sanctuary
arch, the arcading round the apse, the triple window openings in the central space,
&c., all displaying the Pre-Lombartlic manner of the VHIth and IXth centuries.
Among these capitals, the old ones still preserved, and the casts taken of the others
before they were re-worked, suggest a Lombard hand. Unless, indeed, some
Fre-nch carver had learned to handle his chisel in Comacine fashion, after the
introduction of the Pre-Lombardic cubical capital in the crypt of Flavigny (755-768)

56 LOMBARDIC ARCHITECTURE

origin, but an Italian by birth (" erat Theodulfus natione Italus " l ), procured from
Italy the craftsmen qualified to erect and decorate the church which was his pleasure

Fig. 422. Uermigny des Pres. Church (801-806).

, ' " - J*o i i i i J i .-i . A1113

capitals, clumsy adaptations of Corinthian, used in the

1 Mabillon, Annales OrJ. S. Benedicti.

NORMANDY

Fig. 423. Germigny des Pres. Church (801-806).

More evidence is to be found in the capitals of the shafts belonging to the sanctuary
arch, the arcading round the apse, the triple window openings in the central space,
&c., all displaying the Pre-Lombardic manner of the VHIth and IXth centuries.
Among these capitals, the old ones still preserved, and the casts taken of the others
before they were re-worked, suggest a Lombard hand. Unless, indeed, some
French carver had learned to handle his chisel in Comacine fashion, after the
introduction of the Pre-Lombardic cubical capital in the crypt of Flavigny (755-768)

LOMBARDIC ARCHITECTURE

where, as some think, it was due to imitation of foreign models, or actually to the
work of Comacine masters. 1

Again, there are the remains of the mosaics in the half-dome of the principal
apse, which are believed to be original. 2 We must remember that when Theodulfs
church was erected, mosaic work still maintained a position of some importance in
Italy, as is shown by the precious examples in the Roman basilicas of Santa Cecilia,
Santa Maria in Domnica, and Santa Prassede, all put up by Pop e Paschal I

(817-824), and as would be illustrated,
had they survived, by the mosaics, prob-
ably the work of artists from Ravenna,
with which not a few sumptuous struc-
tures of the Lombard age were originally
decorated.

Then there is the arcading which
decks the interior of the sanctuary apse
(Fig. 423), an arrangement perhaps de-
rived from the pairs of arcades in the
sanctuary of the baptistery at Poitiers
(Vllth century), which in their turn may
have come from the arcading round the
apse of San Giovanni Evangelista at
Ravenna (425).

Lastly, there are the pairs of deco-
rative angle shafts in the sanctuary arch,
a feature taken from the chapel (now
crypt) of Saint Laurent at Grenoble
(VI th century), where it was employed
just about the time when the architect
of the basilica at Kalb-Lauzeh (Vlth
century) was springing the outer archi-
volt of his chancel arch from shafts
supported by brackets.

But though we may infer that
Theodulfs basilica was raised by the
combined efforts of Italian and French
workmen, we may also be allowed to
believe that it was carried out under
the advice of an architect from the
East, perhaps the one who designed the
chapel at Aachen in accordance with the
wishes of Charles the Great. This would explain the presence of domes and barrel
vaults in the church, and also the feature of horse-shoe arches. For though recesses
of this form occur (sporadically) in the ruins of the villa known as the " Sette Bassi "
on the Via Latina near Rome (Ilnd century), the type came from the East. I may
notice that the sides of a sarcophagus in Sant' Apollinare in Classe (not earlier than
the second half of the VHIth century, owing to the motive of branches ending in
leaves joined at the centre of each group by a boss so as to form a sort of series of
wheels : a motive which does not appear in Italy before that time) afford an early

1 Bordet et Galimard, op. tit. * Prevost, op. cit.

Fig. 424. Home. Villa Mattel. Detail from
sarcophagus (Illrd or IVth Century).

NORMANDY

decorative example of horse-shoe arches for Italy and the lands beyond the Alps.
The oldest instance in the West is furnished by the well-known sarcophagus in the
grounds of the Villa Mattel on the Celian at Rome (Illrd or IVth century), the
sides of which have arcades with arches larger than semicircles (Fig. 424).

THE CRYPT OF SAINT AVIT AT ORLEANS. This is a miniature basilica, with
nave and aisles divided by four octagonal piers surmounted by polygonal capitals
chamfered at the angles. The vaulting has been reconstructed (Fig. 425). It is
reached through an ante - crypt
with intersecting vaulting spring-
ing from two cylindrical columns
with capitals in the same style as
those just mentioned.

The history of this structure is
not known. Some think that it
belongs to the days of Childebert I
(511-558), others that it is of the
Carolingian period. With this
latter view I am in agreement,
and the most probable date will
be the reign of Charles the Fat
(881-887), and later than the crypt
of Saint Aignan, inasmuch as the
art displayed in the mouldings in
Saint Avit is more decadent than
that of the capitals of the latter.

Let us now proceed to review
the distinguishing characteristics
of the French ecclesiastical build-
ings of Merovingian or pre-Mero-
vingian and Carolingian times.

I. Up to the time of Charles
the Great (768-814) chapels were
covered with barrel vaulting, while
structures of larger size, such as aisled basilicas, had timbered roofs. Thus, barrel
vaults occur in the chapel of Saint Laurent at Grenoble (Vlth century), and in the
oldest part of the chapel of St. John Baptist in Saint Benigne at Dijon, going back
to an earlier period than the IXth century, and possibly even belonging to the
age of Gregory, bishop of Langres (5O7-539). 1 For though the vaulting of the
chapel at Grenoble has been restored, and that of the one at Dijon rebuilt, its form
was not altered in either case.

On the other hand, to take a few instances, there were wooden ceilings in the
basilicas of St. Martin at Tours as erected by Bishop Perpetuus (460-490), and of
SS. Peter and Paul, also founded by him there. 2 The same was the case with the
church built by St. Namatius (446-462) at Clermont ; 3 and with that of Holy Cross
and St. Vincent (Saint Germain des Pres) at Paris, erected by King Childebert
(511-558), and consecrated in 558 by St. Germanus, which was of cruciform plan and

1 3 Mon. Germ. Hist. Gregorii episcopi Turoiiensis historia Francarum.

Fig. 425. Orleans. Crypt of Saint Avit (IXth Century).

60 LOMBARDIC ARCHITECTURE

had its roof sheeted with gilded copper. 1 Again, the baptistery of Saint Jean at
Poitiers (Vllth century), not excepting the sanctuary, had, apparently, a wooden roof;
for when the scheme of painting was carried out in the Xllth century, the decoration
was confined to the side walls, whereas in the redecoration of the following century
the paintings extended over the vaulting, so that we must infer that the latter was
not in existence in the Xllth century.

In all these wooden-roofed buildings the apse must be excepted. Apart from
one here and there, of more complex form, and presenting greater difficulties in
construction, like the eastern apse of the baptistery at Poitiers, they must have been
covered by half-domes in masonry, though Gregory of Tours never mentions such.

The lantern-tower over the crossing, where there was one, must also have had a
wooden roof. Gregory, in fact, says in his account of the destruction of the tower
over the altar of the basilica of St. Antolianus at Clermont, that the operation began
with the removal of the timbers of the roof : " iussit tegnos asseresque vel tegulas
amoved." 2

The crypts mentioned by him are described as vaulted. The one at Dijon stands
for all. Later, those of basilica form, of the age of Pippin (752-768), had continuous
unraised cross vaulting springing from isolated supports or wall shafts, after the manner
seen in the crypt of Saint Pierre at Flavigny (755-768).

In the reign of Charles the Great, churches of central plan, as his palace chapels
appear to have been generally, were vaulted, as we see at Germigny des Pr6s ; while
those of basilica plan had wooden roofs over nave and aisles. The abbey church of
Saint Riquier, built by Angilbert between 793 and 798 (" fulgentissima ecclesia, omni-
busque illius temporis ecclesiis praestantissima" 3 ), was supported by columns brought
from Rome, and had a wooden roof, which accounts for its destruction in 1131.

The two large round towers of this church, one in front of the apse, the other
between the church and its vestibule, had imbricated wooden roofs with an opening
at the top, over which rose light structures, also of wood, in three stages with corre-
sponding roofs. An identical tower, but of smaller dimensions, stood beside the
adjacent church of St. Mary, also built by Angilbert, as may be seen in the priceless
view of the monastery of Centula, preserved for us by Mabillon 4 (Fig. 426). This type
of very lofty towers with wooden spires, passing from an interior square base into a
circular form on the exterior, is of Prankish origin. Venantius Fortunatus, 6 describing
the new cathedral of Nantes, erected by Bishop Felix (552-582) about the year 570,
mentions a tower-like structure, square below and round above, rising to a point, and
soaring into the air with a series of arcaded stories :

" In medium turritus apex super ardna tendit

Quadratumque levans crista rotundat opus.
A/this, lit stupeas, arce ascendente per arcus
Instar mantis agens aedis acumen habet "

Cross vaulting continued to be without visible transverse arches, as is proved by
that of the ground floor of the Palatine Chapel at Aachen (796-804) not indeed a
French building, but the most famous structure of its time. Such arches do appear in
the upper story, but in connection with barrel vaults and vault cells.

1 Mabillon, Ada Sanct. Ord. S. Benedicts Vita s. Droctovei abbaiis basilicae S. Vincenti in suburbia.
1 Man. Germ. hist. Historia Francornm.

3 D'Achery, SpicilegiumChronicon Hariulfi monachi S. Richarii Ceniulensis.

4 Acta Sanct. Ord. S. Benedicti Vita s. Angilberti abbatis Centulemis.
6 Man. Germ. hist. Venantius Fortiinattis, Carmina.

NORMANDY

After the time of Charles the Great the practice of vaulting only the sanctuary
continued. The first of the two chambers forming the chapel of St. John Baptist in
Saint Benigne at Dijon (IXth century) has an unraised intersecting vault. The chapel,
too, of Sainte Blandine at Lyons (about 966) was given an unbroken barrel vault.
Basilicas also still had wooden roofs. The abbey church of St. Gall (822-829) (outside
France like the rotunda at Aachen), which was a colonnaded basilica, had a roof of
this nature.

It was only after the advent
of the Capets (987) that aisles
of churches were occasionally
vaulted. Thus, for instance, while
the old cathedral of Beauvais
(987-997) had its nave and aisles
roofed with timber, the church
of Saint Front at Perigueux,
founded later than the year 988,
had a wooden roof over the
nave, but ramping barrel vaults
in the aisles.

It is true that it has been
suggested that the abbey church
of Saint Pierre de la Couture at
LeMans had a cross-vaulted am-
bulatory with radiating chapels.
The rebuilding of this church is
chronicled byMabillon underthe
year 997 l at the hands of Abbot
Gauzbert I (990-1007) (appar-
ently in consequence of a dona-
tion by Hugo Count of Le Mans
in 990 2 ), replacing the older
church erected by Bishop Ber-
tram (587-623), and still stand-
ing in 996 when Bishop Segen-
frid of Belleme (971-996) was
buried in it. 3 But the oldest
parts of the church of " La
Couture" must be subjected to
a fresh examination, for the
only fragment of the rebuild-
ing of 997 which survives is the crypt, and that is no longer in its original
condition.

Whether the works begun by Gauzbert I were interrupted, and then resumed
with a fresh architectural design by his successor Ingelbaud (1010) ; or whether, at a
later date, the choir above the crypt was pulled down in order to rebuild it with
an ambulatory and radiating chapels, it is impossible to say. It is certain, however,
that the shafts barely touching the outer walls of the crypt, with their stilted arches,

Fig. 426. Saint Riquier (Centula). View of the Churches
(793-798). (from the " Acta Sanct. Ord. S. Benedicli.")

Annales Ord. S. Benedicli. a Delisle, Ex aclibus pontijicuni Cenomannoisium.

1 Cartulaire des abbayes de Saint- Pierre de la Couture et Saint-Pierre ae Soltsitics.

Kig. 427. Angers. Church of
Notre Dame de la Charite.
Capital (1028).

62 LOMBARDIC ARCHITECTURE

are a later addition. The capitals, too, belonging to similar shafts and to the corre-
sponding half-wall-piers, are different from the others in the crypt, and must have been
set up in the second quarter of the Xlth century, and that will be the date of the
remodelled ambulatory with its ring of chapels : a date, in other words, later than that
of the original capitals in the crypt and in the abbey church of Notre Dame de la
Charite or "du Ronceray " at Angers, founded by Fulco III Nerra (1012-1040) and
his wife Hildegarde, and consecrated in 1028. And we say this while taking into account
the absence of extreme archaic character in the animal figures in the ambulatory of
" La Couture " compared with those in " La Charite." These latter, being original,
exhibit foliage, birds, quadrupeds either in pairs facing one another in a threatening

manner (Fig. 427) or engaged in throttling a dove,
human figures in a circle holding hands, a man into
whose ear the Holy Spirit whispers in the form of a
dove, the Flight into Egypt. The representation of
the living beings reveals a lower stage of art than that
of the contemporary capitals in San Flaviano at Monte-
fiascone.

We may observe here that the crypt of Saint Martin
au Val at Chartres belongs in our opinion to almost the
same date, viz. the episcopate of the celebrated Fulbert
(1007-1029). The church itself was destroyed by the
Normans in 911, and again by Duke Richard I (943-
996) in 965.! As a matter of fact, its Lombardic
capitals belong to the first quarter of the Xlth century,
with the exception of one here and there of the Merovingian age, which has been
utilized : e.g. a Composite capital with handles, recalling another of the same kind
in the crypt of Saint Paul at Jouarre (653).

II. The buildings of the Merovingian age illustrate a whole new Grammar of
Ornament. One fresh motive is that of shafts placed one above the other to decorate
the sanctuary arch, in the manner we noticed in the chapel of Saint Laurent at
Grenoble (Vlth century). Entirely new, again, are the coupled blank arcades in the
sanctuary of the baptistery of Poitiers (Vllth century), which were the model for the
continuous arcades of the principal apse in Theodulf s church at Germigny des Pres
(801-806), and the variant in the apse of the chapel of Sainte Blandine at Lyons
(about 966). Original, too, is the form of the triple arch enclosing the frontal arch of
the principal apse of the baptistery just mentioned. Of quite new design and concep-
tion are the two blank arches with a triangular-headed one between them, suggested
by the colonnades of alternate arches and pediments, or arches, architraves, and
pediments, which sometimes decorate the fronts of sarcophagi belonging to the first
Christian centuries (Fig. 428), and also tombstones. 2 Possibly this architectural and
decorative pediment design of the baptistery at Poitiers may have given suggestions
to the French builders of the epoch of 1000 ; unless, indeed, they got the idea from
some structure of the Gallo-Roman age, for the Romans were acquainted with the
motive of a series of isolated or continuous pediments, as we saw in our account
of San Giovanni Evangelista at Ravenna. In the same way, the design of blank
triangular-headed arcading may have given the idea to the German builders of the
IXth century.

In Italy, where the round arch has always been in favour, and has formed the

1 Clerval, Chartres So. cathidrale Ses monuments. 2 Ramsay, The Cities of St. Paul.

NORMANDY 63

basis of all the chief architectural styles, examples of this motive are very rare in
mediaeval buildings, and have only a secondary importance.

In France the pediment motive may be seen employed on the exterior of the
nave of the church at Saint Gen6roux, of the origin or history of which nothing is
known, though some think that it was built before the Norman invasions, while
others, with Gailhabaud, 1 believe that it probably belongs to the reign of Charles the
Bald (843-877). I should place it, approximately, in the last years of the Xth or
the first of the Xlth century, on the grounds suggested by Choisy - and by Dehio
and Von Bezold. 3 In any case it is certain that it is not earlier than the reign of
Louis III the Saxon (876-882). In the time of Charles the Great (768-814) and
Louis the Pious (814-840) ecclesiastical buildings had no external architectural
ornament, and decoration was confined to the interior. Thus, the exterior of the
rotunda at Aachen is plain, with the exception of the drum of the dome, where the

Fig. 428. Rome. Sarcophagus in the Lateran Museum (IVlh Century).

angle buttresses form part of the construction. And the exteriors of the original
church at Saint Riquier (Centula), and of the existing ones at Germigny des Pr6s and
Steinbach near Michelstadt (815-819), and of the sepulchral rotunda at Fulda
(818-822), were equally unadorned. North of the Alps the first building exhibiting
a scheme of architectural decoration on its outer face was the chapel at Lorsch
(876-882).

Moreover, the presence of dentils, not of the ordinary oblong form, but cylindrical
(i.e. billets), in the moulding which frames the windows and runs below the gables in
the church at Saint Generoux, suggests a date which, though not the I Xth century,
when a decorative member of that form was unknown, is not far from the revival of
art which dawned on France at the opening of the epoch of 1000. These billet
mouldings were largely used in the exterior decoration of churches in the Lombardo-
Norman style.

III. As late as the reign of Pippin (752-768) capitals were either Roman ones
used over again, or else imitations of them so far as could be achieved in that more
or less disturbed period. But under Pippin the Pre-Lombardic cubical capital came
on the scene. This characteristic member, with or without chamfering of the lower

1 L' architecture du V e au XVII' sitclc.

- Histoirt de I 'architecture.

s Ot. cit.

6 4 LOMBARDIC ARCHITECTURE

part, the earliest Transalpine example of which exists in the crypt of Saint Pierre at
Flavigny (755-768), soon became the fashion in France, and in the days of Charles
the Great we find it in full possession at Germany des Prfe. It remained so all
through the Xth century: the old church of Saint Pierre at Jumieges (940), and the
chapel of Sainte Blandine in Saint Martin d'Ainay at Lyons (966), still contain
capitals of the purest Pre-Lombardic type.

The typical Pre-Lombardic cubical capital in France before the epoch of 1000
was embellished with foliage, flowers, interlacing, cauliculi, but not human or animal
figures, whether real or imaginary. For these the French had to wait till they were
imported from Lombardy ; and the first results of this importation we saw in Saint
Benigne at Dijon.

After this preliminary study we will now proceed to examine the few but
important churches built from the designs of William of Volpiano or his pupils, and
with their help to trace the gradual development of the Lombardo-Norman style
which marked the revival of architecture in Normandy.

THE ABBEY CHURCH OF BERNAY was founded and dedicated to the Mother
of God in 1013 by Judith (1008-1017) wife of Duke Richard II (996-1026). The
convent was dependent on the abbey of Fecamp, and placed under the supreme
authority of William of Volpiano. The work was completed by the duke, who gave
the abbey a liberal endowment. 1

With all its mutilations and alterations ; reduced at one time to a barrack, and
now permanently converted into a corn exchange, a fire-engine house, shops, a prison,
&c. ; partly concealed by structures which have grown up against it, this ill-starred
Lombardo-Norman church still forms, in spite of the great difficulty of making an
even incomplete examination of the structure, a very valuable piece of evidence,
dealing as we are with the earliest known specimen of the Lombardo-Norman style
in its infancy.

To begin with, it had the form of a perfect Latin cross (" crux immissa," in
which the transept is not placed at the extremity of the upright limb but cuts it
some way below the top), and was divided into nave and aisles which were continued
east of the crossing. The nave ended in a semicircular apse. Each arm of the
transept has an apse projecting from its eastern side.

All that is left standing of the original structure is the nave arcades of five
arches on either side, and the corresponding southern aisle ; the south arm of the
transept, and the crossing ; and, thirdly, the south aisle of the presbytery. The
northern one has been altered and spoiled, and all that can be seen on the outside is
a fragment of wall and a window. The northern aisle of the nave has been rebuilt
in the style of the XlVth century, probably after the damage which the church is
believed to have suffered in the siege of 1357, when the parish church of Sainte Croix,
adjoining the fort within which the abbey stood, was destroyed. It was rebuilt
in 1374. An old view 2 shows the church reduced to five bays and deprived of its
apse.

The interior has an effect of rugged, severe majesty (Fig. 429). The nave has a
wooden roof, and the piers are compound with two engaged columns. The vaulting

1 Du Monstier, Neuslria Pia Bernayum. - Peigne-Delacourt, Monasticon Gallicamtm.

NORMANDY 65

in the south aisle is work of the XVIIIth century. Above the aisles are galleries
covered by very low roofs, and originally lighted from the nave by two-light openings
enclosed by an arch. Above the arches of the crossing rose the lantern tower, now
destroyed.

The south arm of the transept, which belongs to the original construction, has on
its west side a passage or service gallery with lofty open arches (Fig. 430) ; while the

Fig. 429. Bernay. Nave of Abbey Church (1013).

similar one on the eastern side has low arches. An ace of hearts is carved on the
face of one of the piers belonging to the latter. With regard to this decorative
motive of the ace of hearts, sometimes taking the form of an ace of spades, I may
remark that it has been derived, 1 at least when it occurs later than the Vth century,
from Syria. But this is not the case. These aces of spades and hearts, which are
merely reproductions of the conventionalized outline of ivy or vine leaves with or
without the stalk, had been used by Christian artists in the West from the earliest

1 Courajod, Origines de Fart reman et gothiyae.
VOL. II

LOMBARDIC ARCHITECTURE

Kg. 430. Bernay.

Abbey Church.
(1013).

South arm of transept

times of Christianity, either as a
symbol, or as a full stop, or in
separating or abbreviating words.
A defective tablet with an inscrip-
tion of 269, 1 still exhibits ten aces
of clubs, and originally fourteen or
more. Nor was there any need for
these artists to go to the distant
East (e.g. the temple of Baalsamin
at Siah founded about 23 B.C.) to
find conventionalized ivy or vine
leaves, for there were plenty of ex-
amples in the West. They appear,
to take one instance, on a mosaic
of the Romano-British period dis-
covered at Silchester (Calleva Atre-
batum) (Fig. 43 1). 2 And in Italy,
not to mention the Romans, the
Etruscans had used them in their
tomb decorations from early times
(Fig. 432).

The nave and aisles are con-
tinued for two bays east of the
crossing, the central space having a
wooden roof, while the aisles had
roughly constructed cross vaulting

(Fig. 433). The original rude capitals in the nave and choir are Corinthianesque,
with stiff, plain, crocket leaves which take the place of the angle volutes. Some

Fig. 431. Silchester. Portion of Roman mosaic. (From " Archaeologia," Vol. LV.)

have a crown of leaves, with a boss or rude human head instead of the flower

1 De Rossi, Inscnptiones Christianae urbis Komae.

2 Archaeologia, Vol. LV. St. John Hope and Fox, Excavations on the site of Hie Roman city at Silchester,
Hants, in 1895.

NORMANDY

(Fig. 434). Some of them have been re-worked, probably in the days when the
abbey was ruled by Vitalis (1060-1076), originally a monk at Fecamp and afterwards

Fig. 432. Corneto Tarquinia. Wall painting in an Etruscan Tomb.

abbot of Westminster (1076-1082), when it attained its greatest prosperity. Some
have even been treated with drapery, knobs, and Cherubim heads, in the fantastic
baroque style of the XVI I Ith century.

To judge by what is left, the external decoration of the building (Fig. 435) was
confined to a stringcourse at
the foot of the windows, and
occasionally billet mouldings
round the archivolts. The
walls were entirely without
buttresses. In short, it was a
structure of noble proportions,
but poor in mouldings and
carving.

The architect was William
of Volpiano (" Haec enim
auctore Guillelmo abbate
Fiscannensi . . .") 1 who per-
sonally directed the works at
their outset (" qui in locandis
fundamentis non modicum
praestiterat consilii auxi-
lium"). 2 As to the builders, I
imagine they were masons and
workmen from Normandy, as-
sisted by Italian or Burgundian
master builders, and acting
under their direction. What

1 Sammarthanus.Xc., GalliaChrist-
iana Abbatiae dtoecesis Lexouiensis
Bernaicus.

Du Monstier, op. cit.Bernayum.

- 433- Bernay. Abbey Church. South aisle of presbytery (1013).

F 2

LOMBARDIC ARCHITECTURE

Fig. 434. Bernay. Capital in the Abbey
Church (1013).

has survived of the work of Norman builders at a
date not far removed from that of the erection
of this church (e.g. at Jumieges) is more than
sufficient to prove the impossibility of their under-
taking a structure of this extent, or of constructing
the vaulting which we see in the presbytery aisles
at Bernay. A letter sent from Fecamp to Dijon
in the time of abbot William : mentions the
sending of master builders to Fecamp, that is to
the centre of the religious, intellectual, and artistic
life of Normandy. " De artificibus aedificiorum
nostrorum quae coepimus, vos obsecramus quo
. . . mittere ad nos festinetis, quia valde nobis
necessarii sunt." This request must have been
due to the small numbers and, still more, the
lack of skill of the Norman workmen.

As for the carving (I refer, of course, only to
the original work) I think we shall be right in

ascribing it to Norman artists, perhaps inspired by artists among the monks who had

learned how to handle the chisel

during the erection of the great

Burgundian edifice. At any rate,

it is not from an Italian hand

The crocket leaf is characteristic

of the Norman School, which

introduced it subsequently into

England.

This church provides mate-
rial for the following interesting

observations :

I. The plan of a Latin cross
may come from that of SS. Peter
and Paul, now Sant' Abondio
outside Como (Vth century), that
of the mausoleum of Galla
Placidia (about 440), and, even
more probably, that of Santa
Croce (about 449) at Ravenna.
These buildings, in their turn,
were derived from cruciform
Roman structures with a central
cupola supported by a circle of
columns, such as one illustrated
in Montano. 2 In our case the
plan shows two features worth
notice.

1 Paris, Bibliutheque Rationale, Coll.
de Bourgogne, torn. XI.

Fig- 437- Florence. Baptistery (Xlth Century)

NORMANDY

43S- Beinay. South side of Abbey Church as
seen from the prison yard (1013).

The first is that of the aisles prolonged beyond the crossing. There was a
precedent for the idea in the church of the Nativity at Bethlehem (IVth and Vlth
centuries), where the aisles were con-
tinued into the choir.

The other is that of the subordi-
nate apses projecting from the transept
outside the line of the choir aisles.
This arrangement seems to have some
relation to the recesses taken out of
the thickness of the eastern walls in
the transept of the underground church
of Saint Bt^nigne at Dijon ; if indeed
it were not borrowed from the Con-
stantinian Vatican Basilica, where, as
we learn from the plan made and pub-
lished by Alfarano in 1590, the tran-
sept had niches on its western side (the
side of the high altar), and in the end
walls, all used for chapels and altars.

These two features were afterwards
copied in the great abbey churches built
under William of Volpiano's immediate
supervision, or produced under the
influence of the School of Fecamp, of which he was the founder and for many
years the director.

II. The church affords the earliest instance of an arcaded wall-passage. Such

passages, designed partly for
purposes of communication, and
partly for decorative effect, were
suggested by passages like those
in the Aurelian walls of Rome
(Fig. 436). There is no trace of
them in any church earlier than
Bernay. Later, when they had
been put at the level of the
clerestory, in imitation of the
arcading in the interior of San
Pietro at Toscanella (739), and
embellished, they formed one of
the most striking and truly origi-
nal features of Lombardo-Nor-
man ecclesiastical architecture.

It has been suggested, in-
deed, that in the interior of the
Baptistery of Florence a wall-
passage with pairs of openings

Fig. 436. Rome. Aurelian Walls near the Porta I'inciana
(Illrd Century).

(Fig- 437) was constructed between the last years of the IVth century and the early
ones of the Vth. 1 But this celebrated building was really a result of the material

1 Nardini Despotti Mospignotti, // duomo di San Giovanni o&'i battisttro iti Firenze.

LOMBARDIC ARCHITECTURE

prosperity and religious zeal exhibited by Florence about the end of the Xth century. 1
It was not finished by 1057 or 1058, for otherwise it would have been consecrated by
Pope Victor II (1054-1057) or his successor Stephen IX (1057-1058), both of whom
died at Florence. The date of its completion was 1059, on the 6th of November
of which year it was consecrated by Pope Nicholas II (1059-1061) ; and the dedication
festival is still kept on that day. 2

The pairs of openings have no intimate connection with the masonry of the
building, and are therefore a later addition. The assertion that the presence of
Ionic capitals in these openings is decisive against a mediaeval origin both for this
baptistery and for all the other structures in Italy and the various European
countries, where they occur, 3 is wrong. I will only mention here the Ionic capitals
made expressly for the positions they occupy in the portico of San Lorenzo outside
the walls (1216-1227) at Rome, and in the external open galleries of the facades
of San Pietro (Xllth century) and Santa Maria Maggiore (1206) at Toscanella.

III. The church of Bernay tells us, thirdly, that in Normandy at the beginning
of the Xlth century great churches built in the new style, which was still in its infancy,
had wooden roofs, except for the ground floor of the aisles. And further, that the
characteristic Lombardic compound piers made their appearance in Normandy only
after they had been in use for years in Italy, where they were first produced.

THE ABBEY CHURCH OF MONT SAINT MICHEL is known under the names of
St. Michael "in monte tumba," St. Michael "in periculo maris," and St. Michael " de
monte " (Figs. 438, 439, 440). Originally a mere chapel or oratory built by St. Aubert

(about 708), bishop of Avranches,
it was rebuilt after a fire (1001) in
the time of Abbot Maynard II
(991-1009) with a subsidy from
Duke Richard II (996-1026), and
was re-founded in 1020, when
Hildebert II, a disciple of William
of Volpiano, was abbot (1017-

Fig. 438. Mont Saint Michel in the Xlth Century, (from
the Bayeitx Tapestry. )

1022). This date, coming between
the building of Bernay (1013) and
the rebuilding of the church of
Cerisy la Foret (1030), would be considerably more important for us than it is
had not the erection of the fabric been connected with dissensions between the
monks of Mont Saint Michel and the abbot of Fecamp, which delayed its completion,
and had it not in course of time undergone extensive alteration and rebuilding.

Suppo, abbot of Fruttuaria, was appointed (1023) by William of Volpiano to
succeed Hildebert II. But the opposition of the Norman monks, who objected to the
rule of a foreigner (and Suppo was William's nephew), prevented his taking possession,
and Almodus (1023-1031) and Theodoric (1031-1033), another nephew of William,
were appointed to administer the abbey. His installation did not take place till 1033
by Johannelinus, abbot of Fecamp (1029-1078). Suppo remained abbot, in spite of
the continual opposition of the monks, till 1048, when he went back to Fruttuaria.

These events were not of a kind to facilitate the progress of the new buildings
then in course of erection. The works, in themselves, had already demanded a great

1 Villari, 1 prim i due secoli della storia di Firenze.
3 Nardini Despotti Mospignotti, // duomo di San Giovanni oggi baitistero di Firenze.

NORMANDY

deal of time and trouble, on ac-
count of the massive substruc-
tions required for carrying out
Hildebert's design of construct-
ing on the summit of the conical
rocky eminence a platform on
which the church and conventual
buildings were to stand. So that
in 1048 the choir was hardly
finished, and, in fact, Abbot
Raoul de Beaumont (1048-1058)
was still engaged in 1058 in
erecting the piers and arches
for the central tower. Abbot
Ranulphe I de Bayeux (1060-
1084) worked on the nave (ac-
cording to some he entirely Fig 439 ._ Mont Saint Michel in the XVIIUh Century. (From
finished it), 1 which his successor, the " Annales Ordinis S. Benedicli.")

Roger I (1084-1106), completed

in its upper part (Labbe 2 says that he put on the roof), only to see the north side
collapse (iiO3), 3 Huynes stating that the nave fell down on that side. 3 The disaster

Fig. 440. Mont Saint Michel in 1904.

was repaired by Roger II (1106-1123), under whom the abbey was set on fire by
lightning (i 1 12), the vaulting and walls being left without any covering. Bernard

1 Sammarthanus, &c., Gallia Christiana -Abbaliae dioecesis Abrincensis S. Michael in periculo marts.
Nova bibliolheca maniiscriptoruin librarian De abbatibus M.S. Michaelis in periculo marts.
Robillard de Beaurepaire, Huynes, Histoire ge'ne'rale de tabbave du Mont Saint-Michel an Pt'ril de la mer.

7 2

LOMBARDIC ARCHITECTURE

Fig. 441. Mont Saint Michel. Abbey Church.
Capital (1048-1058).

de Bee (1131-1149) raised a bell-tower above the four piers of the crossing, and
Robert de Torigni (1154-1186) built two western towers, one of which fell shortly

afterwards, and added a porch to the west
front.

Such is, in brief, the generally accepted
history of the church of Mont Saint Michel in
the Xlth and Xllth centuries, as it may be
found in the latest publications on the subject. 12
Of this church (in the form of a perfect Latin
cross, like William's design for Bernay), damaged
on several occasions by fire, by the fall of the
central tower and of the choir which was re-
built in the Pointed style of the XVth and
XVIth centuries, deprived in the XVIIIth
of three out of the seven bays of the nave
and aisles which threatened to collapse, the
only portions that survive are the transept and the four easternmost bays of the nave

and aisles.

This history requires correction in one point, viz. as to the works carried out
under the eighth abbot, Raoul de Beaumont, originally a monk at Fecamp, and
appointed by William of Volpiano in
1028 to take charge of the abbey of
Bernay. For Raoul's operations can-
not have been confined to the piers
and arches forming the central bay
or crossing, but must have extended
to a part of the body of the building,
in order to abut the piers and enable
them to resist the weight and thrust
of the arches. An examination of
the structure during the last restora-
tion has in fact made it clear to me
that the two bays nearest to the
crossing are, owing to various fea-
tures which they present, evidently
the work of a different period from
that of the next bays ; and this
period is nearer to the first than the
second half of the Xlth century. In
these bays, the imposts of the two-
light openings in the south gallery
have a different outline from that in
the next bays. Further, one of the
original crocket capitals with a
human head (Fig. 441) is very simi-
lar to the earlier ones at Bernay
(1013) and Cerisy la Foret (1030-1066); while it is itself clearly older than those

Fig. 442. Mont Saint Michel. Abbey Church (Xlth and
Xllth Centuries).

1 Gout, Uhistoire et I' architecture francaise an Mont Saint-Michel.
- Corroyer, Description du Mont Saint-Michc.

NORMANDY

Fig. 443. Mont Saint Michel. Triforium in Abbey Church (Xlth and XI Ith Centuries).

in the crypt of the

T finite (1064-

1066), and others

in Saint Etienne

at Caen (1066-

1086), and must

therefore be dated

before 1064.

The body of

the church is

divided into nave

and aisles (Fig.

442) by cruciform

piers of uniform

size. The nave

had a wooden

roof, while the

aisles have rudely

constructed u n -

raised cross vaulting. The triforium (Fig. 443) gets its light from the nave, and is

covered by a sloping roof. The exterior is marked by lesenas which to some extent

have suffered from alteration (Fig. 444).

The low triforium without direct lighting, more lofty, however, than the one

at Bernay, which is also without windows, but not of large size with windows in the

aisle walls which are carried up
above the vaulting, as at Cerisy la
Foret ; and the absence of wall-
passages in the clerestory, as at
Bernay, while that system of com-
munication and decoration is em-
ployed at Cerisy la Foret, prove
that in the construction of the body
of the church the design of the
original choir (1020) was followed.
And that is the reason why we deal
with the building in this place, so
that it may form a link in the chain
of the history of the Lombardo-
Norman basilica.

Mont Saint Michel was the work
of the School of Fecamp founded
by William of Volpiano, and was
modelled on the abbey church of
Bernay, from which it differs by its
piers with roofing shafts, not only
in the choir but in the transept and

the whole of the nave ; by the lesenas

which strengthen the walls on the
Fig. 444. Mont Saint Michel. South side of Abbey ., , . / . ..

Church (Xlth and xilth Centuries). outside ; by the triforium with its

LOMBARDIC ARCHITECTURE

pairs of two-light instead of single openings, and greater space gained by carrying
up the aisle walls ; and by the absence of wall-passages in the transept. It exhibits
only one advance in scientific construction over its original, viz. the use of buttresses
at the points most in need of support, with the object of compensating for the
reduced thickness of the outer walls.

We will conclude by remarking incidentally that of the two well-known
churches with the title of Mount St. Michael the one just described, and that
at Chiusa (Fig. 445) on the summit of the Monte Pircheriano in the Val di Susa the

Fig. 445. San Michele della Chiusa. Abbey Church (Xllth Century).

latter no longer provides any materials for our studies. Of the original church
on Monte Pircheriano, founded, according to some, in 966, or between 999 and 1002
according to others, by Hugues de Montboissier, all that is left is the plan of a
basilica of small dimensions and irregular outline, with nave and aisles ending
in apses, beneath the floor of the present church, which was erected about the second
half of the Xllth century.

THE ABBEY CHURCH OF CERISY LA FORET was erected in 1030 by order of
Duke Robert II, called by some "the Magnificent" and by others "the Devil"
(1028-1035), to replace the church founded by St. Vigor, bishop of Bayeux
ar >d destroyed by the Normans. It was dedicated to him in 1032.

1 Du Monstier, op. cit. Cerasium.

NORMANDY

Fig. 446. Cerisy la Foret. Abbey Church,
in the nave (1030-1066).

Capital

The primary operations were directed by the abbot Durandus (1030-1033), previously a

monk of St. Oucn at Rouen, and after him by Almodus (f 1033) who had been

in charge of the abbey of Mont Saint

Michel (1023-1031). After the death of

Robert II the building was continued

and finished (with the exception of the

west front, where the towers were left

incomplete) by Duke William II, the

Bastard (1035-1066), afterwards King

William 1, the Conqueror (1066-1087),

through the instrumentality of the third

abbot Garin (1032-1066) who had been

a monk of Mont Saint Michel.

Altered in the XHIth century by

the vaulting of the presbytery, nave, and

transept, at which time the west front

was modified by the addition of a porch;

damaged in the XVIIIth century by the collapse of a part of the north arm

of the transept, and by lightning in the XVIth, XVIIth, XVIIIth and XlXth

centuries, it was partly demolished in 181 1, and reduced to its present condition.

Generally speaking its plan is
a copy of those of Bernay and Mont
Saint Michel. The only difference
is the feature of the piers which
bisect the extremities of the tran-
sept and carry galleries. Of the
structure of Robert II and Wil-
liam II there remain, three of the
eight original bays of the nave, the
last of which was flanked by the
two western towers, the whole of
the upper limb of the cross, and
the transept, of which the northern
arm has been mutilated.

In the nave the arches spring
from compound piers, alternately
larger and smaller, cruciform in
section with four half-columns and
the same number of angle shafts.
The capitals are Lombardo-Norman
Corinthianesque, not continuous,
however, like the Lombardic, but
confined to the columns. They
occasionally recall those with figure
subjects in San Flaviano at Monte-

Fig. 447. Cerisy la Foret. Abbey Church.
(1030-1066).

Triforium

fiascone, and the vegetable forms
are treated fairly well, though

without much undercutting. The animal representations, however, are very barbarous,

especially those of human beings (Fig. 446).

LOMBARD1C ARCHITECTURE

Fig. 448. Cerisy la Foret. Abbey Church. South arm
of the transept (1030-1066).

The aisles have cross vaulting.
Over them extends a spacious tri-
forium with wooden roofs (Fig. 447).
Above, in the clerestory, a service
gallery is constructed, with triplet
openings enclosed by a single arch
after the Ravennate fashion, and this
runs the whole way round the build-
ing. The nave and choir had origin-
ally a wooden roof, and the walls
were not tied together by transverse
arches as some have imagined. 1 2 Any
one can satisfy himself of this >who
takes the trouble to mount to the
present vaulting. The two lowest
stages of the central tower belong to
the original work. The south arm of
the transept (Fig. 448) still retains its
gallery carried on arches, and termi-
nates on the east in an apsidal chapel
opening out of both floors, not in its
original condition.

The aisles and triforium extend
for two bays east of the crossing. The
deep semi-elliptical apse (Fig. 449)
has a range of blank arcading round
its base. The middle stage contains
a wall-gallery in continuation of the
triforium, while round the clerestory
runs a service gallery. Originally it
had a wooden roof, but this was
replaced in the XHIth century by
vaulting in radiating sections.

The great display of tiers of
multiplied arches in this apse, giving
it an effect of severe majesty, has a
decorative as well as a practical ob-
ject which, it seems to me, was sug-
gested, on the one hand, by the desire
to give- a magnificent appearance to
the most sacred part of the church,
and, on the other, by the difficulty of
procuring marbles for lining the walls,
or artists capable of decorating the
vault with adequate paintings or
mosaics. The motive of loftv arches

Fig. 449. Cerisy la Foret. Choir of Abbey Church
(1030-1066).

Cerisy.

Dehio and von Bezold, op. cit.

De Farcy, Abbayes de rh.'c(hi lie Bayeiix-

NORMANDY

to decorate the internal wall of an apse is an old one. The apse of the " Bishop's
chapel," going back to the time of Frugiferus, first bishop of Trieste (524-about
568) and founder of the cathedral, 1 provides an instance of a range of arches
supporting the half-dome (Fig. 450).

The church is lighted by round-headed windows, splayed on the inside. The
walls of the exterior (Figs. 451, 452) show frequent use of herring-bone work, and
are strengthened by
buttresses. The lan-
tern tower is em-
bellished with blank
arcading. The fact
that this is partly
hidden by the main
roofs of the church
has suggested the
idea z that the nave
and transept had
originally a flat roof
covered with sheets
of lead on the out-
side, and with a
ceiling within. But
the XI th century
builders certainly
used gabled roofs in
such cases ; and the
architect would
never have followed
any other plan, con-
sidering the region
in which the church
was situated with all
its rain and snow,
under whose weight
a flat roof might
have given way.

It is not known

whether the plans Fig. 450. Trieste. Duomo. "Bishop's chapel" (524-about 568)

for Cerisy la Foret

were made by Durandus, or by him with suggestions from William of Volpiano,
who, just at the time when the rebuilding was begun, had come back from
Fruttuaria to Fecamp, where John of Aglie, called " Johannelinus," was then abbot
(1029-1078), and where William himself died in 1031. It is, however, certain that
the designs were modelled on two buildings of the School of Fecamp, Bernay, and
Mont Saint Michel, from which it proclaims its direct and immediate descent.

The builders were apparently Normans ; at least, there are no signs of any
Italian gild in any part of the work, from the rough, unraised, and heavy cross

1 Gams, op, cit.

2 De Farcy, op. cit.

LOMBARDIC ARCHITECTURE

vaulting, to the carving on the capitals. The craftsmen of Normandy were at last,
for better or worse, self-sufficing, though they may have had the aid of carvers who
were monks or came from some other part of France, e.g. Anjou, where as early as
1028 representations of living beings were being produced, as we noticed in the
church of Ronceray. Very different was the cross vaulting produced at the same
period in Italy, sometimes with bold ribs after the Lombardic fashion, as in San
Flaviano at Montefiascone (1032), which is nearly contemporary with Cerisy. Very

different, too, was
the carving, as San
Flaviano again
bears witness.

Cerisy shows a
notable advance
over Mont Saint
Michel in organic
forms, by the adop-
tion of two of the
typical character-
istics of the Lom-
bardic church. On
the one hand, the
triforium is lighted
direct! y from the
raised outer walls,
which by their
additional weight
strengthen the
lower part corre-
sponding to the
aisles. The other
Lombardic feature
is the alternation
of large and small
piers. Lombardic
influence is further
illustrated by the
characteristic
arcaded wall - pas-
sages. And, finally, it appears in the two stages of wall-passages round the apse,
continuing the triforium and clerestory passages. It thus marks another notable
step towards the perfection of the Lombardo- Norman basilica.

The characteristic arcaded clerestory passages of the nave, choir, and apse, which
combine their immediate utilitarian object with a constructive and static purpose,
viz. that of diminishing the weight of the upper part of the walls, and also with a
decorative intention, have the merit of absolute novelty. We saw their beginnings at
Bernay and in San Pietro at Toscanella (739). Before the erection of Saint Vigor,
service passages were either mere gangways protected by a railing, carried along
projecting cornices above the architraves of colonnades and galleries in large naves,
or else below the impost line of domes and half-domes. Such passage-ways are

Fig. 451. Cerisy la Foret. South side of Abbey Church (1030-1066).

NORMANDY

illustrated by the
old St. Peter's at
Rome (IVth cen-
tury) (Fig. 453), and
by St. Irene (VII Ith
century) and St.
Sophia (532-537) at
Constantinople.

Another novel-
ty at Cerisy is the
arcaded passage
formed in the thick-
ness of the apse
walls, and con-
tinuing the tri-
forium. It was sug-
gested by apsidal
galleries like the
upper stage of the
apse of Santo Ste-
fano at Verona (Xth
century), and by the
blank arcade in the
apse at Germigny
des Pres (801-806).

We should also
notice the feature of
the transept gal-

Fig. 452. Cerisy la Foret. Abbey Church from the south-east (1030-1066).

leries, which first appeared in the Lombardo-Norman basilica, and was suggested by
the two porticoes at the ends of the transept in the Constantinian Vatican basilica.

In the same way we should note the first appearance in Normandy of animal
figures on capitals. In Italy, after making a timid display in San Babila at Milan

fv . i} early in the Xlth cen-
tury, they had at-
tained considerable
development by 1032
in San Flaviano at
Montefiascone ; but in
France they did not
appear till somewhat
later. Notre Dame do
la Charite at Angers.

^_-_ consecrated in 1028,

affords a very early
dated example. If its
capitals betray Lom-
bardic influence, the
handiwork of French
carvers is no less

A I B

Fig. 453. Rome. Old St. Peter's (IVth Century). (From Bonanni,
" Tcnipli Vaticani Historia.")

LOMBARDIC ARCHITECTURE

obvious. To mention only one point,
the crocket leaves were not the work
of any Italian chisel.

THE ABBEY CHURCH OF
JUMIEGES, begun in 1040 by the
well-known Robert II, one of the
advisers of Edward the Confessor
(1042-1066), for a short time bishop
of London (1044), afterwards arch-
bishop of Canterbury (1051-1052), and
from 1037 onwards abbot of Jurnieges,
was finished in 1066 and consecrated
in 1067, in the presence of William
the Conqueror (IO66-IO87), 1 - 3 by
Maurilius, archbishop of Rouen (1055
1067), a disciple of William of Vol-
piano and an Italian by birth accord-
ing to one account, though others say
that he was born at Rheims.

After the alterations of the XlVth
century, and the destruction in the first
years of the XlXth, all that is left of
the original building is the west front
(Fig. 454), the body of the church as
far as the crossing, and a portion of
the latter and the central tower (Figs.
455, 456). Within the front, flanked

by a pair of very lofty towers, a narthex of two stories is formed. The carving

of the two Lombardic capitals (Fig. 457) in the upper one is later. In the towers,

the square base passes into an octagon, and the octagon into a circle, by means of

conchiform squinches or hood-shaped raccords. The nave, which was designed for

a wooden ceiling, has eight rou nd arches on either

side, and square piers with three half-columns

attached (the half-piers of Pointed character on

the nave side are a XlVth century addition), alter-

nating with cylindrical piers which are not tapered.

In all these supports, and the corresponding ones

in the outer wall, the columns are surmounted by

quadrangular funnel-shaped cubical capitals, which

sometimes have a ridge down the middle, and the

angles hollowed out. The left aisle still retains its

original unraised cross vaulting of rough construc-

tion, above which is the triforium with similar

fig. 454. Jumieges. West front of Abbey Church
(1040-1066).

1 Rolls Scries Willelmi Ulaltiicsbiricnsis monachi ges/a regum
Anglorum.

- Du Monstier, op. dt.Gemetictun.

3 Loth, Histoire dc fabbaye royals de Saint-Pierre de
jumiiges.

g. 457._J U mieges. Abbey Church.
Capital in upper narthex (Xlth or
Xllth Century).

NORMANDY

vaulting. The latter, on the nave side, has triple openings enclosed by a single arch
after the Ravennate fashion.

On the outside, the north wall of the aisle and triforium, which is visible and in
good preservation, is strengthened by massive buttresses corresponding to the wall
piers within (Fig. 458). Their present stepped form is the result of later alterations.
The central tower, which is square and not octagonal as stated by Cordero, 1 had

Fig. 455. Jumieges. Nave of Abbey Church (1040-1066).

originally a wooden roof. The ruins of the transept show that the side walls were
pierced in their highest part by passages.

Notwithstanding all the injuries inflicted by time and human agency, the latter
being by far the most serious, enough is left of this impressive church to compel the
spectator on the first view to pause and admire, not indeed the elegance of its
decoration or the grace of its forms, but its severe lines, its noble proportions, and the
grandeur of the whole effect.

1 op. at.
VOL. II G

LOMBARDIC ARCHITECTURE

The church of Jumieges, rude but excellent example as it is of the method of
building prevalent in Normandy about the middle of the Xlth century, displays a

Fig. 456. Jumieges. Nave and transept of Abbey Church (1040-1066).

marked advance on the way towards the perfection of the Lombardo-Norman
basilica, by the extension of cross vaulting to the aisles and triforium, the angle
thrust being met by substantial buttresses outside. Choisy 1 speaks of this as a

risky undertaking, and
one that must appear so
when \ve think of the de-
pression of the vaulting
and the consequent in-
crease of thrust, and also
of the time when it was
done. And it could only
be carried out at the ex-
pense of the direct lighting
of the triforium.

Next, we may notice
the hood-shaped raccords

F 'g- 459- Corneto Tarquinia. Tainting in Etruscan tomb.

Histoire de F architecture.

NORMANDY 83

in the western towers, the earliest instance of this form north of the Alps to which I
can point. The central tower is also notable because, though the explosives intended
to destroy it at one blow in 1802 have spared only one of its four sides, what remains
makes it clear that it had a roof of wood and not of masonry, and, with the traces left
of the nave roof, enables us to restore its form and that of the ceiling below it
details which are of interest for the history of Norman architecture.

Fig. 458. Jumieges. North side of Abbey Church (1040-1066.)

Nor should we omit to notice the presence of a decorative form not previously used
in ecclesiastical architecture, viz. the bands of chequer pattern, so frequently reproduced
later in Normandy and England, and finally applied by the architect Lanfrancus to the
capitals in the cathedral at Modena (1099-1106). This chess-board motive was a
favourite one with the Etruscan artists, who often employed it in tomb-paintings
(Fig 459). The Romans applied it specially in mosaics.

On the other hand, we must not ignore the retrograde step taken by the designer

G 2

84 LOMBARDIC ARCHITECTURE

in going back to some of the forms of the church at Bernay (e.g. the absence from the
nave piers of roofing shafts rising to the point where they could carry the main beams
of the roof, and the lack of wall-passages with open arches springing from thick piers)
which are a negation of the progress we remarked at Mont Saint Michel and Cerisy
la Foret. This retrogression may, perhaps, have been due to the theories of construc-
tion which Robert II had learned in the school of William of Volpiano, and also,
in the case of the piers, to the desire of keeping the nave freer by suppressing the
wall shafts with their bases, and relying on the considerable thickness of the nave
walls and the substantial buttresses outside for resisting the weight and thrust of the
main beams of the roof.

Lastly, we must notice the alternation of massive piers with squat, untapered
columns, an arrangement which demands some remark.

The organic conception of supports alternately substantial and slight, which was
introduced in the first Lombardic vaulted churches, had no connection with the
device of pairs or triplets of columns alternating with piers, as in Santa Maria in
Cosmedin and Santa Prassede at Rome, in spite of the reiterated assertions to the
contrary of so many writers, one of whom, Enlart, 1 adds that the expedient was the
origin of a whole constructive system. For the last restoration of Santa Maria in
Cosmedin has revealed the fact that the church of Hadrian I (772-795) was rebuilt
in the Xllth century, only some of the oldest parts being preserved. And the piers
in Paschal I's (817-824) church (in the shape of a "crux commissa" where the
transept forms the horizontal limb of a T) are the result of one of the remodel-
lings which the building underwent. The alternate large and small supports found
in Normandy are really due to Lombardic influence. In Italy it occurs as early as
985 in SS. Felice e Fortunate near Vicenza, and in 1013 at San Miniato near
Florence.

The church of Jumieges forms the climax of the series of Lombardo-Norman
buildings erected under the auspices of William of Volpiano. We have now to
watch the rise of another series, essentially as important as the first, and superior
to it in the intrinsic value of its results ; one, too, in the course of which the
style in question will be seen to take fresh and important steps towards
completion.

Lanfranc, born at Pavia (1005) and educated in its venerable and flourishing
school, the centre of Latin culture at the time, had left his home for France accom-'
panied by a band of colleagues and disciples, and opened a school at Avranches
(1039). Then, quite unexpectedly, he abandoned teaching, shut himself up in the
monastery of Bee (1042), erected about that time by Erluin, previously abbot of
Burneville (1034), and there received the coarse habit of a Benedictine monk.
Thanks to him, this obscure convent soon became the intellectual centre of the
Christian world. To the school which he there opened, and of which he was the life
and soul, laymen of every station flocked from the most distant regions ; the great
sent their sons to it ; the most famous masters, the most profound dialecticians
regarded it as a special privilege to be allowed to frequent it. 234

1 Op. cit.

2 Duchesne, Hist. Nonnann. script. antiquiWillclmi Calculi historia Normannonun.
8 Le Prevost, op. cit.

1 Mabillon, Ada Sanct. Ord. S. BenedictiCrispimis Milo Vila B. Lanfranci.

NORMANDY 85

Erluin's buildings at Bee being found too small for such a concourse, new and
more spacious ones took their place, and were consecrated in 1077. Lanfranc, who
had been appointed prior (1045-1066), designed them and began the work, which was
finished by Anselm. 1 Nothing is left of this new structure, which is all the more to
be regretted because it was the first essay in architecture of the creator of Saint
Etienne at Caen and many other famous churches. And so it comes about that we
are obliged to make Saint Etienne itself the point of departure for the architectural
epoch which succeeded that of William of Volpiano, and must be called the epoch of
Lanfranc.

THE ABBEY CHURCH OF THE TRINITE AT CAEN. The exact date of its
erection is not known. Du Monstier 2 puts it in 1064. It was dedicated in

Fig. 460. Caen. Abbey Church of the Trinite. Crypt (1064-1066).

io66. 3 But the dedication clearly implied
Duke William* ("coniux mea Mathildis . .

the

by the well-known charter of
construxit basilicam ") can only

Fig. 461. Caen. Abbey Church
of the Trinite. Capital in ihe
crypt (1064-1066).

refer to the choir and crypt beneath it. It must
have been finished by 1082, for in the deed of
foundation which appears in Du Monstier we read
that William the Conqueror and Queen Matilda
had built the church (" ecclesiam . . . pro salute
animarum nostrarum coaedificavimus "), and by the
same deed the abbey was endowed with a noble revenue
for its support.

The only part of the original building preserved
intact is the crypt (Figs. 460, 461). The church as a

1 Migne, Pair. Lat., Vol. i$oCAronii Bccccmis abbaliae.
' l Op. fit. Cadomits.

* Le Prevost, op. fit.

* Sammarthanus, &c., Gallia Christiana fiislriu/ienta Etc/csiac

, usis.

86 LOMBARDIC ARCHITECTURE

whole (speaking of the parts that are old) is the result of a general remodelling
in the Xllth century (Figs. 462, 463).

THE ABBEY CHURCH OF SAINT ETIENNE AT CAEN was, like the Trinite,
founded in expiation of the marriage contracted by Duke William II with Matilda
of Flanders in face of the prohibition decreed by the Council of Rheims (1049). About
the dates when it was begun and when it was consecrated, there is disagreement

Fig. 462. Caen. Abbey Church of the Trinite (Xllth Century).

among the contemporary writers and chroniclers. Du Monstier 1 places the former
in 1064. But seeing that Lanfranc, the architect of the building (his subordinate
was his pupil Ralph, afterwards prior of Caen), was appointed the first abbot of
Saint Etienne in io66, 2 we may infer that the latter year is the more probable date,
if not of laying the first stone, at any rate of the beginning of building operations
on a large scale. As to the consecration, the date 1077, given by Ordericus Vitalis, 3
must be that of the choir, transept, and part of the nave ; while that of 1081 or

1 Op. cit. Cctdoinus.

Le Prevost, op. cit.

NORMANDY 87

1086, which appears in Du Monstier, will refer to some minor ceremony l relating to
the completion of the building and the alterations effected in the west front and the
towers.

What is left of the original church comprises the western limb of the cross with
the west front and internal narthex and the transept. The eastern limb (except
part of the triforium) and the apse, together with the apsidal chapels in the transept
were sacrificed to the rage for the Pointed style.

The body of the church, entered through a two-storied narthex confined between
the western towers, is divided into nave and aisles by an arcade with compound

Fig. 463. -Caen. Abbey Church cf the Trinite (Xllth Century).

piers alternately larger and smaller (Fig. 464). The half-columns belonging to these
supports (Fig. 465) and those in the triforium (Fig. 466) are surmounted by
Corinthianesque capitals supporting a continuous abacus, and ornamented with plain
crocket leaves, carved foliage, projections either plain or graduated, and with human
or semi-human heads. These capitals, though their conventional treatment betrays a
want of imagination in the carvers, are nevertheless well designed, with free movement
and bold relief, but not completely undercut. The same, however, cannot be said of the
figures ; for no sooner do the carvers leave the treatment of foliage than they fail and
become clumsy. The heads, something between man and beast, with erect pointed
ears, should be noticed, for this stylistic representation, with another which we shall

1 Freeman, The History of the Norman Conquest of England,

LOMBARDIC ARCHITECTURE

see on the piers of the central tower, crossed the Channel, and served as a model
for the carvers of Great Britain, e.g. those who worked in Durham Cathedral
(1093-1133).

The aisles, originally covered by unraised cross vaulting, have triforium galleries
over them, roofed by ribbed half-barrel vaults (Fig. 467). If this barrel vaulting is
not original, everything shows that it merely replaced an older roof of the same kind.

Fig. 464. Caen. Abbey Church of Saint Etienne (1066-1086).

Nor can Ruprich-Robert's x idea that it was unnecessary be maintained, for at Jumieges
(1040-1066) it was precisely the triforium that was vaulted, while the nave had a
timbered roof. Its object is easily explained by the architect's desire to increase the
stability of the nave walls, which, though about 4 feet 6 inches thick, were considerably
weakened by the wall-passages with their wide arches, and also by the external blank
arcading. The expedient adopted at Saint Etienne was afterwards followed by Serlo
in the choir of Gloucester, rebuilt in 1089.

1 Les voiites de rablaye-aux-hoiiunes a Caen.

NORMANDY

Above the triforium on either
side runs a clerestory passage with
triplet openings, the result of an
alteration which is believed to have
been carried out in the Xllth cen-
tury, when the existing vaulting
was substituted for the original
ceiling of the nave and choir. In
the course of this, new clerestory
arches were opened lower down,
with the object of enabling the
walls to offer more resistance to
the thrust of the vaulting. Their

Fig. 465. Caen. Abbey Church of Saint Etienne. Capitals in
nave (1066-1086).

Fig. 466.
Etienne.

Caen.
Capitals

Abbey Church of Saint
in triforium (1066-1086).

verted into shafts from which
spring the ribs of the groining.

Above the present vaulting may
still be seen the continuous impost
cornice at the height of the original
ceiling ; and in the north wall there
are traces of two of the original
openings of the service passage, each
of four lights, with roll mouldings,
and uniform in size like those at
Cerisy la ForSt.

In the central tower, the arches
of which are not in their original
state, the capitals of the piers exhibit,
besides the usual single or combined
leaves, a human head, and another
characteristic semi-human one with
pointed ears and no lower jaw, but a
row of dog's teeth in the upper one.

capitals are in the form of truncated and inverted
half-cones and wicker baskets, or else they show
monsters with serpents' tails, foliage, interlacings,
lions, monsters with demon heads and limbs
ending in cauliculi. The archivolts are ornamented
with an embattled moulding sunk in the thickness
of the wall. All the instances of the embattled
and zigzag moulding which appear on the in-
terior walls of the church are later additions. At
the same time was inserted the continuous course
of billet moulding carried round the capitals of
the vaulting shafts ; and the capitals themselves
were carved. The half wall-piers were also con-

Fig. 467. Caen. Abbey Church of Saint Etienne.
Triforium (1066-1086).

LOMBARDIC ARCHITECTURE

The ends of the transept are occupied by galleries with unraised cross vaulting
(Fig. 468). Here the figure carvings are barbarous in design and no less rude in
execution, and teach us that if the Norman craftsmen had by this time learned to
treat pure decoration with some grace, they were still far from being at home in dealing
with the human form.

In the triforium of the existing choir are to be seen part of an arch parallel to the
axis of the building, and some Corinthianesque capitals, both belonging to the original

work.

Bouet 1 argues that the original choir had cross vaulting, but there is not the
smallest evidence to support the idea.

Fig. 468. Caen. Abbey Church of Saint Etienne.
Gallery in north arm of transept (1066-1086).

Fig. 469. Caen. North side ot Abbey Church of
St. Etienne (1066-1086).

The exterior of the nave, aisles, triforium, and central tower is embellished with
blank arches and arcading (Fig. 469). The west front (Fig. 470) contains three
Lombardic portals (Fig. 471) which have been partly rebuilt. The three stages of the
towers, with their wall arcading, must be assigned to between 1081 and 1086; the
upper portions were completed at later dates.

Saint Etienne, though, on the whole, it belongs to the type established by
William of Volpiano, presents certain organic and decorative features which
differentiate it from that type ; and some of these mark a new step towards the
completion of the Lombardo-Norman basilica.

In this church Lanfranc remained content to carry the vaulting shafts up to

1 Analyse architectm-ale de Tabbaye de Saint-tienne de Caen.

NORMANDY

the roof, hinting at a desire to cover the
nave with cross vaulting, which, however,
he never fulfilled (in spite of Bouet's l
idea to the contrary), deterred perhaps
by the inefficiency of his master masons.
For the primatial church of Canterbury,
which he began shortly after (1070), had
a wooden roof; and similar roofs, without
even the support of transverse arches,
were provided for the first great churches
built in England by his pupils. With this
new arrangement goes that of the single
arch openings of the triforium. Other
peculiarities are the following :

I. The Ravennate motive of ex-
ternal arcading corresponding to the nave
arches in the interior. Such applied
arcading appeared in Saint Benigne at
Dijon, but it does not occur in the naves
of the great Norman abbeys.

II. The blank external galleries
forming a finish to the side walls of the
church. I cannot remember any earlier
example of this feature in Normandy.
In another part of France, an older in-
stance is presented by the abbey church of
Roncerayat Angers (1028). Some writers

give
the
Nor-
mans credit for its invention, whereas it is a crea-
tion of the Lombardo-Ravennate style.

III. The arched corbel course forming a
continuous cornice : the first of its kind to appear
in a Lombardo-Norman church.

IV. Lastly, there are the portals of the west
front. This is the first occasion on which we can
verify their appearance in a building of the
Lombardo-Norman style.

Further, we must not omit to notice the
remarkable progress shown in the carved foliage
of the capitals in Saint Etienne, compared with
work of the same kind in earlier Norman build-
ings. These capitals are differentiated by their
artistic quality, not only from the contemporary
ones in the crypt of the Trinit^ at Caen, but also
from those of the same date in the crypt of the

Fig. 470. Caen. West front of Abbey Church of
Saint Etienne (to66-lo86).

Fig. 471. Caen. Abbey Church of Saint
Etienue. Portal in west front (1066-1086).

1 De Caumont, Bulletin monumental, 1863 Seconde lettre
M. de Caumont an sujet des vaults de Saint- tienne de Caen.

LOMBARDIC ARCHITECTURE

cathedral of Bayeux, rebuilt by Bishop Hugh about the year 1044, finished and
decorated by Odo I (1050-1097), and finally reconstructed after the fire of 1105.

This latter crypt is not the work of Hugh or Odo, as has been thought, 1 z
some 3 actually believing that the existing church is Hugh's work and not a
reconstruction. Of the time of Hugh or Odo I (who was, according to Gams, 4
the builder of the church, which he consecrated in 1080) nothing is left save the
outer walls, and some of the isolated supports with their bases and capitals showing

Fig. 472. Bayeux. Cathedral (Xllth and Xlllth Centuries).

plain, stiff, turn-over leaves and crocket leaves. In some cases they have been
re-worked and embellished by carving the leaves. The rest is a reconstruction
carried out by Bishop Philippe d'Harcourt (1142-1164), and of this there remains the
Lombardo-Norman part of the nave (Fig. 472), the church having been largely
rebuilt in the Xlllth century. To the time of Bishop Philip belong, for instance,
a capital with foliage and human heads with cauliculi protruding from their mouths,

> Mylne, The Cathedral Church of Bayeux. 2 Freeman, op. cit.

Beziers, Memoire$ pour servir a fetal historique et gtographiqw au dioche de Bayeux. 4 Op. cit.

NORMANDY

and another with inverted half-
cones surmounted by a denticu-
lated cornice, this last type of
capital not having been introduced
when the crypt was built, as it
appears in France for the first time
at Boscherville. To the same date
belong the bases of Pointed charac-
ter, and the beautiful continuous
unraised cross vaulting. The Nor-
man builders of the Xlth century,
though fairly skilful on the whole,
were still indifferent constructors
of vaulting. Ruprich- Robert ] has
called attention to their weakness
in this respect, even during the next
century.

THE CHURCH OF SAINT
NICHOLAS AT CAEN, though not
erected from the designs of Lan-
franc, is immediately derived from
his church of Saint Etienne, and
is traditionally said to have been
planned by monks who were his
pupils, the same builders being
employed who erected the abbey
church. It will, then, be convenient
to describe it next after the latter.

It is said to have been begun
about the year io8o, 2 when the
work on the west front of the
neighbouring Saint Etienne was
being brought to a conclusion, and
is believed to have been finished in
1093. It is now a military store.

The plan shows that it be-
longs to the type which we con-
nect with the name of William of
Volpiano. It has a nave and aisles
of seven bays, divided by six com-
pound piers similar in section to
those at Cerisy la Foret. The
capitals have foliated volutes or
crockets, and crowns of leaves going
all the way round, with projections
representing the flower. This type

1 Op. dt.

2 Bouet, op. fif.

Fig. 473. Caen. Saint Nicholas. South aisle (1080-1093).

Fig. 474. Caen. Saint Nicholas.. Chapel in south arm of
transept (1080-1093).

LOMBARDIC ARCHITECTURE

of capital prevails all through the building. Occasionally we find a human or
half-animal head with erect ears (rough work, evidently from the same chisel that
produced those in Saint Etienne), that of a lion, and of some indefinable creature.

The aisles (Fig. 473) have unraised cross vaulting. In the XVth century
the timbered roofs of the nave and transept were replaced by Pointed vaulting.
The interior of the central tower is embellished with two tiers of wall-passages.
The transept, on the other hand, has ranges of deeply recessed arches without

communication, just
below the clerestory.
Similar blank arcading
decorates the apsidal
chapels which project
from their eastern
sides (Fig. 474). The
choir, together with
the nave, has a sham
triforium correspond-
ing to the aisle roofs,
and has cross vaulting
like the aisles. The
apse has arcading in
two tiers, one of which
forms a passage. The
exterior of the church
^^ is strengthened by sub-

stantial buttresses, and
is decorated with ar-
cading (Figs. 475,
476). The west front
is flanked by two
towers, originally left
unfinished.

Saint Nicholas
marks another step
towards the perfection
of the Lombardo-Nor-
man basilica by having
the parts east of the
crossing covered by

cross vaulting with visible arches springing from compound piers, and supported by
buttresses outside. Further, it exhibits an entirely new feature in the triforium, which
is not an open gallery but a mere series of tall, narrow openings, almost like loops..
This arrangement (followed at Boscherville), which had the advantage of consoli-
dating the walls by the omission of wide openings, combined with the shortness
of the eastern limb, and the support provided by the adjoining transept and
central tower (in itself an important element of stability), must have given
confidence to the architect and builders in taking the step, which for them
was a bold one, of covering the whole of the eastern limb of the church with cross
vaulting. Lastly, in the apse we find for the first time the decorative treatment of

Fig. 476. Caen. Saint Nicholas. Apse (1080-1093).

NORMANDY

large arches enclosing pairs of
smaller blank ones. It was copied
immediately afterwards at Boscher-
ville, and later in other churches.

THE ABBEY CHURCH OF
SAINT GEORGES AT BOSCHER-
VILLE. As Saint fitienne was the
parent of Saint Nicholas at Caen,
so Saint Georges at Boscherville is
the immediate descendant of the
latter, and that is the reason for
its inclusion in our survey.

Some writers would make out
that the existing building is identi-
cal with the cruciform church men-
tioned in a well-known deed, 1 exe-
cuted between 1053 and 1066, of
Duke William II (1035-1066), and
believed to have been erected about
the year 1050 by Raoul de Tancar-
ville to replace an earlier chapel, of
uncertain age, dedicated to St.
George. Others, on the contrary,
regard it as a reconstruction of

Raoul's church, carried out either when the Benedictine
monks of Saint Evroult d'Ouche were installed in it
(1114), or in the last years of the Xlth century, or
rather its last quarter.

The ground plan (Fig. 477) follows the type of
'"A William of Volpiano: a Latin cross; nave and choir
j with apsidal termination ; aisles prolonged east of the
crossing, and flanking the presbytery ; a transept with
minor apses projecting from the east side, and galleries
at the ends. With the exception of the clerestory
above the triforium, the scheme is a copy of that of
Saint Nicholas at Caen: the same triforium, without
direct light and formed under the slope of the roof,
and similar vaulting for the choir.

The piers of the nave arcades have the same
section as those at Cerisy la Foret and Saint Nicholas
at Caen. The pillars in some cases have Lombardic
cubical capitals of the Sant' Abondio (1013-1095) type,
but mainly capitals of a new pattern derived from a
combination of the cubical Byzantine melon form and
the aforesaid Lombardic capital. This form, with an
entirely new inspiration, displays, in place of the

- 475- Caen. Saint Nicholas. South side (1080-1093).

F 'g- 477- Boscherville. Plan of
Abbey Church (Xlth and Xllth
Centuries).

1 Besnard, Montgrepkit de Nglise et de Fabbaye de Saint-Georges de

Boscherville.

9 6

LOMBARDIC ARCHITECTURE

Fig. 478. Jumieges. Church of Saint Valentin (Xllth Century).

segments of a
melon, a sort of
undulation con-
sisting of shallow
inverted semi-
cones in relief.
The primitive
form of this capi-
tal is to be traced
to England,
where it appears
in the Anglo-
Lombard ic cubi-
cal capitals of St.
Albans (1077-
1088). There we
find cubical capi-
tals with each
face exhibiting
not a single half

round, as in the Lombardic type, but two portions of rounds. This new capital

(the real origin of which has not hitherto been pointed out), described in France as

"godronne" and in England as

" scalloped," while my name for it

is the Anglo-Lombardic cubical

capital, appears in its earliest form

in the abbey church of St Albans,

and in more elegant and perfect

shape in Gloucester Cathedral. The

oldest instance of a scalloped capi-
tal, though without the refinement

of that at Gloucester, which I can

point to in France is to be found

on the cruciform piers which carry

the first transverse arch in the

parish church of Saint Valentin at

Jumieges (Fig. 478), built by abbot

Ursus (noi-1127), 1 and original

from the west front as far as the

said arch.

The new capitals of Boscher-

ville, in some cases merely blocked

out, are either quite plain, or else

exhibit pointed, stiff, smooth leaves,
stars, studs, and other ornaments.

We also notice capitals decorated

with the usual crockets, bunches of

1 Loth, Jumieges.

Fig. 480. Boscherville. Abbey Church. North arm of
transept (Xlthand Xllth Centuries).

NORMANDY

cauliculi, human heads, stylistic palmettos, interlaced circles, discs, cables, and
monsters of all sorts. The foliage is fairly well treated, but not undercut ; the figures,
on the other hand, are very uncouth and barbarous. The bases, in some cases
decorated with zigzags and other ornaments, are almost all provided with angle-spurs
in the shape of claws, leaves, discs, &c.

The body of the church (Fig. 479) was originally spanned by transverse arches
with the object of resisting any inward tendency of the walls. Its roof was of timber,

Fijj. 479. Boscherville. Abbey Church (Xlth and Xlhh Centuries).

the present vaulting having been constructed in the XHIth century. The side aisles
had from the beginning unraised cross vaulting. Above the aisles runs a shallow
triforium, constructed under the sloping roof, and originally lighted by narrow oblong
openings contained within arcades of four arches to a bay. These have spurred bases
and capitals ornamented with semi-cones, foliage, human and animal heads. These
openings were blocked up later, perhaps when the present Pointed vaulting was
constructed. A service passage is formed in the clerestory. The galleries at the
ends of the transept (Fig. 480) have interesting figure capitals (Fig. 481).

The deep apse, encircled by blank and open arcading, was originally covered with
VOL. II H

LOMBARDIC ARCHITECTURE

Fig. 481. Boscherville. Abbey Church.
Capital in transept (Xlth and Xllth
Centuries).

a half-dome, to which ribs have been added at a later date. The two bays of the
choir in front of it have ordinary unraised cross vaulting.

The most remarkable features of the ex-
terior (Figs. 482, 483) are the west front (Fig.
484) with its Lombardic portal, the chief
ornament of the church, and the arcaded apse.
And now let us turn to the vexed question
of the date of Saint Georges. Attempts to
fix it in an approximately definitive or, at
any rate, an acceptable manner have hitherto
been generally unsuccessful owing to a failure
to bring the right buildings into comparison
with it.

In my opinion, the existing church is
certainly not the one mentioned in the
foundation charter, which must have been
a building of modest dimensions, but a new

structure, carried out on a larger scale and under more advanced artistic conditions,
before the year 1114, i.e. while the Augustinian Canons were still in possession, and
after the completion of Saint Nicho-
las at Caen (1080-1093) ; in other
words, between the last years of the
Xlth and the first of the next cen-
tury. With the exception of the
west front, it was built all of a
piece, and on a predetermined
plan : the roll mouldings employed
throughout are sufficient to prove
this. The violent contrasts some-
times exhibited by the carving are
the result of later work, executed
when the west front was erected
under a new artistic impulse.

It must be later than Saint
Nicholas at Caen, as it shows an
advance in construction beyond
that church in the form of the
transverse arches of the nave. Its
Anglo - Lombardic scalloped capi-
tals must be later than the erec-
tion of the cathedral of St. Albans
(1077-1088), where their primitive
form appears for the first time.
Moreover, the date of the body of
the church is earlier than that of the
west front. There is too great a
gulf between the carvings of the

latter with their free movement, and the clumsy forms in the church itself, for them
to be contemporary. The front in its turn is considerably later than that of Saint

Fig. 482. Boscherville. North side of Abbey Church (Xlth
and Xllth Centuries).

NORMANDY

Fig. 483. Boscherville. East end of Abbey Church (Xlth
and Xlhh Centuries).

Etienne at Caen, for the carvings of the portal of Saint Georges (which exhibits
multiplied mouldings of much greater richness than the doors of Lanfranc's church)
reveal a more developed art and
a surer hand than the doorways
of Saint Etienne. It is also de-
cidedly later than the chapter-
house of the abbey, erected by
Abbot Victor (1157-1211) after
1 1 57, where (making allowances
for the Pointed forms of the
structure) the beautiful carved
foliage and figures are very differ-
ent both in style and execution
from the work on the portal in
question. Hence we may rea-
sonably assign it to the first
years of Abbot Louis (i 1 14-1 157)
(" . . . . ingenii sanctitatisque
fama celeberrimum clarissimum-
que . . ." J ) ; and to it, together
with the re-working of some
of the carving in the church,
will refer the words "ecclesiam ampliare et meliorare " in the well-known
charter of Henry I of England (i IOO-H35). 2

Saint Georges marks an ad-
vance in the principles of equi-
librium and construction beyond
Saint Nicholas at Caen. The in-
troduction of cross arches with
walling above them served to bind
the walls together, and also to
provide a solid support for the
roof; in case of fire, too, they
would prevent the flames from
spreading, and the nave walls from
falling in. A very early instance
of such arches, which form the
germ of a revolution in the system
of covering a church the substi-
tution of stone for timber is to be
seen in San Miniato al Monte, near
Florence (1013) ; and the primitive
church of SS. Felice e Fortunate,
near Vicenza (985), provided a still
older one.

1 Gaignieres, Travail tie aom Michci Ger-
main S. Georgii de Bacherivilla monasttrii
primordia et nostrum ad aetatem eventus.

8 Deville, Essai hist, et descript. sur Ctglise
et Fabbaye de Sainl-Georgcs-de-Bos(hervilIe.

H 2

Fig. 484. Boscherville. West front of Abbey Church
(Xlth and Xlllh Centuii. i).

1OO

LOMBARDIC ARCHITECTURE

We should notice also on the bases the free use of the characteristic protective
spurs, which are not met with in Normandy before the second half of the Xlth

century.

Lastly, the blank arcading which decorates both the interior and exterior of the
apse of Saint Georges recalls an ancient circular structure treated in the same way
the vestibule of the "Piazza d'Oro" in Hadrian's Villa at Tivoli (125-135) (Fig. 485).

This vestibule is of no
small importance in the
history of architecture, for
we may be sure that some
Roman building of the
same kind, derived from
that in Hadrian's Villa, was
the original which served
as a model to the builders
of the first mediaeval
apses decorated with blank
arcading.

The mention of
Hadrian induces me to
lay before the reader an
original view of the Em-
peror and of the manner
in which he influenced the
buildings erected by him.
My object is to bring into relief one of the eminent qualities of an illustrious
ruler whose deficiencies are usually more noticed than his estimable sides, viz. his
ability as a master architect. This has, indeed, been previously suggested, 1 but
as yet the facts have never been established.

Aelius Spartianus 2 states that Hadrian was well versed in arithmetic, geometry,
painting, and every department of art. His capacities as a geometrician and an
artist are confirmed by other ancient authorities. 3 * From a passage in Dio Cassius,
where he relates that Hadrian sent his plans for the Temple of Venus and Rome
to Apollodorus, we learn that he made his own designs for his buildings. There
is therefore no room for doubt as to his architectural endowments.

Among the structures, wholly or partially preserved, which he created and
erected, the most important is the Pantheon at Rome, proved by recent investigations
to be the result of a rebuilding carried out between 120 and 124. The fact might
have been arrived at long ago if the formulas used by Spartianus " instauravit, fecit,
aedificavit, exaedificavit, extruxit " had been correctly interpreted. For the Pantheon
falls under the formula " instauravit," which is used by Spartianus in the sense of
" restoration " only in the case of the Forum of Augustus. " Romae instauravit

1 Lanciani, The Ruins and Excavations of Ancient Rotne.

- Soiptores Historiae Augustae (ed. Teubner) Aelii Spartiani de vila Hadriani.

3 Historiae Romanae Scrip/ores Sexti Aurelii Victoris Epitome,

4 Dionis Cassii Cocceiani Historia Romana (ed. Teubner).

Fie 48? Tivoli. Villa of Hadrian. Vestibule of the " Piazza d'Oro "
(125-135).

NORMANDY

101

If"

Fanthcum, Saepta, basilicam Neptuni, sacras aedes plurimas, forum August!,
lavacrum Agrippac." Whereas, in the case of the other pre-existing buildings which

we are told that he rebuilt, it has the meaning of
" building afresh " or " reconstruction." As a matter
of fact, rebuilding took place in the case of

the Pantheon, which had been finished by

Agrippa in 27 B.C. ;
the Saepta Julia, begun by Julius Caesar

and completed by Agrippa in the last

named year ;
the Basilica of Neptune, another work of

Agrippa of the year 26 B.C. ;
the Baths of Agrippa, erected by him in

19 B.C.

The other expressions, on the contrary, are used
for the structures of which Hadrian was the creator
as well as the builder. Such, for instance, as

his Mausoleum with the bridge leading to

it (136);
the Temple of Trajan in that Emperor's

Forum ;

his Villa at Tivoli (125-135) ;
the Basilica which he erected in honour of

Flotina near Nimes ;
the monument raised by him to the memory

of Fompey at Felusium.

Fig. 486. Rome. Pantheon. System
of relieving arches (120-124).
(From Heltrami, " 11 Pantheon")

The lofty dome of the Fantheon was an object of admiration from ancient times
onwards. Ammianus Marcellinus l bears
testimony to this in his account of the
notable things seen at Rome in 357 by the
Emperor Constantius (337-361). The in-
vestigation of its organic structure made
in 1 892 and 1 893 2 revealed the masterly
system of relieving arches which rendered
possible the construction and ensured the
stability of the wonderful pile (Figs. 486,
487). I will only remark that, so far as
my researches have gone, I have never
found in a dome or half-dome of earlier
date than the Fantheon any use of
relieving arches or even of mere ribs ; so
that the employment of a skeleton frame-
work in vaults of circular form is due to
Hadrian.

Fig. 4Sj. Kome. I'antheon (120-124).

Next to the Fantheon, in the new constructive and static features which it

1 Ammianus Marcellinus (ed. Teubner) Res gestae, xvi, IO.
- Ueltrami, // Pantheon.

102 LOMBARDIC ARCHITECTURE

presents, comes the Villa Tiburtina, which excited the wonder of Spartianus. Among
them we note :

I. The idea of cross vaulting springing from corbels, to be seen in the Great
Baths, for which we may refer to Fig. 120.

II. The half-dome in the Serapeum, with alternate concave and flat compartments.
There is no record of any earlier instance of this treatment. And we cannot
accept the suggestion that it is a reproduction of some original in the temple of
Serapis at Canopus. It is true that Hadrian derived his general idea from the
famous Egyptian sanctuary, but that does not imply that he copied it in every detail.
Spartianus, in fact, does not refer to any such slavish imitation. " Tiburtinam
villam mire exaedificavit, ita ut in ea et provinciarum et locorum celeberrima nomina
inscriberet, velut Lycium, Academian, Prytanium, Canopum, Picilem, Tempe vocaret.
Et, ut nihil praetermitteret, etiam inferos finxit."

III. The conical raccords to which I have called attention in connection with
the baptistery of Galliano.

IV. The vestibule of the " Piazza. d'Oro." It is an octagonal hall, with recesses
alternately round and rectangular, two of which form the entrance and exit. The
plan reproduces that of two rooms in the " Domus Augustana" on the Palatine. In
the interior, the re-entrant angles contained shafts (now gone, but the sockets remain)
supporting the semicircular arches which carried the compartments of the cupola.
The latter was pierced by a round opening in the centre. On the outside, correspond-
ing to the shafts within, are piers (about I ft. 4 in. x I ft.), from which spring arches
rising as high as the base of the cupola. These arches served four purposes. They
met the oblique thrust of the arches inside ; they strengthened the cupola at its base ;
they formed a facing to the walls, supplementing their very moderate thickness of
some I ft. 8 in., with considerable economy of materials and expense ; and they added
grace and elegance to the structure.

I am unable to point to any earlier vaulted building in which such sound prin-
ciples of making the forces of resistance depend on their distribution and not on
mass were illustrated. Nor can I refer to any earlier example of a dome composed
entirely of compartments.

On the model of the vestibule rose another work of Hadrian's (in all probability),
which would naturally follow the first experiment made at Tivoli, viz. the rotunda,
called by A16 Giovannoli 1 in a view published in 1619 the "Tempio di Siepe," which
stood near Hadrian's Temple of Neptune. This building was converted into Santo
Stefano del Trullo, and demolished in the time of Pope Alexander VII
(1655-1667). ^

Lastly, comes the Temple of Venus and Roma at Rome (121-135), which was
finished by Antoninus Pius, damaged by fire in 307, and restored by Maxentius
(307-312).

The singular plan of two sanctuaries with their apses back to back, strengthened
at their point of contact with massive buttresses, and the roof formed through-
out of barrel vaulting some 68 ft. in diameter, make up a whole of absolute
originality for that age, so far as temple architecture is concerned. And it is not
difficult to understand the unfavourable judgment passed on it by Apollodorus when
Hadrian, during the progress of the work, sent him the plans in order that he might
see that buildings on a vast scale could be erected at Rome without his assistance.
Some ill-feeling already existed between them. We may be sure that Hadrian had

1 Vedute degli antichi vestigj di Roma. * Armellini, Le chiese di Roma da! sec. IVal XIX.

NORMANDY 103

not forgotten or forgiven the insult he had received in the presence of Trajan when,
according to the story in Dio Cassius, the Syrian, unaware of the young man's genius,
had told him to "go away and paint pumpkins."

Moreover, the Greek architect brought up in the theory of elegant flat-roofed
architectural compositions, so far as basilicas and temples were concerned, of
which he had produced so notable an example in the Forum of Trajan at Rome
(" singularem sub omni caelo structuram " 1 ), must have thought the design conceived
and the constructive methods adopted by the Imperial architect at least inharmonious
and possibly extravagant. And this quite apart from the unfavourable criticisms
which Dio tells us he made on the building.

Here was more than a quarrel between two great architects : it was a battle
between two schools, the Roman and the Greek, placed as they were at opposite
poles. The one, the representative of a new people, serious, sturdy, practical ;
capable of conquering a world and impressing a unity upon it : the other, the
expression of an old race, restless, unstable, but penetrated with the sense of
proportion and beauty. It was the clash of two architectural styles. The one
new, finding its chief expression in Baths and Palaces ; based essentially on
vaulting and its combination, stability, and equilibrium ; deriving its source of
vitality from its own nature and not from the liberal aid of the minor arts ; capable
of inexhaustible development leading to the production of new styles. The
other, mainly an architecture of temples, carried by the Greeks to a superlative
degree of beauty, but by this time fossilized and incapable of giving birth to
new treatments.

The buildings raised by Hadrian which we have examined are the main
exponents of vaulted construction, carried out scientifically, in the Roman Empire
during the Ilnd century. And it was to the impulse given by the Emperor-architect
to the building art ("in omnibus pene urbibus aliquid aedificavit " 2 ) that a feature
of capital importance for vaulted architecture was due. For it was in the time
of Hadrian (who has been described as the only man of genius among the Roman
Emperors 3 ) that the ribs which had previously been used by the Romans in
arches and barrel vaults, were first applied to cross vaulting. And this is not
all. These diagonal ribs, very interesting examples of which are provided by
the substructions of the Palace of Septimius Severus (203) and the Imperial
Pulvinar looking on to the Circus Maximus at Rome, with double or single
" chains " of brick, while the Baths of Diocletian present striking instances with
two chains and compartments in brickwork filled in with rubble, were also used
in a new way by making them stand out on the intrados of the vault, as may
be seen in the " Sette Bassi " Villa near Rome. And this form was the precursor of
the principle of vaulting in the Lombardic and Pointed styles.

Our account of the Norman churches must be supplemented by the description
of another which, though not in Normandy but close to the borders of southern
Burgundy, nevertheless belongs to the Lombardo-Norman style, and was the work
of the same monastic order which produced the founder of that style, William
of Volpiano. This building, as being the most complete expression of the science
of construction as practised by the Cluniac monks and French builders at the

1 Ammianus Afarcellinus, xvi. " Scriptores Historiac Augustae I'ita Hadriani,

* Stuart Jones, The Roman Empire.

104

LOMBARDIC ARCHITECTURE

close of the Xlth and the opening of the following century, furnishes a sure
guide for estimating the real conditions of French ecclesiastical architecture at
that time, and also for testing the much-vaunted influence of the School of Cluny
and of French art generally on the evolution and perfection of the vaulted Lombardic
basilica.

THE ABBEY CHURCH OF CLUNY. The original church of the abbey founded
in 910 l by William the Pious, Duke of Aquitaine (887-918), and the abbot Berno

(910-927), was a mere chapel. When this
became too small for the ever-growing com-
munity, a larger structure (still preserving the
old one) was begun under Abbot Aymar
(942-965), finished by Majolus (948-994), and
dedicated in 982. This was known in later
times as " Saint Pierre le Vieux," and re-
mained standing, though it had suffered a
good deal, till the days of the abbey's destruc-
tion. This church, in addition to its central
tower, had two flanking the west front, an
idea which Majolus may have derived from
the bell-towers of the northern facade of St.
John Lateran at Rome, at least as old as the
time of John XIII (965-972).

But even this in course of time no longer
sufficed for the great concourse of monks for
whom room had to be found in the choir, nor
was it consonant with the splendour of the
monastery. It was under these circumstances
that Abbot Hugo (1049-1109) took in hand
(1089) the erection of the colossal edifice which
forms our subject. The monk Hezzel, formerly
a canon of Liege, is said to have the credit
of being the author of the designs. In 1095
Pope Urban II (1088-1099) consecrated the
high altar, which stood a little east of the
second crossing a fact throwing light on
the progress of the works, which, according to
Marrier, took twenty years to finish. Really
they must have taken longer, seeing that the
dedication by Innocent II (1130-1143) only
took place in 1 1 30, though one of his pre-
decessors, Callixtus II (i 1 19-1 124), had been
at Cluny in i I2O. 2 In front of Hugo's church
there was added in 1220, under Abbot Roland I, a narthex planned like a church
with nave and aisles.

Of the building, the largest monastic church in Christendom, and once thought
fit to be a " deambulatorium angelorum," as well as of Roland's addition, the
whirlwind of the Revolution in the XVIIIth century swept away like dust so large

1 Bruel, Recrteil des chartes de fabbaye de Cluny. - Jaffe, Regesta pontificum Romanorum.

Fig. 488. Cluny. Abbey Church. South arm
of main transept (1089-1130).

NORMANDY

105

a part that in 1811 all that remained was the portal of the west front of the narthex
with its rose window flanked by two square towers, some arches of the nave vaulting,
three bell-towers, the columns of the choir, the apse in an almost perfect state, and,
finally, some of the chapels which had been added later to the aisles. 12 ** 1

At the present day Hugo's church is represented by so much of the south arm
of the main transept as projects beyond the line of the outer aisle, with its octagonal
tower and staircase tower ; the angle where the outer aisle wall meets the south arm
of the lesser transept, and some remains of the latter. The main transept (Fig. 488)
is decorated with arcading, and
has a barrel vault. From its
eastern face one of the two
original apses still projects. The
octagonal tower belonging to
it rests on hood-shaped pen-
dentives. The capitals are
Corinthianesque with conven-
tional foliage, and, in some
cases, roses and monsters. These
figures disprove the arbitrary
assertion that the monster figures
of the Lombard ic School were
excluded by the School of Cluny.
The capitals are designed and
carved with freedom, and are
boldly and sometimes com-
pletely undercut. One in par-
ticular, with foliage and rampant
monsters, is admirable.

On the exterior (Fig. 489)
the tower is enriched with
arcading and arched corbel
courses. The surviving apse,
on the other hand, has shafts
round its circumference, while
the side walls of the transept
have lesenas in their upper part.
A large staircase tower projects
from its south-west angle.

With these data, with the illustrations and only too brief description of
Mabillon (Fig. 490), and with the information given by Lorain, 7 we may form
a clear idea of Hugo's basilica. It consisted of a nave and choir with double aisles,
regularly orientated, and had the form of an archiepiscopal cross, i.e. it possessed
two transepts, one larger than the other. The length from the west front to the
end of the apse was some 443 ft, and the breadth about 131 ft. The nave was
about 32 ft. wide between the piers, the inner aisles over 19 ft, and the outer about

1 Marrier, Bibliothcca ClnniacensisVita sanctissimi falris Hugoiiis ablntis Ciuniaccnsis.

3 Mabillon, Annales On/. S. Benedicti. * Lorain, Histoire dc fabbayc dc Cluny.

* 1'ignot, Histoire de tOrdre de Cluny. > Sacktir, >L' Cluniacenser.

" Annales Ord. S. Bencdicli. 1 Op. (it.

Fig. 489. Cluny. Abbey Church. South ami of main
iranst-pt (1089-1130).

io6

LOMBARDIC ARCHITECTURE

13 ft. They were separated by compound piers with a uniform diameter of
about 8 ft. The nave, which had pointed arches, was covered by a barrel vault
rising to a height of over 98 ft. above the floor of the church. Over each of the
longitudinal arches were two tiers of round-headed arcading, three arches to a bay.
According to Lorain, the lower, which contained windows, had piers, and the upper,
which was blind, had shafts. The inner aisles had also barrel vaults nearly
60 ft. high, and pointed arches. The same was the case in the outer aisles, which
were over 39 ft. high. Vaulting of the same kind (though Pignot l erroneously talks
of cross vaulting) was continued in both transepts, of which the larger measured
over 217 ft. by about 32 ft, while the smaller was above 108 ft. long, and somewhat
wider than the other. The tiling of the roof rested directly on the extrados of the
vaulting without any intervening timber work. The presbytery, with an apsidal
termination, was supported by lofty ancient columns brought from Italy, with pointed
arches springing from them, and was covered by a barrel vault and half-dome.

Fig. 490. Cluny. Abbey Church (1089-1130). (From the " Aimales Ordinis S. Benedicti.")

The pointed arch was used for all the structural parts of the building, while the
round arch was confined to the doors, windows (splayed on both sides), and
internal arcading.

From the ambulatory surrounding the presbytery and forming a prolongation of
the inner aisles opened five radiating chapels, each vaulted with an elongated half-
dome. Three towers rose on the line of the main transept : a square one exactly over
the crossing, and two octagonal ones above the arms. The latter, to judge from the
one that remains, formed at their base an octagonal cupola resting on hood-shaped
pendentives. In the case of the square tower the dome was probably circular,
and at a height of about 118 ft. above the floor of the church. A single tower rose
above the second crossing which formed the upper bar of the cross. All these towers
had pyramidal roofs covered with slating.

The exterior of the walls of the nave, lesser transept, and part of the larger one,
was broken by pilasters. Both nave and aisles were strengthened by substantial
vertical buttresses corresponding to the arcades within. Those of the nave and outer
aisles were further supported by flying buttresses, added, it is thought, in the
XHIth century to increase the resistance to the pressure of the vaulting, over-
weighted by the tiled roof resting directly upon it, and consequently of weakened
stability.

Of the three towers of the principal transept, the surviving one has been described
above. The view in Mabillon shows that the corresponding tower over the north arm

1 Of. cit.

NORMANDY 107

had three stages of arcading, instead of only two like the southern, and that the central
tower also had two. The tower over the minor crossing was of one story only, and
undecorated.

The west front contained a Lombardic portal with four jamb shafts on either
side, surmounted by a range of arcading resting on pilasters, of which the central arch
was pierced and lighted a small chapel partly formed in the thickness of the wall,
while part of it projected into the nave. The arches on either side of it formed niches.
This range of blank arcading was the earliest appearance of this form of decoration on
a church front in France.

Having completed our description, let us see what new features are presented by
Hugo's basilica. I confess that I cannot discover a single one. The double transept
is not one ; here due to the necessity of providing room for a large number of monks
in the choir. As a matter of fact, the view of Angilbert's basilica of Saint Riquier in
Mabillon 1 makes it clear that the two great lantern towers, flanked in either case by a
turret staircase, rose above the crossing of two transepts, which were obviously of the
same date as the central towers, as both show the same motive of " oculi," these being
blank in the towers and open in the transepts, and also the same type of roof.
Again, the church of Saint Remy at Rheims, rebuilt by Archbishop Turpin (756-802)
and finished by Hincmar (845-882), who dedicated it, according to Flodoard, in 852,
had two transepts, one of which included a tower, and a large two-storied tower at the
west front, as we learn from an illustration of a bas-relief on the tomb of the said
Hincmar (reproduced in Marlot's " Memoires," published in 1895 by the Academy of
Rheims) which appears to represent the dedication, and contains a figure of the
Emperor Charles the Bald (843-877) holding a model of the church. Lastly, St.
Michael at Hildesheim, completed in 1033, possesses two transepts with, in each case,
a tower over the crossing.

Nor is the arrangement of chapels radiating from the arcaded ambulatory round
the apse original, for its prototype is to be found in the destroyed basilica of Saint
Martin at Tours, the most famous sanctuary in the whole of Gaul, though not the
church erected by Bishop Perpetuus (460-490) to replace the earlier one built over the
tomb of St. Martin, first bishop of Tours (371-397), by St. Britius (396-443), and
dedicated about the year 470 the subject of the eulogies of Sidonius Apollinaris, 2
and of a brief description by Gregory of Tours. 3 For though the testimony of the
Frankish historian shows that the " absida tumuli," or " absida corporis," or " absida
sepulchri " in the church was surrounded by an atrium " hoc in atrio quod absidam
corporis ambit " * it does not follow that the atrium itself was surrounded by radiating
apsidal chapels, as has been imagined. 5 Perhaps this " absida " was something like
the " absida lignea " 6 over the tomb of St. Benignus at Dijon, but constructed in
masonry, seeing that, during the fire of 796, Alcuin was able to prostrate himself on
the pavement in front of the saint's tomb and pray that the flames might be stayed,
without receiving any harm. 7 Unless, indeed, the "absida" were an arcaded apse
enclosed by an ambulatory, after the fashion of that in San Sebastiano outside the
walls of Rome (367-384). In any case, only one apse projected from the body of the

1 Ada Sanct. Ord. S. Benedicti Vila S. Angilberti abbatis Centulensis.
a Man. Germ. hist. Sidonius Apollinaris, Carmina.

3 Man. Germ. hist. Gregorius episcopus Turonensis, Historia Francorum.

4 Mipnc, Pair. Lat., Vol. 71. S. Gcorgii Fhrentii Gregorii de miraculis S. Martini episcopi.
' C. Chevalier, Les foiiillts de Saint-Martin de Tours.

8 D'Achery, op. cit. Chronita S. Bcnigtii Divionensis.
7 Man. Germ, hist. Vita Alcuini.

io8 LOMBARDIC ARCHITECTURE

basilica of St. Martin, and that formed the chancel. This alone is mentioned by
Gregory of Tours, and its existence is implied in the story of the robbers who broke
into the church and carried off a rich booty. 1

The church of Perpetuus, which suffered from fire in the time of Bishop
Euphronius (555-572) and was restored by him, was sacked by the Saracens in 732.
Scarcely had it been restored, when a new fire damaged it in the days of Alcuin, who
had been prefect of the palace of Charles the Great, and afterwards became abbot of
Tours. He gave it a new roof.

A coin of the time of Charles before his coronation as Emperor (768-800) 2
shows the front of the church as it existed in the second half of the Vlllth century
(Fig. 491). Later, the Emperor himself rebuilt it, but not entirely, as may be
inferred from another coin struck under Louis the Pious (814-840), showing the
same church front as that of the days when Charles the Great was still only a
king. On this occasion a lofty bell-tower was erected, 3 which some would make

Fig. 491. Coin of St. Martin of Tours (768-800). Fig. 493. Coin of St. Martin of Tours (893-929).

out to be identical (except for the topmost story) with the existing "Tour
Charlemagne" (Fig. 492) which stands near the present church of Saint Martin.
This, however, is due neither to the Emperor, nor to the treasurer Herve, as
others argue, but is rather one of the results of the work undertaken after the
fire of 1096. The architectural and artistic decoration of the two stages above
the basement exactly suits the last years of the Xlth and the first of the next
century. The highest stage in the Pointed style is the result of alterations in
the XHIth or XlVth centuries.

On several later occasions Saint Martin was set on fire and injured by the
Normans between 853 and 903, and consequently several times repaired, and between
904 and 918 radically restored or rather rebuilt, as was the monastery. A coin
struck in the time of Charles the Simple (893-929) shows quite a different front
from that of the Vlllth and IXth centuries, and implies a reconstruction (Fig.
493). The basilica of the time of Charles the Simple was devastated by a fire in
994, according to Maan 4 ; but Herve de Buzangais, treasurer of Saint Martin
from 1014 onwards, 5 at once took in hand its rebuilding, and the new structure
was dedicated by Bishop Hugo (ioo7-about 1023) in 1014, or, according to Ademar,
I020. 6 Mosnier 7 states that Herv6 rebuilt only that part of the basilica which
covered the saint's tomb, i.e. the east end. Maan, on the contrary, describes the
old walls being pulled down before the new church was built up. However this
may be, what is of importance for us to notice is that to Herve must be assigned the
oldest traces of the choir with its five radiating chapels, which came to light during
the recent erection of the new Saint Martin, when I was able to inspect them.

1 Man. Germ. hist. Gregoriits episcopus Turonensis, flistoria Francorum.

1 C. Chevalier, of. cit. 3 4 Maan, Sancta </ metrofolitana ecdesia Turonensis.

5 U. Chevalier, Repertoire des sources hisloriqtus du moven Age.

8 Men. Germ. hist.Ademanis, Historiae. ^ Historia S. Martini.

NORMANDY

109

Maan, again, can scarcely be trusted when he refers to an increase of length in
the new church. This choir was afterwards rebuilt, first after the fire of 1096,
then again after 1175, and finally after 1202. In Italy the plan does not occur
till 1032, in an elementary form, in San Flaviano at Montefiascone, and in its
full development in the cathedral of Aversa.

Nor was there anything original in the small apses which formed the termination
of the transepts. Rome and Roman Italy afford very
ancient examples of cruciform buildings with the arms
ending in a curved projection. An examination of Mon-
tano's l work will show several such.

To my mind, an impartial estimate of the church
of Cluny reveals nothing really notable except its im-
posing proportions and the immense vaulted space
covered by it immense, however, only in the sense of
length, for the nave was not so wide as that of Sant'
Ambrogio at Milan. But this does not detract from
the great achievement of Hezzel in raising at that date
a building of such size entirely in masonry. The carving
is also remarkable.

The church itself, then, shows that the influence
which it is supposed to have exercised on the develop-
ment and completion of the Lombardic organism is
imaginary. But there were directions in which it had
an influence, sometimes of considerable importance.

Thus, the free use of the conical pendentive, which
previously had barely made an insignificant appearance
at Jumieges (1040-1066), was the starting point for its
diffusion through the country, where in the district be-
tween the Loire and the Garonne it prevails. It is
precisely to the end of the Xlth century that the
cupola of Saint Etienne at Nevers (1097) with its pen-
dentives of this type belongs; while in the Xllth
century we get them in Notre Dame at Avignon,
Saint Philibert at Tournus, Notre Dame at Beaune, the
church at La Charite, Notre Dame du Port at Clermont
Saint Hilaire at Poitiers (which originally had a wooden
roof), &c.

The use of the pointed arch, though it was anomalous
and not systematic, gave rise to the Transitional or
Lombardo-Pointed style, and was thus the starting point
for the Pointed style proper.

Again, we may be sure that the lofty walls of the nave, with their two tiers
of arcading rising above arches of excessive height, formed a subject of stud}',
of reflection, of imitation, for the earliest architects of churches in the Pointed style
with their soaring naves, quasi-triforiums, and wall galleries.

As for the carving, to judge by the little that has been preserved, it is certainly,
in the case of some of the figure subjects, not a whit inferior to the best Italian work
of the same date. I refer especially to a number of capitals, eight superb ones

1 Op. dt.

Fig. 492. Tours. So-called " Tour
Charlemagne" (Xlth, Xllth,
and XHIth or XlVth Centuries).

1 1. 1

I.OMHAKDIC AKCHITI'XTUkK

among them, now collected in the Museum near to where Roland I's narthcx once
stood. They -are of Corinthian type with figures, the conventional foliage being

Fig. 494. Mudrna.

('rypt (1099-1106).

treated in high relief and sometimes completely undercut, and are executed in a very

easily worked stone.

On one of them appear Adam and Eve before the Fall, wearing the

characteristic cylindrical Jewish cap, also to be seen on the common head belonging

to the bodies of a pair of
winged centaurs carved on a
capital in the old part of
the crypt of the cathedral at
Modena (1099-1 106) (Fig. 494),
and represented on many of the
sculptured sarcophagi in the
Latcran Christian Museum.
The motive of two quadrupeds
with a single head was derived
from the Etruscans, who em-

49S > ~''''" renci '- Arcliiu-olocicul MUM-IIMI. Portion
Klniscuti curving from C orncto Tanjuiiiia.

ployed it not only on vases but
also on carved panels. One of
the compartments of a piece
of archaic Etruscan carving in

the Archaeological Museum at Florence contains a pair of lions with a single head (Fig.

495). The Romans also used it, as may be seen on a capital illustrated by Montano, 1

1 ('/. til.

NORMANDY

ii i

showing two winged animals united by a single head which Ukes the place of the
flower on the abacus.

On another capital, which seems to me the finest of all, David is represented
playing the harp (Fig. 496). The figure is treated with grace, in an easy and lifelike
manner, showing refinement and correct propor-
tions, especially in the extremities. The pose is

natural, and the

drapery intelli-
gently arranged.

Taken all in all

it is not inferior

to the figures of

the same date

executed by

Wiligelmus or

Gugliclmus for

the facade of the

cathedral of Mo-

dcna (Fig. 497).

In these, while we

can admire, for

instance in the

subjects from the

beginning of

Genesis, the grand style and majestic air of the figures, which sometimes (e.g. the
man killed by Lamech's arrow) have a natural look not found in the faces of the
figures at Cluny with their more forced expression, on the other hand one cannot help
noticing the general lack of proportion between the heads and bodies, and also in the

Fig. 496. Cluny. Muneum. Capital
fr<mi Abbey Church (1089-1130).

Fig. 498. Cluny. Muoeum. Capital
from Abbey Church (1089-1130).

l"''g. 497' Modena. Sculpture on the facade of the Duomo (Xllth Century).

feet, even when covered. Moreover, they have a rigidity from which the David of
Cluny is free. In a third capital, with God calling Adam after the Fall (Fig. 498),
the figures are inferior both in quality of line and modelling to the best of the work
at Modena.

112

LOMBARDIC ARCHITECTURE

The church of Cluny was not alone in its failure to show any contribution
towards the evolution and perfection of the Lombardic vaulted basilica ; for the
same thing is true of the rest of the French ecclesiastical buildings. And this
we shall have to substantiate by the evidence of some of the most important
dated churches of the country in the Xlth and early years of the following century,
though not in the Norman districts which we have hitherto dealt with. These
we shall briefly compare, taking note of their chronology and special features,
with the Lombardic buildings of Italy; thus obtaining tangible proof that all these
French churches were behind the times so far as that development is
concerned, when confronted with the creations of the Lombard gilds which showed
the way.

THE ABBEY CHURCH OF SAINT PHILIBERT AT TOURNUS is the result
of Abbot Bernerius's rebuilding between 1008 and 1019 (the latter being the year
of the consecration) of the previous church erected by Abbot Stephen between
960 and 980, and damaged, together with the adjoining convent, by a fire in 1006

in the days of Abbot Vago. This
Xth century church in its turn
replaced an earlier one. 1

Five bays of the church of
Bernerius remain. The choir with
its crypt, and the two-light open-
ings on the interior face of the
two-storied narthex, indicate with
their Corinthianesque and varied
Composite capitals the partial re-
building in the early years of
the Xllth century mentioned by
Robert, 2 and followed by the dedi-
cation performed by Calixtus II
(1119-1124). The original bays
are divided into nave and aisles
by very lofty cylindrical piers,
each of which carries above its
capital a vaulting shaft (Fig. 499).
From these shafts spring the un-
couth transverse arches which sup-
port the barrel vaults over the
nave. The aisles, which are
of great height, have roughly

Fig. 499Toumus. Saint Philibert (1008-1019).

constructed, ramping, unraised

cross vaulting. The side walls
are strengthened by external buttresses corresponding to the piers within
(Fig. 500).

The church of Tournus is important not only for its place in the history of the
development of vaulting, but also because it presents two features, one of which
is worthy of special notice.

The first and most important is that of the great cylindrical piers which separate

1 Meulien, Histoire de la ville et du Canton de Tourmts.

' 2 Gallia Christiana Tormtsimu.

NORMANDY

the nave from the aisles. This form influenced the builders of Sant' Abondio
at Como (1013-1095) and of Jumieges (1040-1066), and it was copied at Malvern
(1085), where it became the model which was reproduced in so many abbey and
priory churches of Great Britain.

The second is the arrangement of the vaulting shafts. It was used later by the
builders of Sant' Ambrogio at Milan on the piers, with the object of supporting
an arched corbel course to decorate
the nave. Afterwards the architects
of the cathedrals of Worms and
Mainz employed it in the form of
half-piers from which to spring the
pairs of wall-arches in the nave.

The chief value of Saint Phili-
bert consists in its system of con-
struction and equilibrium. As we
saw, the second church had to be
rebuilt after little more than a
quarter of a century ; and the
reason, as the injury caused by fire
suggests, was that it had a wooden
roof. It was accordingly decided
to rebuild it completely in masonry.
And it is this decision which demon-
strates the real merit of the Bur-
gundian builders.

Anxious to give their work a
character distinct from that of the
Lombardic style, which at that time
was being illustrated by Saint
Benigne at Dijon, then in course
of erection (1002-1018), and San

Babila at Milan, and desirous of lighting the nave directly, but afraid to spring a
longitudinal barrel vault from such elongated piers, they fell back on the inartistic
but more stable device of springing from the transverse arches of the nave as
many barrel vaults as there were bays, the nave walls being strengthened by
buttresses corresponding to the arches. This system, which is of very rare
occurrence, had been already followed in the original church of Saint Front at
PeYigueux, the erection of which was begun 1 by Bishop Froterius (988-991), who
was buried in it, 2 a proof that the work must have been then well advanced.
The dedication took place in 1047 during the episcopate of Geraldus, who died in
1059. The facts are that the parts of this church preserved in the later rebuilding
show that while the nave, which was barely 17 ft. wide, had a wooden roof, the
aisles were covered with ramping barrel vaults, one to each bay, parallel to one
another and at right angles to the nave, and resting on transverse arches springing
from isolated piers and half wall-piers. 3

It was undoubtedly from Saint Front that the architect of Saint Philibert
derived his idea when he adopted ramping vaulting in the aisles. Its function

Fig. 500. Tournus. Saint I'hilibert. North side
(1008-1019).

1 Sammarthanus, &c., Gallia Christiana.

1 Robert, Gallia Christiana.

3 Verneilh, L'archi/eiltire byzantinc en France.

VOL. II

u 4 LOMBARDIC ARCHITECTURE

in the latter case was to abut the transverse arches of the nave, while at the same
time the nave was not affected by its thrust. The result was a three-lobed
framework of resistance and of thrust (" trilobo di forza ") discharged on to the outer
walls, almost without need of intermediate supports.

The present church of Saint Front in the Byzantino- Venetian style, with its
Oriental derivation through St. Mark's, is a new erection of 1120, and not identical
with the fabric of 1047 as some even recent writers imagine. 1 The church of
Froterius was burned, together with the whole of the wooden-roofed convent,
in the terrible fire of 1 120, when even the bells were melted. 23 The only parts which
escaped were the vaulted aisles.

And all said and done, San Babila at 'Milan, but little older, already possessed
aisles with beautiful raised cross vaulting springing from compound piers, while
the nave (though without windows) had a barrel vault, and the thrust of the ribs
which interrupted it and supported the other vaulting, was met by an original
and scientific system of buttressing.

A great deal of water will have to run down the rivers of France before
the French builders succeed in covering directly lighted naves with longitudinal
barrel vaulting. And when this does happen, the hour will have almost struck
for them to abandon a foreign style, and devote all their energies to the creation,
under the guidance of the ruling principles of the Lombardic basilica, and with
the assistance of the improvements in vaulting introduced in the nave of Durham
Cathedral, of a new national architecture, viz. the Transitional Lombardo-Pointed,
out of which the Pointed style pure and simple was developed.

THE CHURCH OF SAINT HILAIRE AT POITIERS was rebuilt after the disastrous
fire which destroyed most of the churches of the city in 1018, and dedicated
in io59- 4667 Originally it consisted of a nave with double aisles, the former
separated from the latter by compound piers and columns alternately, and had
a wooden roof. The aisles were separated by columns, and had cross vaulting.
The interior face of the side walls presents a cluster of three shafts corresponding
to the nave piers, from which spring transverse arches, and a shaft answering to the
columns of the aisles and nave, to support the cross vaulting. On the outside these
wall-supports are met by buttresses alternately larger and smaller.

When, in II3O, 8 it was decided to vault the nave, owing to the fact that it was
some 46 ft. wide, the plan was adopted of lining its interior with arcades resting
on clustered piers connected by transverse arches at the level of the clerestory, which
was strengthened by an additional facing, and then raising over each bay
of this inner structure an octagonal cupola resting on hood-shaped pendentives
(Fig. 501).

It is obvious that the only Lombardic features presented by Saint Hilaire
are the alternate large and small supports, and the form of the piers, the capitals
of which, however, are neither continuous nor Lombardic. The figure capitals
(where old) must be ascribed to the Xllth century. The original ones were of two
kinds : those with plain crocket leaves, and those with similar foliage above a
continuous crown of leaves.

1 Peyre, Histoire ginirale des beaux-arts. ~ Delisle, Ex chronico S. Maxentii.

3 Id., Exfragmento de Petragoricensibus episcopis. 4 Delisle, Ex chronico Ademari Cabanensis.

6 Id., Ex fragmento historiae Monasterii Novi Pictavensis. e Id., Ex chronico S. Maxentii.

~' Labbe, op. cit. Chronicon S. Maxentii. 8 Enlart, op. cit.

NORMANDY 115

THE ABBEY CHURCH OF SAINT SAVIN was erected in the time of Duke
William III (980-1030) and Abbot Gumbald (f 1040).' It was designed with nave
and aisles separated by cylindrical piers, and the aisles had cross vaulting (which,
being the result of the intersection of two perfectly semicircular barrel vaults, is
without transverse arches) springing from the nave piers and the corresponding half
wall-piers, the latter being strengthened externally by powerful buttresses. The
nave, on the other hand, is covered with a continuous barrel vault rising from very
low nave walls, and therefore without windows, the thrust of which is met by the
aisle vaulting. The
whole is covered by
a single gabled roof.

The only Lom-
bardic feature in this
church is the familiar
idea, derived from the
Roman tradition, of
strong external but-
tresses to resist the
thrust of the cross
vaulting. Whereas,
in the contemporary
San Flaviano at
Montefiascone we
already find used in
the aisles the essen-
tial element of Lom-
bardic architecture
diagonal ribs applied
to the cross vaulting.

THE CHURCH
OF SAINT-REMY AT
RHEIMS was rebuilt
by Abbot Theoderic
(1036-1048) in the
fifth year from his
appointment, after
demolishing, owing to Fig. 501. Poitiers. Saint Hilaire (Xlth and Xllth Centuries).

its excessive size, the

portion of the structure begun by his predecessor Airardus (1005) to replace the
older church which had been completed by Hincmar (845-882) in 852. This
demolition spared nothing but some of the foundations : " quam basilicam
Theodericus, destructa ob nimiam aedificii molem ea quae ab Airardo incoepta
fuerat, aedificare aggressus, Herimaro successori perficiendam reliquit" 2 It was
finished by his successor Herimar, who completed the north arm of the transept
and erected the southern one. He also constructed the roof. The consecra-
tion was performed by Pope Leo IX in 1049. The church was enlarged by

1 Sammarthanus, &.C., Gallia Christiana Abbatia S. Savini ad IVartinfani.
" Mabillon, Annales Ord. S. Benedict!.

I 2

LOMBARDIC ARCHITECTURE

Abbot Pierre de Celles (1162-1182), who replaced the narthex by two bays. He
also remodelled the west front, raised the side walls of the nave, and threw across
them, as also in the transept, Pointed vaulting, which involved an alteration in the
nave piers. Lastly, he rebuilt the choir. 1 2 3 *

Of the original church there survives the nave, the transept, and the remains
of the choir. The nave and aisles are separated by massive pillars formed of

Fig. 502. Rheims. Saint Remy (Xlth and Xllth Centuries).

a number of shafts clustered within a circle, and two compound piers, preceded
by two bays in the Pointed style (Fig. 502). Everything indicates, as has been
suggested, that the piers were at first plain and cylindrical instead of being encircled
by clustered shafts as now. This would explain the difference between the rude
capitals with triangular indentations (Fig. 503) and those with foliage and figures,
which would be the result of alterations between 1049 and 1162, when also the

1 Marlot, Metropolis Remensis historia.

3 Id. , Ex historia dedicationis ecdesiae S. Remigii.

- Delisle, Ex chronico S. Benigni Divionensis.
4 Cosset, La basilique de Saint-Remy ft Reims.

NORMANDY

117

single-arch openings of the triforium springing from rectangular piers were divided
into two by shafts with foliage capitals evidently of another date, though earlier than
the capitals of the time of Pierre de Celles.

The aisles (Fig. 504), which were after-
wards cross vaulted and bear traces of altera-
tion on two occasions, originally had wooden
roofs like the triforium. The vaulting we see
in the nave is not that of Abbot Pierre, but a
sham vault put up in the last century, when
at the same time some bays of the transept
vaulting were rebuilt, as well as a considerable
part of that in the aisles and triforium.

The transept also has aisles, with galleries
above. The piers were originally either cylin-
drical, or quadrangular with attached mem-
bers : in the one case a half-column and a
pilaster, in the other a simple pilaster. The
ground floor had barrel vaulting interrupted by transverse arches springing from
cylindrical or quadrangular piers. The galleries and central space had timbered
roofs.

The aisles terminated at the east in small apses (partly destroyed) flanking the
rectangular sanctuary.

Externally, the south side of the nave (the only one that is free) is strengthened
by semi-cylindrical buttresses, which were altered when the nave walls were raised

Fig. 503. Rheims. Saint Remy. Capitals
in the nave (Xlth Century).

Fig. 505. Rheims. Saint Remy. South side (Xlth and Xllth Centuries).

for the construction of the vaulting. At the same time the range of round windows
was formed at the top (Fig. 505).

The old Saint Remy exhibits no Lombardic features, displays an art which

is old-fashioned compared with that of the contemporary church at Cerisy la Foret

LOMBARDIC ARCHITECTURE

(1030-1066), and contains no ele-
ment of service to the Italian gilds
for the development of the Lom-
bardic vaulted basilica. We find,
however, one feature which is new
for France : the transept aisles and
galleries continuing those of the
nave. This arrangement must have
been suggested by St. Maria im
Capitol at Cologne, which, though
it too was consecrated in 1049, ap-
pears to have been rebuilt between
1024 and 1030.

THE MONASTIC CHURCH OF
SAINT SERNIN AT TOULOUSE was
rebuilt after the Cluniac monks in
1083 took the place of the canons,
and was consecrated in 1096 by
Pope Urban II. 1 Its plan is that
of a nave and choir with double
aisles, divided by piers of uniform
thickness with an engaged vaulting
shaft to carry the transverse arches
of the nave. The latter is barrel
vaulted and without windows. The
aisles have ordinary cross vaulting.

The transverse arches over the two outer ones started from piers and half wall-piers
strengthened by substantial buttresses outside. Above the inner aisles is a triforium
covered by a half barrel vault which meets the thrust of the nave vaulting. The
outer aisles have half barrel vaults springing from the floor level, forming dark
spaces. To each arch of the nave corresponds a double opening above, divided by
a shaft, and enclosed by a single arch. Round-headed windows light the outer aisles.

THE MONASTIC CHURCH OF SAINT ETIENNE AT NEVERS was rebuilt by
William I Count of Nevers (1040-1097), and given, with the convent, to Cluny in
\ogj.~ Its plan is that of a nave and choir separated from aisles by compound piers
of uniform size, consisting of a square pier with four engaged half-columns, corre-
sponding to which are wall-shafts and external buttresses. The aisles have ordinary
cross vaulting, but the nave is covered by a barrel vault divided into sections by
large transverse arches kept up by buttresses outside. The nave walls from which
it starts are carried up high enough to allow of windows, formed partly in the walls
and partly in the base of the vault. Over the aisles runs a triforium with a semi-
circular vault, which opens on the nave by pairs of arches, divided by a column, and
enclosed by another arch.

The church of Rivolta d'Adda had already been given to Pope Urban II (1088-
I0 99). ar >d Sant' Ambrogio at Milan was nearing completion, when Saint Sernin at

1 Jaffe, Regesta por.tificiim Romanontm.

2 Marrier, op, cit. Carta fitndationis sett dotatwnis Monasterii S. Stephani Nivernensis.

Fig. 504. Rheims. Saint Remy. South aisle ol nave
(Xlth and Xllth Centuries).

NORMANDY

119

Toulouse and Saint Etienne at Nevers were finished. Now, any one can see at the
first glance how different was the organic and constructive conception in the minds of
the architects of the two Lombardic churches from that which was before the authors
of the two French ones, and how far ahead the former were of the latter in constructive
and statical knowledge. This is abundantly proved by the rational system of
buttressing employed in the church of Rivolta, which has maintained the integrity
of the building in such a wonderful way through the centuries ; whereas the
cracks in Saint Etienne at Nevers are evidence of unsound construction. And
so we have reached the close of the Xlth century without having found a single
building of certain date in France which can be said to have in any respect shown the
way in the creation of the Lombardic style.

THE ABBEY CHURCH AT LA CHARITE appears to have been begun by the
prior Girardus, who was also its architect, in IO56, 12 or io6o., 3 and was dedicated in
I ID/. 4 It was erected with ordinary cross vaulting for the aisles, supported by internal
and external buttresses ; while the nave had a barrel vault, crossed by transverse
arches, and starting above the clerestory with its round-headed windows.

Fig. 506. Vizelay. Abbey Church of the Madeleine. Nave (Xlth and Xllth Centuries).

THE ABBEY CHURCH OF THE MADELEINE AT VEZELAY, erected by Abbot
Artald (1096-1106), was dedicated in 1104. It was damaged by a terrible fire
in II2O, 56 but repaired by Abbot Rainald of Semur (1106-1128). The narthex,
built and dedicated about II32, 78 must have been the work of his successor.

The body of the church, which belongs to the original work, is divided into nave
and aisles (Figs. 506, 507) by compound piers of uniform size, and has unraised

1 Delisle, Ex chronico A ndegavensi altero.

2 Id., Ex chronico Willelmi Godelli man. S. Martialis Lemovicensis.

3 Id., Ex chronico S. Maxentii. 4 Id., Sugerius abb. S. Dionysii Liber cU vita LuJovUi Grossi rcgis.
5 Delisle, Ex chronico Vizeliacensi. 6 Id. , Ex chronologia Koberti moil. S. Mariani Altissiodorensis.

7 Petit, Descriptions des miles et campagnes dn department de F Yonne.

8 Anthyme Saint-Paul, yiollel-le-Diic, ses travaux fart et son systeme ttrtUohgiqtu.

I2O

LOMBARDIC ARCHITECTURE

cross vaulting. The bases of the piers are in some cases ornamented with foliage,
ovolos, cauliculi, fluting, animals, scroll work, &c. The Corinthianesque capitals

Fig. 507. Vezelay. Abbey Church of the Madeleine. South aisle (Xlth and Xllth Centuries).

display, besides foliage, real or imaginary animals of every description, demons, human
beings in repose or conflict, and scenes from sacred story. These carvings, evidently

Fig. 508. Vezelay. Abbey Church. Portal (Xllth Century).

the work of the same school which produced those at Cluny, with complete
undercutting in some cases, while they will not bear comparison with those of the

NORMANDY

121

cathedral of Modena (1099-1106), are superior to those at Rivolta d'Adda and
in Sant' Ambrogio at Milan, though we must of course remember the rather easily
worked stone used for the church of Vezelay.

The representations of monsters and of hunting and fighting scenes, which run
riot on the capitals, show how entirely without foundation is the assertion that the
School of Cluny avoided such subjects. On a capital of the first pier to the right,
which being near the west end must date from about 1104, appears a centaur with
the characteristic cylindrical cap which we noticed in the Museum at Cluny and
in the crypt of the cathedral of
Modena.

On the outside, the walls had
simple pilasters, converted later into
the present massive buttresses.

The imposing portal of the nave,
divided after the French fashion by
a central pier to relieve the weight
of the lintel (Fig. 508), shows a
marked advance artistically over
the carvings inside the church, and
is to be referred to the beginning of
the works on the narthex. For the
fineness of its sculpture its only
rivals in Italy are the contemporary
portal of the cathedral at Ferrara
(Fig. 509), dated by the inscription
on the innermost arch in 1135, and
that of the church of San Zeno
Maggiore at Verona, the rebuilding
of which was finished in H38, 1 both
works of Master Nicholas. The front
of San Zeno was afterwards pierced
by a large rose window (the middle
cornice being cut through for the
purpose) symbolizing the wheel of
Fortune, the carvings of which are certainly to be referred to the works carried out
before 1178. This rose window is one of the earliest to be found in Italy, the birth-
place of this type of opening. Santa Maria del Vescovado at Assisi furnishes
an example of 1163. It was only in 1220 that the church of Cluny received one in
the west front of Abbot Roland's narthex.

The two-storied narthex at Vezelay is constructed with arches some of which
are round and others pointed, Lombardic piers, and ordinary cross vaulting
constructed in rubble, either groined or else ribbed. The narthex has a portal of its
own, almost entirely re-worked, the outer archivolt of which rests on the back
of a lion with a monster between its paws on one side, and a bull surrounded
by foliage on the other. It is the oldest example of an arch supported by
half-figures (" protome "), after the Etruscan (as we shall see when we come
to Deerhurst) and Lombardic fashion, that France can show.

Some of the. capitals in the church and narthex have a suggestion of the Pointed

1 Cipolla, Per la storia if Italia e tie' suoi conqti is/atari nel media evo piit anlico, ricercHe varie.

Fig. 509. Ferrara. Portal ot Duomo (1135).

122

LOMBARDIC ARCHITECTURE

style in the play of the foliage. Similar capitals, still more advanced in character,
are found in the cathedral at Modena (a fact to which attention has hitherto not been
called), which, with the slender bell and the play of the foliage, anticipate or may

Fig. 510. Modena. Duomo. Capital in
the crypt (1099-1106).

Fig. 511. Modena. Uuomo. Capital in
the facade (1099-1106).

style

even be said to be treated in the manner characteristic of the Pointed
(Figs. 510, 511, 512).

It cannot be said that the first attempt made by the School of Cluny to cover
an entire structure of considerable size with cross vaulting achieved a perfectly
happy result, for not only did the vaulting of the nave develop cracks and have
to be held in by tie rods, the attachments of which are still to be seen, but it became
necessary, after the lapse of little more than a hundred years, to secure the stability
of the entire structure by the external buttressing to which we have already called
attention. While, on the other hand, the church of Rivolta d'Adda, with its ribbed
raised cross vaulting, concave at the crown, lighted by windows in the lunette
wall spaces, and with its ramping buttresses pierced by arches, still stands to bear
. __ witness to its inherent soundness.

Some have wanted to make out that the idea of
covering the nave of Vezelay with cross vaulting
came from Palestine. On the contrary, it was the
natural solution of a problem which was always
stimulating the energies and efforts of the School of
Cluny, eager as it was to pursue an independent
course, and anxious to carry out, by ways and means
distinct from those laid down by the great Lombardic
School, the revolution in the principles of construc-
tion and equilibrium which the North Italian gilds
had brought about in the Latin basilica. The fact
rather is that it was the monks of Cluny who carried
the idea to Palestine. New styles of architecture
are not produced by magic. Now, neither Palestine
nor Syria afford, after the Roman period, a single dated example of a basilica
completely covered by cross vaulting in which the cells are constructed in coursed
and dressed masonry, and the piers are of the Lombardic type, as for instance in
St. Anne's at Jerusalem before their introduction by the Franks after their capture of
the city, for there can be no question that they were the builders (with the assistance

Fig. 512. Modena. Duomo.

in the facade (1099-1106).

Capital

NORMANDY 123

perhaps of local workmen) of the churches there which have roofs of that kind.
We can thus explain the appearance in Palestine of such structures of perfected
type, without having to look there for earlier ones illustrating the necessary gradual
development.

The abbey church of V^zelay is a proof that the School of Cluny was still
in the first stages of the solution of the problem how to cover churches of large size

F 'g- 513. Angers. Cathedral (Xllth Century).

with a complete system of cross vaulting, at a time when the church of Rivolta
d'Adda was ready to serve as the introduction to the first chapter of pointed-arch
construction. It was only at the dawn of the Pointed style that this much-vaunted
School resolutely entered upon the difficult path leading to the proper arrangement of
simple or ribbed cross vaulting ; and that period, so far as the northern countries
are concerned, did not begin in France, but rather in England. France may lay
claim to the imperishable honour of having seen how to unite the Lombardic
organism to that fusion of the pointed arch with ribbed cross vaulting which was
achieved in Durham Cathedral between 1129 and 1133, perfecting and trans-

I2 4

LOMBARDIC ARCHITECTURE

forming it in a transitional system which later found its perfect balance and entire
harmony in the Pointed style.

Of this Transitional style I will, in conclusion, indicate a few typical
examples.

(1) The western towers of the cathedral of Chartres as rebuilt after the fire
of 1 1 34. The northern tower was begun in the same year, and it is known
that about 1 144 the works on the southern one were in progress. 1 The two cross
vaults of the ground floor are ribbed, the ribs having a moulding of three rolls.

(2) The abbey church of Saint Denis, founded by King Dagobert about the year
630, refounded by Pippin and Charles the Great, and erected by Abbot Suger
(1122-1152), who was also the architect (" instauravit ecclesiam eo schemate").
The narthex, which was the first work undertaken, and the choir were finished
in 1 140 and 1 144 respectively, while the nave was built later than 1 144. 2 S4 5

(3) The cathedral of Sens, begun about 1140.

(4) The cathedral of Angers (Fig. 513). The vaulting over the aisleless nave,
which is nearly 50 ft. wide, was constructed by the bishop Normand de Doue
(1150-1153), under whom we find a payment made for this purpose. 7

1 Clerval, op. cit. 2 Robert, Gattia Christiana.

8 Musier, Hisloire de Suger abbi de Saint-Denis.

4 Migne, Pair. Lat,, Vol. 186. Sugerius abb. S. Dionysii, Libettus ae amsecratione ecclesiae a se aedificatae.

e Id., Liber de rebus in administratione sua gestis,

6 Lefevre-Pontalis, L 'architecture religietise de Fancien diocese de Soissons.

7 Tresvaux, Histoire de feglise et du diocese a" Angers.

CHAPTER III

ECCLESIASTICAL ARCHITECTURE IN ENGLAND FROM CON-
. STANTINE TO THE NORMAN CONQUEST

is little to be said about the ecclesiastical architecture of Britain
in the period between Constantine's Edict of Milan (313) giving peace
to the Christians, and the withdrawal of the Roman legions from the
island (411) by the Emperor Honorius's rescript of 410, and from this to

the invitation given to the Anglo-Saxons (449). The only vestige that has survived

is the remains, barely rising above the ground, of the

small basilica which in old days stood near the Forum

of the Roman town of Calleva Atrebatum (Silchester).

It followed the early Latin plan of a cross without the

upper limb and with a very short transept, like the old

St. Peter's at Rome. The nave and aisles were reached

through a narthex, and the former terminated in a semi-
circular apse, while the latter were closed by two sacristies

at the upper end.

The basilica at Silchester, the plan (Fig. 514) of

which unquestionably indicates the Christian purpose for

which it was erected, has been dated between 313 and

41 1, 1 and I think rightly, particularly on the ground of

the orientation, for it was only after the erection of the

Basilica Ursiana at Ravenna (370-384) that the new

Ravennate arrangement of the apse at the east end was

generally adopted. On the other hand, Britain was rapidly Christianized in the

course of the IVth century. 2

The church of Silchester affords a very early specimen of an apse flanked

by two " secretarial In Italy it is illustrated by three examples, also of early date,

two of which have been already mentioned : the Basilica Pammachiana at Porto near

Rome (about 398), and the large Basilica of St. Symphorosa on the Via Tiburtina,

thought to be not later than about the Vth century. The third is San Salvatore

at Spoleto, which belongs to the IVth century.

* * *

We are able to say even less about church building in the years following
the calling in of the barbarians down to the arrival of the monk Augustine and his

1 Archaeohgia, Vol. LIII. Fox and St. John Hope, Excavations on the Site of the Roman City at Silchester,
Hauls, in 1892.

2 Harnack, The Mission and Expansion of Christianity in the First Three Centuries.

125

Fig. 514. Silchester. Plan of
Christian Basilica (IVth or
Vtli Century).

126

LOMBARDIC ARCHITECTURE

companions (597). The feeble ray of light (one might almost say moonlight)
shed by the remains of the little suburban church of St. Martin at Canterbury
is insufficient to dissipate the settled darkness which envelops the conditions of the
art of building during the period when the unhappy country was being trampled
under foot by its new masters. Those remains are in all probability to be referred to
the work of Queen Bertha, 1 and this explains the dedication of the church to the
wonder-working saint of Tours.

When Bertha, daughter of Caribert I (561-567), was married to Ethelbert King
of Kent (560-6 1 6), 2 a condition was made, according to Bede, 3 that she should
be accompanied by Liudhard so that she might keep her Christian Faith. Hence,
it is natural that as Ethelbert had a sanctuary for his pagan worship, his wife should

Fig. 515. Canterbury. Old Church of St. Martin. South side (Vlth Century).

demand another for her Christian service ; and so St. Martin's was erected on the
spot where she was wont to offer her prayers. This supposition would be confirmed
by the eastern orientation of the structure, a fact which is fatal to Bede's assertion
that it was built in the days of the Roman occupation. Whatever its form may have
been, the chancel was certainly at the east end. It has been thought 4 that it was
semicircular, on account of traces of buttresses belonging to the original building
discovered at the south-east angle ; but the evidence is insufficient, though it is quite
true that everything is in favour of an apsidal end. Moreover, the use of Roman
bricks in the construction suggests a date contemporary or nearly so with St.
Pancras at Canterbury, built by Augustine. 5 Unless the traces of a junction with
the south wall of the church, near the second door, and the remains of a pavement

1 The Archaeological Journal, December, 1906. Micklethwaithe, Something about Saxon church building.

2 Rolls Series The Anglo-Saxon Chronicle. 3 Historia ecclesiastica gentis Anglorum.

4 Routledge, The Church of St. Martin, Canterbury Appendix C.

5 Archaeologia Cantiana, Vol. XXV. St. John Hope, Excavations at St. Austin's Abbey, Canterbury The
Chafe! of St. Pancras.

ANGLO-SAXON ARCHITECTURE

127

in front of the latter, belong to a porch, 1
in which case the church itself would
date from the time of Augustine, and
belong to the type which has been con-
veniently named " Augustinian."

Of the primitive church there re-
mains the rectangular nave (Fig. 515),
which has been tampered with and
originally extended further west. It
forms the chancel of the present Anglo-
Saxon church, probably built after a
bishop was established there by the
primate Theodore (668-690), as we are
told in " The Black Book of the Arch-
deacons of Canterbury." This episcopal
see is also mentioned by Dugdale. 2

St. Martin's contains a font which
may be supposed to have been originally
a well-head, and belongs to the Xllth
century. The decorative motives on it
are derived from the simple arcading
found as an ornament on other well-
heads, e.g. the one in the cloister of St.

John Lateran at Rome (Fig. 516), the execution of which indicates the same date
as that of the puteal in the ancient atrium of San Giovanni a Porta Latina, in
other words, the Xth century.

Fig. 516. Rome. Well-head in cloister of St. John
Lateran (Xth Century).

We are more fortunate when we come to the years after 597, and the
constructive period of the time of Augustine and his companions, for here we have
important descriptions and remains of building.

THE CATHEDRAL OF CANTERBURY. We know from Bede 3 that the monk
Augustine, afterwards first Archbishop of Canterbury (603-605), recovered, with the
help of Ethelbert, a church said to have been built on the site by Roman Christians,
and dedicated it to the Saviour. Augustine's work, however, cannot have been
confined to a consecration, but must have taken the form of rebuilding, for the altar
was at the east end, an arrangement which shows that the church cannot have been
erected before the Anglo-Saxon conquest. And we cannot suppose that in the years
between that conquest and 597 the Christians would have erected, under Pagan rule,
a church of such importance, which was a reproduction, up to a certain point,
of St. Peter's at Rome. As a matter of fact, Mabillon 4 mentions the rebuilding
of the previous church, and places it in 602. Goscelin 5 also confirms the fact
of a construction by Augustine ; and a short description of it is given by Eadmer."
According to this it consisted of a nave and choir with aisles. The apsidal chancel

1 The Archaeological Journal, Vol. LVIIL Peers, On Saxon churches of the St. Pancras type.

2 Monasiicon Anglicanum. 3 Hist. eccl. * Annales Ord. S. Benedieti.

5 Mabillon, Ada Sanct. Orel. S. Benedieti Vita s. Augustini episcopi Cantuariensis primi Anglorunii/ue
apostoli.

* Kolts Series Gcrvasii Cantuariensis opera hislorica.

128

LOMBARDIC ARCHITECTURE

at the east
at Rome,

, raised above a crypt or confessio formed in imitation of that in St. Peter's
and reached through the choir, was faced at the west end by a chapel
dedicated to the Virgin, also elevated above the floor
of the church, and containing the primatial chair.
From the aisles, at a point more than halfway down
the church towards the west, projected two towers,
which at first must have been mere porches, after-
wards raised in height by Archbishop Odo (942-959),
when he increased the walls of the church to a height
of 50 ft. There is no record of tower porches in
England before the Xth century. The southern
tower contained an altar dedicated to Pope Gregory,
while the northern had an altar of St. Martin.

On these main facts, and some secondary ones
derived from Gervase, Professor Willis l based a plan
of Augustine's church, imagination supplying the gaps
where Eadmer's description fails. But we shall deal
with this when we come to discuss St. Mary's at
Abingdon (675).

THE ClIURCH OF ST. PANCRAS AT CANTER-
BURY. Thome 2 and Elmham 3 say that this was a
pagan temple which Augustine converted into a church and dedicated to St. Pancras.
This transformation (" quod
phanum . . . mutavit in eccle-
siam ") certainly took the form
of rebuilding, for the existing
remains show no trace of a
temple, but belong to a church
with an elongated apse at the
east end.

The little church had an
aisleless nave terminating in a
semicircular apse with a bay in
front of it, separated from the
nave by four columns (Fig. 517).
It possessed three porches, on
the west, north, and south sides.
What remains visible above
ground consists of the body of
the church, and very scanty
relics of the presbytery and
apse (Fig. 518).

The plan of St. Pancras,
with its apsidal chancel pro-
jecting by more than a semi-
circle from the outer nave wall to provide room for a presbytery ; with its chancel,

1 The architectural history of Canterbury Cathedral.

- Twysden, Historiae anglicanae scrip/ores decent Chronica Guill. Thome monachi S. Augiistini Cant.

3 Rolls Series Historia inonasterii S. Augnstini Cantitariensis.

lg. 518. Canterbury.

Remains of St. Pancras (Vlth or Yllth
Century).

ANGLO-SAXON ARCHITECTURE

129

divided from the nave by a row of columns ; with its small porch at the west front,
and the side chapels or porches, exercised a notable influence on the form of later
English churches, right up to the Norman Conquest. The origin of the plan is involved
in great obscurity, and we shall endeavour to give an explanation which may throsv
some light on it. The idea of a row of columns dividing the nave from the sanctuary
was borrowed from a church with which the monk Augustine must have been well
acquainted, as it stood but a few steps from his monastery of San Gregorio ad
Clivuin Scattri at Rome. This church, the relationship of which to the other has not
hitherto been suggested, is the building which some archaeologists identify with

THE PALATINE CHURCH OF SAN CESAKIO. The " Ecclesia S. Caesarii in
Palatio " is first mentioned in the reign of Phocas (603), and for the last time in the
XlVth century. Though its site has not been identified,
Lanciani 1 is inclined to place it in the ruins of the so-
called Baths of Heliogabalus on the Via Sacra (Illrd
century). Others think that it was an oratory of the IVth
century fitted up in the " Domus Augustana," and after-
wards converted into a church. 2 Without attempting to
decide between such conflicting views, we will confine
ourselves to the simple statement that a church was un-
doubtedly fitted up in these ruined buildings about the
end of the Vth century. The rough construction of the
inserted masonry, consisting of alternate courses of
used-up bricks and blocks of tufa, with a liberal use of
mortar, points definitely to the period following the days
of Leo I (440-461) and the terrible sack of the city by
Genseric's Vandals (455), and preceding the revival of
building under Theodoric (493-526). The remains of
Santo Stefano on the Via Latina, erected in the time of
that great pontiff, are very instructive in this connection,
for they show the same style of masonry as the inserted
walls of the reputed San Cesario, though it is not so rough.

The fragmentary remains of the church in question,
of which I append an approximate plan (Fig. 519) as it
appeared when excavated in 1872, consist of a nave
ending in a spacious presbytery marked off by two
columns and two pilasters (Fig. 520). The presbytery
is flanked by two sacristies, as in the large basilica of
St. Symphorosa on the Via Tiburtina, and is provided
with an apse having a passage round it behind the altar.
An exactly similar arrangement is presented by an

ancient dated example, the basilica at Benian in Algeria Fig. 519.- Rome. Plan of so-called
, . ~ . ' , L r i San Cesario al Palatino (Vth

(434-439). 3 facing the apse at the further end of the Century).

church is the tank for baptism.

It will be readily seen that this church is related to those of the Anglo-Saxon
period, not only by its pillared chancel screen, but also by the cruciform shape given

1 The riiins and excavations of ancient Rome.

2 Nuovo Bull, di Archeohgia Cristiana, 1907 Bartoli, Sccferta delf oratorio e del monaslero cii San Cesario
sul Palatino.

3 Gsell, Les monuments antiques de t ' Algtric.

VOL. II K

130

LOMBARDIC ARCHITECTURE

to the presbytery and
found, for instance, in
the churches of Repton
(XthorXIth century),
Deerhurst (Xth cen-
tury), and St. Mary's
in the Castle at Dover
(Xlth century).
Another point of con-
tact is the prolongation
of the apse, a feature
already presented by
St. Symphorosa.

The large rectan-
gular chambers at the
sides of the presbytery
of the reputed San
Cesario, transferred to
the aisleless nave of
St. Pancras form the
typical chapels which
some English writers 1 2
believe to be the germ
of the transept of later
times, ignoring the fact
that the Romans had
constructed not only

buildings in the form of a Latin cross, but also cruciform structures with a dome over

the crossing, as may be seen in Montano 3 and other sources. And they were followed

in this by the builders of Ravenna, as may still

be seen in the mausoleum of Galla Flacidia

(about 440), and as might have been seen a

few centuries ago in the church of Santa Croce

(about 449). Consequently there was no occa-
sion to evolve the idea afresh by means of

tentative experiments.

With regard to the external buttresses of

St. Pancras, we know how this form of support

had been already developed in Italy. The

single-bay porch at the west end was derived

from the one in front of the portico of the

Constantinian St. Peter's. Lastly, as to the two

side porches, we may remark that the scheme

of giving more than one portico to a building

had been already introduced by the Romans.

Montano 4 and Bramantino 5 give illustrations of buildings with several entrances

each approached through a porch (Fig. 521). We may also remember that

Fig. 520. Rome. Remains of so-called San Cesario al Palatine (Vth Century).

Fig. 521. Rome. Plan of tomb on the Via
Appia. (From Montano, "Li Cinque
libri fii archi/eihira.")

1 F. Bond, Gothic Architecture in England.
3 4 Of. cit.

2 Prior, A History of Gothic Art in England.
6 Op. cit.

ANGLO-SAXON ARCHITECTURE

before this time an example of two lateral porches is presented by the
celebrated

CHAPEL KNOWN AS THE TEMPLE OF THE CLITUMNUS, NEAR SPOLETO.
It consists of a cella measuring inside about 15 ft. x 10 ft., with a semicircular recess
at the east end, and a vestibule or narthex flanked by two porches which have been
damaged. Cella, narthex, and porches, all have barrel vaults, while the apsidal recess
is covered by a half-dome. The walls are constructed of squared blocks of limestone,
and have been tampered with at the sides. At the top runs a dentil cornice. The
cella was originally lighted by an
opening above the entrance : the
windows on the south side have
been made later (Figs. 522, 523).

Inside, above the small apse
(partly buried under the modern
road to Foligno) is a tympanum
with the decussate cross-mono-
gram among scroll work with
roses. An elaborate cornice runs
round the base of the half-dome
at the impost line. The pediment
above the exterior of the apse
contains the cross-monogram be-
tween scrolls with roses, bunches
of grapes, poppy heads, and vine
leaves.

The elaborate pediment of
the facade is supported by four
marble columns, two of spiral
design and two covered with im-
bricated water-leaves, and two
pilasters. They are surmounted
by Corinthian capitals with acan-
thus leaves, or with stiff leaves
combined with acanthus, or, in a
third case, with palmetto leaves,
acanthus, and scroll work with roses. The pediment contains the cross-monogram,
between scroll work bearing roses, vine leaves, clusters of grapes, and poppy
heads.

De Rossi l regarded this architectural gem as a pagan sanctuary transformed
into a Christian church in the Theodosian age (379-450). Grisar, 2 on the other hand,
believes it to be a building of the pre-Christian period, constructed in its lower part
of old materials, and afterwards rebuilt by Umbrian "marmorarii" in the Xllth
century.

My view, on the contrary, is that it is a structure of one date, and that of
Christian times, and contemporary with the basilica of San Salvatore or the

1 Bull, cfarch. fns/iana,lSyi Spicilegio d archeologia cristiana neU' Umbria Del temfietto sulle rive del
Clitunno, consecrate al culto cristiano.
- a Nuovo bull, cfarch. cristiana, 1895 // tcmpio del Clitunno e la chiesa sfoletina di San Salvatore.

K 2

Fig. 5 22 - Temple of the Clitumnus near Spoleto
(IVth Century).

132

LOMBARDIC ARCHITECTURE

Crocifisso at Spoleto, i.e. perhaps the reign of Constantine. Thus, though its original
masonry is composed of materials taken from earlier buildings, it is obviously
the result of a single constructive effort. The architectural ornaments, too,
were executed at one and the same time. A comparison of the carving on the
facade with that in the interior above the apse is quite enough to prove this.
And the capitals of the vestibule, in spite of their differences in design, were executed

at the same moment.

These carvings, in their turn,
are obviously contemporary with
those in San Salvatore. Compare
the continuous impost cornice of
the altar recess in our chapel with
that in the apse and those in the
central square space of San Sal-
vatore, and the correctness of my
statement will at once become ap-
parent. Moreover, the surviving
Corinthian capital belonging to
one of the pilasters on the front of
San Salvatore is clearly of the
same date as the two imbricated
columns in the chapel by the
Clitumnus. And the carving on
the front of the chapel is, on com-
parison, seen to be contemporary
with that on the front of San
Salvatore. At the most one might
say that the latter has a slightly
more classical air about it.

Before leaving the subject we
should notice that here, as in the
case of San Salvatore at Spoleto,
topographical reasons were respon-
sible for the eastern orientation, as the chapel was built against the steep cliff, with its
front turned to the ancient Roman road which must have run below it.

THE CATHEDRAL OF ROCHESTER. The church of St. Andrew at Rochester
was built (604) by King Ethelbert for Justus its first bishop (604-624), subsequently
archbishop of Canterbury (624-627).! Of this earliest cathedral, as is supposed for
there is no record of any rebuilding the foundations of part of the east end were
discovered in 1889 below the west front of the present cathedral and the roadway
before it. The remains showed an aisleless nave ending in an apse at the east ; but
there was no indication whether or no they were separated by a row of columns.

THE CATHEDRAL OF YORK. We learn from Bede 2 that King Edwin (617-633)
began to build (627), under the direction of Paulinus, the first bishop of York
(627-633), a basilica dedicated to St. Peter, which his death at the battle of
Hatfield prevented him from finishing. This was done by King Oswald (634-642).

1 3 Beda, Hist. eccl.

F 'g- S 2 3- Temple of the Clitumnus near Spoleto. Side view
(IVth Century).

ANGLO-SAXON ARCHITECTURE 133

In the course of the works undertaken after the fire of 1829 a large part of the
area occupied by the church of Paulinus came to light. It was found to be a cruciform
building with a very short transept. As no traces of the sanctuary were discovered
it is not known whether the cross was, as is very probable, imperfect, i.e. without the
topmost limb, or of the perfect " Latin" form. Two fragments of the nave walls of
this church can still be seen in the crypt of the present cathedral.

The cathedral of York is supposed to have served as a model for the primitive St.
Peter's at Peterborough, founded by Peada, king of Mercia (655-656 or 657), and
erected by the first abbot Saxulf. Burned by the Danes in 870 it was rebuilt with
the help of King Edgar (957-975) by Ethelwold, bishop of Winchester (963-984) ; l 2 3
and it is not possible, in their present state, to date with certainty the remains of
transept and presbytery walls of some early church existing underneath the present
cathedral.

THE CHURCH OF ST. MARY AND ST. ETHELBURGA AT LYMINGE (KENT)
was built by Queen Ethelburga after the death of her husband, Edwin of
Northumbria (633), or, to be precise, in the year 640.* There exist near the present
church, which is ascribed to Dunstan, some remains of its foundations showing that
the nave and chancel were separated by two columns.

Following the type of the buildings which we have described was erected

THE CHURCH OF ST. PETER ON THE WALL AT YTHANCAESTIR (BRADWELL)
IN ESSEX, of which the walls of the aisleless nave still exist. It is supposed
that these remains belong to the church erected by Cedd of Lindisfarne, bishop
of the East Saxons (6S3-664). 5 This view we cannot accept. We learn from Bede
that the church of the monastery at Lastingham, founded by Cedd himself (648), was
constructed of wood, and that it was only later that a stone church was built in
honour of the Virgin, into which the body of the founder was transferred. We can
hardly imagine that, when he had used timber for the church of Lastingham, which
was so near to his heart, he would go on to erect another in masonry after the
Augustinian type of St. Pancras at Canterbury. We must not forget the tenacity
with which the Irish missionaries clung to wooden construction, even when they went
to Italy and were in contact both with ancient buildings and with the rising
Lombardic School. It is far more likely that the first church at Ythancaestir
was of wood, and that later it was rebuilt in stone under the influence of Canterbury,
after the Synod of Whitby (664).

Some writers (Amico Ricci 6 and Merzario 7 among them) state that the churches
built at the time of Augustine's mission were the work of Italian craftsmen,
particularly those belonging to the Comacine gilds, who followed him together with
the missionaries sent by Gregory the Great. This, however, is an arbitrary assertion.
The Pope sent with Augustine monks only (" misit servum Dei Augustinum et alios
plures cum eo monachos timentes Dominum "), not monks and craftsmen. The
" operarii," whose small numbers compared with the abundant harvest of converts
had been deplored by Augustine in the message sent through Peter and Laurence,

1 Rolls Series The Anglo-Saxon Chronicle. * Beda, Hist. eccl.

3 Sparke, Hisloriae anglicanae serif tores varii Chnnicon Angliae per lohannem abbatem Burgi S. Petri
Hugonis Candidi coenobii Burgensis historia.

* Kails Series Elmham, of. cit. 5 Beda, Hist. eccl.

Op. cit. i Op. cit.

i 3 4 LOMBARDIC ARCHITECTURE

were not material workmen but missionaries, and it was of such that Gregory
sent him a fresh supply (601) (".... misit cum praefatis legatariis suis plures
cooperatores ac verbi ministros ") to the fullest extent that was in his power
("exhaurit monasteria sua et ecclesias, et quidquid potest eruditorum ac religiosorum
virorum in opus Evangelii efflagitare . . . certat delegare "). 1 2 The fact is these
buildings must have been designed and carried out by the monks. Saxulf, the first
abbot of St. Peter's at Medeshamstede ("Peterborough), is actually described in Bede 3
as the " constructor " of the monastery founded by Peada. And though it is reason-
able to suppose that, in consequence of the wars and disorder prevailing after the
Anglo-Saxon invasion, England became practically devoid of any national art,
and the tradition of beautiful things was lost as the artists of the old time went down
to the grave without imparting any lessons to their successors, yet we cannot
imagine that a school of builders, however rude, did not continue to exist ; still less
that the very stamp of it had perished. We may therefore fairly assume that the
builders of the structures in question were English. In any case the entire absence
of architectural ornaments and carving excludes the presence of Italian or French
artists in these works, and so confirms the opinion expressed above.

* *

When Oswald became King of Northumbria (634-642), there was a pause in the
extension of the influence of the Church of Rome, while the activity of the
Irish Church was correspondingly accentuated. The latter dated from the
Vth century, its first two bishops being Palladius (431-432) and Patrick (432-461).*
Oswald as a boy had found refuge within the walls of the monastery founded
by Columba (+ 597) m the island of lona, where he had taken up his abode about the
year 563, and whence he drew Aidan and other clergy belonging to the Celtic rite
in which he had been baptized. The foundation of the monastery at Lindisfarne
followed. Here Aidan fixed his episcopal seat (634-652), and from it soon issued
bands of missionaries who spread over the realm of Oswald (where with his help they
erected churches) and the various other states, but not over Kent which belonged
to the Roman obedience.

Oswald's death at the battle of Maserfield and the consequent supremacy of the
pagan king, Penda(626-654or655), retarded for a space the activity of the missionaries,
but it became more lively than ever the moment that Penda was dead and Oswy had
ascended the throne (642-670). So that the North of England, so far as it was
Christianized, had for its primate, not the archbishop of Canterbury, in spite of the
rights which he asserted, but the abbot of lona. At length the Synod of Whitby
(664) inflicted a mortal blow on the flourishing Celtic community ; and the Latin
Church obtained a firm seat in the island with no rival to dispute her sway.

Of the churches erected by the missionaries of lona and Lindisfarne or its
daughter monasteries I have not been able to find traces of any one constructed of
masonry. Perhaps this is because they were all built " in the Celtic fashion " or " in
the Scotch fashion," that is to say of wood, like the cathedral erected by Finan (652-
662) at Lindisfarne (" quam tamen more Scotorum non de lapide sed de robore secto

1 Beda, Hist. eccl.

* D'Achery, B. Lanfranci Cant, archiep. opera (Migne, Pair. Lat., vol. 150) Appendix Vita S. Augustini
ex veterc MS. Beccensis abbatiae.

3 Hist. eccl. 4 Bury, The Life of St. Patrick and kis place in History.

ANGLO-SAXON ARCHITECTURE 135

totam composuit atque arundine texit"), the reed and thatch roof of which Bishop
Eadbert (687-697) replaced by a complete covering of sheets of lead (" ablata arundine
plumbi laminis earn totam, hoc est et tectum es ipsos quoque parietes eius cooperire
curavit "). 1

We need not be surprised at the Irish clergy erecting structures of this kind in
England, when we remember that the church built by Columban (f 615) in Italy
at Bobbio, in the reign of the Lombard king, Agilulf (590-615), was also of wood :
"ecclesiam in honore almae Dei genitricis semperque Virginis Mariae ex lignis con-
struxit." 2 It has been suggested 3 that the church of Lindisfarne was rebuilt in stone
by King Ecgfrid (670-685), and that its tri-apsidal arrangement was due to Theodore,
archbishop of Canterbury (668-690). But the account in Bede, who lived about the
same time (677-735) and was familiar with the place and circumstances, is fatal to the
theory. The remains of a stone church, which had, as a matter of fact, three apses,
discovered to the east of the Xlth-XIIth century priory church, must be assigned to a
date after its second destruction by the Danes (867), and probably to the reign
of Athelstan (924 or 925-940), who was such a benefactor to Lindisfarne, 4 or even
later.

* *

The struggle for supremacy between the Latin and the Celtic Churches in
Northumbria was conducted on the side of Rome by two energetic champions,
Benedict Biscop and Wilfrid, the latter being the most prominent. And it was
carried on with a policy clear-sighted, determined, at times even high-handed, at
least in the case of Wilfrid, who was always more inclined to carry peace than war
in the folds of his tunic ; but also with the aid of one of the arts which is more closely
connected than others with the instincts, the ideas, the progress, the needs of a people
I mean architecture. And so they set themselves to erect churches in the country,
the work of builders of more skill than could be found at home, and recruiied beyond
the Channel : churches with glass windows, a new thing in Great Britain, and some-
times paintings brought from Rome and intended to form a " biblia pauperum " to
teach the unlearned the facts of the sacred story. One of these churches was
constructed in so elaborate a fashion as to make it the most notable of which
the district could boast for a long time to come. Let us see what remains of
them.

THE CHURCH OF ST. PETER, MONKWEARMOUTH, was built in 675 by Benedict
Biscop, first abbot of Wearmouth and Jarrow, near the monastery which he had
founded (674) at the mouth of the Wear with the aid of King Ecgfrid (670-685).
Bede 5 informs us that the church was built of stone in the Roman style, which was
always preferred by Benedict, certainly as against the Celtic fashion of wooden
construction. And it was carried out by workmen brought on purpose from
France.

Of the original building, which took but a year to finish, there remains only the

1 Beda, Hist, eccl,

' 2 Mabillon, Ada Sand. Ord. S Benedi<ti -Aliracula Columbani scripta a monaeho Bobbiensi.

3 Archaeologia Oxoniensis, 1893 Park Harrison, On an early illuminated manuscript at Cambridge.

4 Rolls Series Symeonis monachi his/aria ecclesiae Dunelmensis.

5 Vita sanctorum abbatum monasterii in Wiramutha et Girvium.

i 3 6

LOMBARDIC ARCHITECTURE

Fig. 524. Monkwearmouth. West
end of church (Vllth Century).

west front and its two-storied porch (Figs. 524, 525),
the outline of whose gable can still be seen in the
third stage of the present tower. The western
opening of this porch is sustained by short baluster
shafts (Fig. 526), turned on the lathe, standing on
high plinths ornamented with intertwined serpents,
whose long, beak-like jaws interlace, like some of
the winged creatures on the famous Bayeux Tapestry
(Xlth century). In the tower which rises above, the
upper part is no doubt due to the restoration carried
out in 1075 12 by the monk Aldwin, with the assist-
ance of Walcher, bishop of Durham (1071-1080),
after the destruction caused by the Scotch king,
Malcolm III (1054-1093), and certainly before 1083,
when the monks of Jarrow and Monkwearmouth
were transferred to Durham. It has, indeed, been
suggested that the addition to the tower belongs to
the years between the foundation of the church and
the devastations of the Danish hordes (867), who
burned the monasteries of Monkwearmouth, Jarrow,
Tynemouth, Lindisfarne, and Whitby: But, as we
shall see when we come to the abbey church of Ram-
sey, the characteristic western tower makes its first
appearance in connection with datable English
churches only in the Xth
century, and then as an
importation from France.
The tower at Monk-
wearmouth is closely re-

lated to that of St. Cuthbert's, Billingham (Fig. 527), which
must not be identified with the church of Ecgred, bishop of
Lindisfarne (83 1-846), 3 but was erected after the Conqueror
had given back Billingham to St. Cuthbert (1072).* It is
also related to those of St. Mary, Ovingham, and St. Mary
Bishophill Junior at York, which belongs to the time of the
Conqueror's rebuilding of St. Mary's Abbey, York (io88). 5
All the towers with two-light openings enclosed by a
common arch standing out from the face of the wall, or
framed by a single arch within a rectangular recess, are
later than the Conquest, which ushered in their appearance
in England. Monkwearmouth, Jarrow, and Billingham
provide instances.

In St. Peter's, Monkwearmouth, we find for the first
time in Great Britain the so-called " long and short work,"
in which the stones at the salient angles and the jambs of

1 Soils Series Symeonis tnonachi hist. Eccl. Dunelin.

2 3 Soils Series Symeonis monachi historia Segum.

1 Rolls Series Symeonis monachi hist. Eccl. Dunelm.
5 Dugdale, op. cit.

Fig. 526. Monkwearmouth
church. Baluster shafts in
outer west door (Vllth
Century).

ANGLO-SAXON ARCHITECTURE

'37

the openings are set al-
ternately horizontally
and vertically. A By-
zantine origin has been
sought for this form,
and a pedestal (Fig.
528) in the Acropolis
at Athens has been
produced as evidence ;
but this structure be-
longs to a monument
erected in honour of
M. Vipsanius Agrippa
in 27 B.C., and has
nothing in common
with the work in ques-
tion. Bonding of this
kind, used in the angles
of irregular masonry,
and forming a source
of weakness rather
than of strength, was
undoubtedly a product

Fig. 525. Monkwearmouth church. Interior of west end (Vllth Century).

of barbarism in the art of quoining. Its introduction into Britain, to judge by what
we know, must have been due to French craftsmen, perhaps from Poitou, as there is

no other locality where we find its use so deeply rooted

and with the tradition of a thousand years behind it ;

for instance, at Poitiers, where the interior of the

narthex of the bap-
tistery of Saint Jean

exhibited it as early

as the first years of

the Xlth century, and

masonry, both an-
cient and modern, in

secular buildings still

provides numerous

illustrations of it for

the observer to-day.

In this connection it

is interesting to note

that, just at the time

of Benedict Biscop,

there is evidence of

direct contact be-
tween the North of

England and Poitou

in a fragment of the

will of Ansoald,

Fig. 527. Hillingham. Tower of
St. Cuthbert's (Xlth Century).

Fig. 528. Athens. Acropolis. Base of
monument to Agrippa (1st Century B.C.).

I 3 8

LOMBARDIC ARCHITECTURE

Fig. 529. Edinburgh. Museum. Top of
Roman altar from liirrens.

bishop of Poitiers (682-696), from which we learn that he appointed a bishop called
Romanus from the land of the Scoti, accompanied by a band of his countrymen, to

govern and occupy the monastery of Mazerolles
sur Vienna which had been restored by him. 1

Before leaving Monkwearmouth I should
like to say a word about the baluster shafts. I
have never been able to discover any of earlier
or even contemporary date with these. They
seem to me to reveal the co-operation of native
workmen in the construction of the church. The
use of supports of this form for decorative pur-
poses was of great antiquity in the island. An

altar of the Roman period discovered at Birrens (Dumfriesshire), and now in the
Museum at Edinburgh (published in the " Transactions and Journal of Proceedings
of the Dumfriesshire and Galloway Natural History and Antiquarian Society,
1895-96") has a carved representation of an arched doorway with the jambs formed
as baluster shafts (Fig. 529). Another small Roman altar from Lanchester in the
Cathedral Library at Durham has also been mentioned in this connection, but the
pediment of the shrine represented on it does not rest on baluster shafts, but on a
bead and reel and cable moulding.

ST. PAUL'S CHURCH,
JARROW, founded by Bene-
dict Biscop and Abbot Ceol-
frid, with the assistance of
King Ecgfrid (670-685), was
consecrated in 684- 2 The
nave of the original building
still exists forming the pre-
sent chancel (Figs. 530, 531).
It was restored in the course
of Bishop Walcher's 3 work
carried out after 1074, at the
same time as the erection of
the existing porch with its
tower. In the north porch
of the present church in the
Pointed style are preserved
some ancient baluster shafts
and fragments of a string-
course with miniature balus-
ters and vertical rolls of
sausage shape in relief, of
the same date as similar
fragments belonging to Wilfrid's church at Hexham, now in the
Library at Durham. There are also various pieces of carving, some

1 Chamard, Histoire eccttsiastique de PoitouMJmoires de la Sociiti des Antiquaires de FOuest,

- Beda, Vita abbatum.

3 Rolls Series Symeonis monachi hist. Ecfl. Dimelm.

Fig. 530. Jarrow. Nave of the original church (Vllth Century).

Cathedral
of which

ANGLO-SAXON ARCHITECT! 'R I

'39

g- S3 1 - Jarrow Church. South side of the original nave
and later tower (Vlllh and Xlth Centuries).

may be ascribed to the French artists
brought over by Benedict Biscop. They
consist of a fragment with interlacing
and birds, and another exhibiting
interlacing with part of a recumbent
figure, and also a man disentangling
himself from the interlacing. They go
with another fragment at Hexham, and
part of a cross from Jarrow at Durham.

THE BASILICA OF ST. ANDREW
AT HEXHAM was erected by Wilfrid
(634-709) between 672 and 678. Its
dedication is the same as that of the
church built by Gregory the Great in
his monastery near the Clivus Scauri,
from which came the missionaries who
evangelized the Anglo-Saxons. Eddius
Stephanus, 1 the biographer and chap-
lain of the energetic bishop, has left
a short account extolling the size
(" mirabilique longitudine et altitu-
dine ") and splendour of the building
("neque ullam domum aliam citra Alpes montes talem aedificatam audivimus ") ;
but these eulogies must be discounted, as it is easy to see that language of this kind
originated in the great poverty of ecclesiastical structures which might serve as
standards for contemporary descriptions. These characteristics are confirmed by
Richard of Hexham 2 (who has left a fairly detailed account, though with some
omissions, of Wilfrid's church), Simeon of Durham, 3 William of Malmesbury, 4 and
Eadmer. 5

Of the original church, which still retained its beauty in the time of William
of Malmesbury (Xllth century), as he himself says in his " Gesta pontificum
Anglorum," but was wrecked by the Scotch in 1296, there is preserved the crypt,
above which rose the sanctuary, as we learn from the metrical biographer of Wilfrid,
Frithegode. 6 It consists of a chamber reached through a vestibule, with three
passages for entrance and exit. There are also some remains of the apse.

With regard to the actual builders of St. Andrew's we are informed in general
terms that Wilfrid procured them from the Continent (" adductis secum ex partibus
transmarinis artificibus " " De Roma quoque et Italia et Francia et de aliis terris
ubicumque invenire poterat, cementarios et quoslibet alios industrios artifices secum
retinuerat "). But we are also told explicitly that they were brought from Rome
(" sed et cementariorum, quos ex Roma spes munificentiae attraxerat,
magisterio . . ."), which finds its natural explanation in the fact that the Roman
workmen, accustomed as they were from the time of Constantine onwards to erect

1 Rolls Series The historians of the Church of York and its archbishops Vita Wilfridi etiscopi.
3 Twysden, op. cit. De statu et episcopis Hagustaldensis Ecclesiae.

3 Rolls Series Hist. Regum.

4 Rolls Series Gesta pontificum Anglorum.

6 Rolls Series The historians of the Church of York and its archbishops Vita Wilfridi episcopi.
* Rolls Series The historians of the Church of York and its archbishops Vita S. Wilfridi episcopi.

140

LOMBARDIC ARCHITECTURE

colonnaded basilicas with ancient materials, were the best fitted for carrying out
the bishop's ideas. 1 2 3

For the works which they were engaged in executing under his direction he had
formed a clear idea during his sojourns in Rome, where, as early as his first visit
(about 654), he was able to admire, among other things, the galleried basilicas of
Sant' Agnese fuori le mura (Fig. 532) and the Santi Quattro Coronati (Fig. 533),

recently rebuilt by Pope Honorius I
(625-638).

The "title" of the Santi Quattro
is mentioned as far back as the time
of Gregory the Great, and dates, ac-
cording to some authorities, from the
Vth century. Duchesne 4 believes that
the church goes back to the I Vth cen-
tury, which would explain its western
orientation, unless indeed this is the
result of reasons connected with the
site. It was rebuilt by Honorius I,
Leo IV, and lastly, after the Norman
fire, by Paschal II, who began his work
on it in 1109, and consecrated it in
1 1 12. It was remodelled and redeco-
rated in the XVth, XVI th, and XVI Ith
centuries. 567

In my opinion the traditional his-
tory of the Santi Quattro does not
quite correspond with the facts ; and
the reconstructions under Leo IV and
Paschal II consisted, in the former
case, of restorations and new additions,
in the latter, of mutilations and addi-
tions, preserving in part the arrange-
ments of the church of Honorius I.
This view is confirmed by the apse, in
which the untouched masonry of the
exterior belongs to the work of
Honorius up to a point above the
large original windows with their double
ring of bricks. If we compare it with
the brickwork of the time of Leo IV

(t.e. the IXth century) to be seen in the apses of Santa Cecilia in Trastevere (817-824),
and San Giorgio in Velabro (827-844), and in the nave of Santa Prassede (817-824),
all apparently the work of the same masons, we shall notice a less marked
decadence, and in any case a difference in date. In the next place, no one would

1 Mabillon, Ada Sanct. Ord. S. Benedicts Vita S. Wilfridi episcopi auctore anonimo sec. XII.

2 Twysden, op. cit. Ricardus prior Hagiistaldensis de statu et episc. Hagust. Eccl.

3 Rolls Series Willelmi Malm, tnonachi gesta pontificum Anglorum.

* Le liber pontificalis. 5 De Rossi, Bull, d'arch. cristiana, 1879.

8 Lanciani, Storia aegli scavi di Roma. 7 Armellini, Le chiese di Roma dal secolo IV al XIX,

Fig. 532. Rome. Basilica of Sant' Agnese outside
the walls (Vllth Century).

ANGLO-SAXON ARCHITECTURE

141

ascribe it to Paschal II. Very differ-
ent were the exteriors of buildings of

his time, especially in the treatment

of the windows, as we see from the

circular and round-headed specimens

in the nave of Santi Giovanni e Paolo,

which he restored after the Norman

fire, and those in San Clemente, the

rebuilding of which was started before

1 126 or 1 128. All that he did to the

apse of the Santi Quattro was to

repair it, and insert the range of

consoles derived from some ancient

building. At Rome, the apses of

Christian buildings earlier than the

epoch of about idOO never have any

but borrowed consoles, and it is a

mistake to assign to the Vlth cen-
tury 1 the eaves cornice on the apse

of San Martino ai Monti (498-523),

formed of carved consoles supporting

sculptured slabs and panels with

masks and other ornaments, for it is

made up of ancient materials.

The apse of the Santi Quattro at

a later date was raised to its present

height. As for the body of the church, Paschal II abandoned the two old aisles, and

walled up their colonnades ; shortened the nave, within which he constructed a new

nave and two aisles ; and in the latter repeated the arrangement of galleries as it

existed in the church of Honorius I. It was not till the times of Pelagius II, in the

case of his transformation
of the Constantinian
basilica of San Lorenzo
in Agro Verano (Fig.
534), Honorius I, and
Hadrian I (as Santa
Maria in Cosmedin
showed), that galleried
basilicas, with two stories
of colonnades after the
pagan type, were erected
in Rome.

To return to Hex-
ham. The three churches
of Pelagius II and Hono-
rius I referred to must
have formed the models
for Wilfrid's structure, for

Old Basilica of San Lorenzo in Agro Verano

(Vlth Century). ' Mazzanti, op. (it.

Fig- 533- Rome. Basilica of the Santi Quattro Coronati
(Vllth and Xllth Centuries).

F 'g- 534- Rome.

I 4 2

LOMBARDIC ARCHITECTURE

it was provided with just the same colonnades in two stories, a feature indicated by

a passage of Prior Richard's chronicle : " parietes autem quadratis et variis et bene

politis columpnis sussultos, et tribus tabulatis dis-
tinctos immensae longitudinis et altitudinis erexit."
And this explains William of Malmesbury's reference
to the likeness between St. Andrew's and churches
at Rome : " Nunc qui Roma veniunt idem allegant,
utqui Haugustaldensem fabricam vident ambitionem
Romanam se imaginari jurent."

We have still to consider who executed the
carving. I do not include, of course, the figured
capitals mentioned by Prior Richard, because I
believe they were Roman antiques brought from
elsewhere. Britain, in Roman times, could not have

Fig- 535. Durham. Cathedral Library, been barren of capable executants of figure subjects.

On the other hand, in the Vllth century the chisels
of Rome and France, which it seems were at work

on Hexham, did not produce figure capitals (I say capitals and not pulvins), for

there is not a single specimen in existence.

Some of the carving from Hexham is preserved in Durham Cathedral Library.

It includes three fragments of stringcourse showing upright

balusters, cable mouldings, and rolls arranged either horizon-
tally or in zigzag fashion (Fig. 535). The rudeness of both

design and execution, as well as the style of ornamentation,

suggest that they are the work of British hands, with whom

baluster shafts were a favourite feature as far back as the

Romano-British period. We cannot imagine that Wilfrid

would have taken the trouble to bring artists from beyond

the seas in order to produce work of this kind. These frag-

-.1 .1 r ii i-i i r-. Fie- 5^6. Hexham. Capital in

ments go with others of the same kind preserved in St. st Andrew's (Xlth Century).

Paul's, Jarrow, and the portion of a stringcourse with a roll

between two cables now in the north arm of the transept of the existing church of

Hexharn.

Of the carving executed by Continental artists for Hexham we have no speci-
mens which can be identified with certainty. We
may, however, ascribe to them some of the frag-
ments gathered in the north arm of the transept,
though their provenance can only be said to be
Hexham and not definitely the church. Among
them are three Lombardic cubical capitals (Fig.
536), which have been made to pass for Anglo-
Saxon, but are really later than the Conquest
(1066). One might search in vain among dated
buildings in England for capitals of similar type
till we come to the abbey church of St. Augustine

Fig. 537> Hexham. Church of St. Andrew. .-. , L !.. i_ A i_i_ o j

Fragment of carving ( VI ith Century). at Canterbury as rebuilt by Abbot Scotlandus

(1070-1087). The earliest trace of simple hemi-
spherical capitals that I have been able to find in this country is the representation
of an arcade with two bulbous capitals and one of hemispherical form in the well-

ANGLO-SAXON ARCHITECTURE

known MS. " Liber Geneseos, caeteraeque historiae sacrae " in the British Museum,
which is thought to be rather earlier than 1066.

To the time of Wilfrid may be assigned a fragment with vine foliage showing a
cock, and the legs and one arm belonging to two human figures (Fig. 537), all in low
relief. The composition, design, and technique, show that it comes from the same
school, and possibly from the same hand, as that which produced the two fragments
with scroll work, birds, and human beings, which we noticed at Jarrow.

With the carving of the time of Wilfrid and Benedict Biscop executed by foreign
artists may be connected the very decayed portions of the upright limb of a cross,
supposed to be that of Acca, bishop of
Hexham (710-740), or rather one of the
two crosses which stood at the head and
foot of his grave : " Duaeque cruces
lapideae mirabili celatura decoratae
positae sunt, una ad caput alia ad
pecles eius." l The fragments, covered
with a very intricate vine stem design,
have been set up in the Cathedral
Library at Durham, to which they were
brought from Hexham (Fig. 538). The
date of the carving may be that which
is generally assigned to it. We know
that Acca embellished St. Andrew's, 2
and we need not be surprised if the
foreign artists of Wilfrid's time were
followed by others at a later date who
produced these gravestones. It is clear
that the carving belongs to a period
which, if not that of Wilfrid, is not far
removed from it ; and it is equally clear
that it comes from a French hand. I
say this because the carvers of Rome
and Ravenna, at that date the best in
Italy, did not produce such complicated
interlacings ; and those of Lombardy,
though very fond of employing them, were unable to treat them with the grace shown
by the cross from Hexham.

All this carving in relief is quite different, both in composition, design, and
technique, from that of the well-known tall cross at Ruthwell, Dumfriesshire (there
is a cast in Durham Cathedral Library), which cannot be dated earlier than the first
half of the Xllth century (Fig. 539).

Lastly, we can connect with Wilfrid's work the ancient episcopal chair (Fig. 540)
known as " the Frith stool," cut out of a single block of stone, which may be seen in
St. Andrew's. The front is outlined by roll mouldings, and on the arms are carved
interlacings ending in knots. The way in which the framing is executed, and the
simple character of the interlacing, suggest a Roman hand ; especially the mouldings,
which recall works of the Roman and Ravennate schools, beginning with the well-
known screen panels in San Clemente at Rome (Fig. 541).

1 a Rolls Series Sytneonis inonadii hist. Regain,

Fig. 538. Durham. Cathedral Library. Fragment
of cross from Hexham (VHIth Century).

i 4 4

LOMBARDIC ARCHITECTURE

In connection with the better carving of the time of Benedict Biscop, Wilfrid, and
Acca, it has been suggested that in the days of Theodore of Tarsus, archbishop of

Canterbury (668-690), artists from the East came
to Britain to ply their trade. But there is no
mention of such in the chroniclers, and in the
course of my long and frequent wanderings up
and down England I have never been able to
recognise their hand in any of the early carving
still in existence. The presence of such artists
in the island may, then, be relegated to the
domain of fable. Others there are who would
attribute it to Irish chisels. Ireland, they say,
in the Vllth, Vlllth and IXth centuries, was
not only a great school of missionaries but also
of art ; and as evidence they produce the illum-
inated manuscripts of the period. For instance,
the " Book of Dimma," l written by a scribe of
the name, who is supposed to be the same as
the Dimanus mentioned in a letter of Pope
John IV (64O-642). 2 Here we see simple in-
terlacing framing figures of the Evangelists John
and Matthew. Or there are the " Lindisfarne
Gospels," 3 written by Edfrid, as is supposed,
before he became bishop of Lindisfarne (697-
721), and illuminated by his successor Ethelwold
(721-737). Here the interlacing is sometimes
very complicated, and better drawn than in the
" Book of Dimma." The latter is also sur-
Fjg- 539.-Durham. Cathedral Library. Part passed by the " Lindisfarne Gospels " in the

of cast of the Ruthwell Cross (Xllth Century). * .

representation of figures, showing a more ad-
vanced stage of art. Or, again, there is the " Book of Durrow," 4 dated in the
Vllth century; but its more com-
plicated interlacing and the representa-
tions of animals differentiate it some-
what from the " Book of Dimma," and
suggest that it belongs to the second
half of the Vlllth century. A notice-
able peculiarity is the long beak-like
jaws of the animals, a feature recalling
the serpents carved by some French
artist in the porch of St. Peter's at
Monkwearmouth (675). Another ex-
ample is the " Psalterium charactere
Hibernico" in the British Museum, be-
lieved to be of the IXth century, with

Fig. 540. Hexham. St. Andrew's.
(Vllth Century).

Episcopal Chair

/VTTt-V. l '..,,i ... i- 1

interlacing, simple in some cases and

very intricate in others. Or, lastly, there is the " Book of Kells," 5 the most important

1 Trinity College, Dublin.
3 British Museum.

3 Gilbert, National Manuscripts of Ireland.

4 5 Trinity College, Dublin.

ANGLO-SAXON ARCHITECTURE

'45

Fig. 541. Rome. San Clemente. I'luteus from the choir
(Vlth Century).

palaeographical and artistic monument existing in Ireland, and regarded as belonging

to the VHIth century. 1 But though the style of its interlacings connects it with the

" Evangelistarium of Mac Regol "

(f 820) in the British Museum, on

the other hand the quality of the

drawing suggests that it is later

than the " Book of Durrow." Its

date will therefore fall in the

years immediately subsequent

to the foundation of Kells (802-

815) by Cellach, who, with the

Columban community, had fled

from lona through fear of the

Danes.

But the interlacing in the
oldest of these manuscripts, the
" Book of Dimma," is evidently
derived from Romano-British
mosaic decoration. There was

no lack in Britain of public and private buildings of that period, rich in polychrome
mosaics. The existence, and also, except in the representations of living beings, the
fine quality of the mosaics, is proved by, to give only one or two instances, those
recently discovered at Calleva Atrebatum (Silchester) and Venta Silurum (Caerwent).
We may mention, too, the mosaic in the Roman Gallery of the British Museum, which
displays, besides knot work and lotus flowers, a circle enclosing eight branches
arranged in the form of the so-called Maltese cross.

Mosaics of this kind might have provided suggestions for the Anglo-Saxon
artists, and, if necessary, for the Irish as well. The fact remains, however, that they
did not do so, for neither England, Ireland, nor Scotland contain any datable carving
of the same type as that which we have classed as the work of foreign artists in the
time of Benedict Biscop, Wilfrid, and Acca, going back to the years between the
calling in of the Northern barbarians (449) and the coming of the craftsmen invited
or brought from the Continent by Benedict and Wilfrid.

The real worth of the carvers at work in Northumbria before 675, or even after
that date but continuing the old style, appears to me to be shown by the well-known
gravestones in Durham Cathedral Library, the Black-Gate Museum at Newcastle-
upon-Tyne, and the British Museum, which were unearthed from the ancient
cemetery of the convent of St. Hilda at Hartlepool, founded by Heiu before 650 2 and
destroyed by the Danes. Each of these gravestones shows a characteristic rude cross
enclosed in a circle, and has a semicircular top ; while the lettering of the inscriptions
corresponds to that of the Irish MSS. of the Vllth century. The fact that a specimen
of this rare type of gravestone has come to light in St. Peter's at Lindisfarne, and
that others have been found at Glendalough and Clonmacnoise in Ireland, demon-
strates its Celtic origin.

The highly complicated interlacing of the Lindisfarne Gospels (even granting
that the illumination comes from an Irish hand, though the name " Ethelwold "
sounds Anglo-Saxon) was done at a time when the Comacine and Lombard gilds
had long been accustomed to produce patterns of the most varied and intricate

1 Abbott, Celtic ornaments from the Book of Kells. z Beda, Hist. eccl.

VOL. II L

146

LOMBARDIC ARCHITECTURE

character in carving. And it was executed under the influence of centres of Latin
culture like Jarrow, Monkwearmouth, and Hexham, where Benedict Biscop, Ceolfrid,
and Wilfrid had formed libraries with books brought from Rome. And this influence
extended in the same way to the artists of the other illuminated manuscripts mentioned
above.

Another source of influence, especially in the case of the intricate interlacing
sometimes with heads of animals which occurs in the " Psalterium charactere
Hibernico,"the "Book of Kells," and the " Evangelistarium of Mac Regol," may have
been the School of St. Gall, which was in touch with the Lombard gilds on the one
side, and on the other kept up relations with the clergy of Great Britain and Ireland.
I have, in fact, seen in that celebrated monastery several manuscripts (one, of local
origin, being of the Vlllth century) with very involved interlacing closely related
to that of the English and Irish manuscripts previously referred to. I may mention
an " Evangelium S. lohannis" (Vlllth century), the " Quattuor Evangelia" (Vlllth

Fig. 542. British Museum. Side ot whalebone casket (Vlllth Century).

century), the "Homiliae S. Gregorii " (IXth century), and the " Psalterium Folchardi "
(IXth century). 1

On the other hand, there is no proof that the artists of that age derived ideas
for carving from the illuminated manuscripts, and still less that they were Irish. It
would indeed have been a singular anomaly if people who were so little used to
stone buildings as to go on erecting wooden structures in the " Celtic " fashion
through the Vllth and Vlllth centuries had been capable of producing carving
of so advanced a character as we find at Monkwearmouth, Jarrow, and Hexham.

Moreover, had not Britain and Ireland suffered from a positive sterility of
artists, Benedict Biscop and Wilfrid would never have incurred the heavy expense
of engaging the services of foreigners. It is more reasonable to believe, as I do,
that it was the instruction and the light derived from the examples left behind them
by the Continental artists which guided the Anglo-Saxon carvers in the work
which they undertook as their successors, resulting in the founding of a National
School.

Certain it is that the productions of the Anglo-Saxon School, particularly in
Northumbria where the said National School was formed, are quite distinct from the
foreign work which provided it with models. In the treatment of figures, and
especially of the human form, there is nothing short of a gulf between them, both

1 Library of the former Abbey of St. Gall.

ANGLO-SAXON ARCHITECTURE 147

as regards design and execution. A convincing proof of this is furnished by the
important whalebone box known as "the Franks Casket" in the British Museum
(Figs. 542, 543). The missing portion (Fig. 544) belongs to the National Museum

Fig. 543. British Museum. Side of whalebone casket (Vlllth Century).

at Florence. This casket, Northumbrian work of the Vlllth century with Runic
inscriptions, betrays foreign influence, e.g. in the scene of the wolf with Romulus and
Remus, and in the canopy with supports showing interlaced ornament and a knot in
a spandrel, evidently derived from a Pre-Lombardic source. The Adoration of the
Magi recalls the precious remains of the oak coffin which once held the body of
St. Cuthbert, now fitted together and exhibited under glass in Durham Cathedral
Library, with its representations of Christ between the Emblems of the Evangelists,
the Archangels, the Virgin and Child, and the Apostles, poor in drawing but freely
cut with the knife or graver, and accompanied by legends in Roman and Runic
characters. Anyone who looks at the human heads represented full face on the

F'g- 544. Florence. National Museum. Missing side of whalebone casket in British Museum (Vlllth Century).

British Museum casket will not fail to perceive the relationship, especially in the
oval outline of the head, between them and those on the remains of St. Cuthbert's
coffin. The latter may very well belong to the year 698, or perhaps 696, as has
been suggested. 1 In any case it cannot belong to a date later than 998, the view

1 The Victoria History of the Counties of England History of Durham Kilchin, The coffin of St. Cuthbert.

L 2

148

LOMBARDIC ARCHITECTURE

Fig. 545. Hexham. St. An-
drew's. Fragment of carving.

being, on every ground, quite untenable which would place it in 1 104, the year of the
translation of the relics of the sainted bishop of Lindisfarne to Durham.

The productions of the Anglo-Saxon School, as reflected in motives of ornament,
are also to be recognized by an interesting feature which
distinguishes English work between the Vlllth and Xlth
centuries. This is the typical complicated interlacing in
which the bands are not given a triangular or merely
rounded surface but have the appearance of intestines.
Numerous specimens are scattered about England, but the
best are to be found in Northumbria because it was there
^^^ that, under foreign influence, the

School had its origin. We may

mention a few of these examples,

the date of which is not always

easy to fix, as this treatment of

interlaced work is not confined

to the Anglo-Saxon period but

sometimes occurs as late as the

XVth century.

(1) The examples preserved at Monkwearmouth, Jarrow,
Hexham (Fig. 545), and in Durham Cathedral Library, which
contains the most important collection of Anglo-Saxon
carving in England.

(2) A sepulchral cross with a Runic inscription com-
memorating one Cynibalth, from the churchyard of St. Mary's,

Lancaster : now in the British
Museum, and assigned to the
late Anglo-Saxon period.

(3) A wheel-head cross
brought from the parish of

Gwinear, and now standing outside the church of
Mawgan in Pyder (Fig. 546). It is considered to be
the finest example of a carved cross in Cornwall, and
might be as-
cribed to the
period following
the submission
of the British
Church in Corn-
wall to the see
of Canterbury in
the reign of

Athelstan (925-940), and the consequent introduction
of Anglo-Saxon influence.

(4) Fragments of sepulchral slabs from the ancient
graveyard of Peterborough Cathedral, found in 1887

under the floor of the north arm of the transept. One of them (Fig. 547)
shows the characteristic Gallic cross, interesting and very early examples of which
in different forms may be seen carved on the tombstones collected in the baptistery

Fig. 546. Mawgan in Pyder.
Wheel -head cross (Xth
Century).

Fig. 548. Hexham. St. Andrew's. End of
top of " hog-backed " tomb (Pre-Conquest).

Fig. 547. Peterborough. Cathedral.
Portion of tombstone in the tran-
sept (before the Xllth Century).

ANGLO-SAXON ARCHITECTURE

149

of Saint Jean at Poitiers. This form of cross was probably imported from France,
as were also the tombstones of the type of that at St. Andrew's, Bolam, with its cross
and fish-bone ornament. National English work can also be recognized in the
characteristic tops of tombstones (difficult to date) of Northumbrian origin, known
as "hog-backed stones," ornamented with crosses, circles, arcading, animal heads,
interlacing. Interesting specimens can be seen in Durham Cathedral Library, and
another in St. Andrew's church, Hexham (Fig. 548).

CHURCH OF ST. MARY AT HEXHAM. Prior Richard of Hexham 1 describes
the form of this church, which was founded by Wilfrid, and formerly stood near
St. Andrew's : " in modum turn's erecta et fere rotunda, a quatuor partibus

F 'g- 55- Perugia. Sant' Angelo (Vlth Century).

totidem porticus habens." In other words, it was a structure of polygonal plan
with four porches or vestibules, like the Roman edifice illustrated in Fig. 521.

It appears, then, that St. Mary's at Hexham was also copied from a Roman
building, either one of the type to which we have just referred, or else a church
with which the bishop had become familiar in the course of his travels, Santo Stefano
on the Caelian, just remodelled by Pope Theodore I (642-649) * on the occasion of
his translation of the bodies of the martyrs Primus and Felicianus from a catacomb
on the Via Nomentana (Fig. 549). It was a concentric circular building with a tower,
and four inserted courts, one of which was made into a chancel. Or the plan may
have been derived from the octagonal Lateran Baptistery with its three chapels and
narthex. In any case it is clear that St. Mary's was not, as has been suggested, an
equal-armed cross with a central polygonal tower, for this does not agree with
Prior Richard's brief description, and is inconsistent with William of Malmesbury's

1 Twysden, op, cit. De statu et episc. Hagust. Eccl.

2 Lanciani, The Kuins and Excavations of Ancient Rome.

1 5 o

LOMBARDIC ARCHITECTURE

account of the new style of Athelney Abbey, of which we shall treat presently ;
unless the shape of a Greek cross was produced by three porches or chapels and

Fig- 549' Rome. Santo Stefano al Celio (IVth and Vth Centuries).

an apsidal chancel, as in the round church 01 Sant' Angelo at Perugia (Fig. 550),
which the latest discoveries show was provided with four projecting arms of this kind.

THE CHURCH OF ST. PETER AT RIPON was erected according to Eddius 1
by Wilfrid between 671 and 678. We gather from his account and that of William
of Malmesbury that it was a basilica with two tiers of arcades with columns, like
St. Andrew's at Hexham. Destroyed by the Danes in the IXth century, 2 it was

rebuilt from the foundations by Roger of
Pont 1'Eveque (1154-1181), archbishop of
York, but Wilfrid's crypt was retained.
This consists of a rectangular chamber
with a barrel vault, approached through a
rectangular vestibule with a half-barrel
vault. Both communicate with two pas-
sages roofed with stone slabs.

ST. ANDREW'S CHURCH, COR-
BRIDGE, is first mentioned by Simeon of
Durham under the year 786. 3 It is sup-
posed that it was built by Wilfrid, 4 and
some remains* of the original structure
support this. They consist of the porch,
the walls of which have been raised, and
the aisleless nave, remodelled at a later

1 Rolls Series The historians of the Church of
York and its archbishops Vita Wilfridi episcopi.

2 Rolls Series IVillelmi Malm, monachi gesta
pontif. Anglorum.

3 Rolls Series Hist. Reguin.

4 The Reliquary, 1893 Hodges, The pre-conquest
F 'g- S5 1 - Corbridge. West end of St. Andrew's churches of Northumbria Cartridge, St. Andrew's

(Vllth Century). Church.

ANGLO-SAXON ARCHITECTURE

date (Fig. 551). There is a large opening in the west end (Fig. 552), the arch of
which must have been transferred bodily from the neighbouring Roman town of
Corstopitum, or from the Roman
Wall, and built into the church
an operation which suggests the
presence of the foreign masons who
worked on St. Andrew's at Hex-
ham.

While the churches which we
have just considered were being
erected through the action of the
Latin Church in Northumbria,
which under Oswald, Oswy, and
Ecgfrid, was the most important
state in the island, others were
rising, or were about to rise,
through the same influence in the
southern districts and in Mercia.

THE CHURCH OF RECULVER
was erected by Bassus under the
patronage of Egbert, king of Kent
(664-673) in 669.* There survive
the foundations, and some frag-
ments of walls above ground.

The original church consisted of a nave and aisles, with a corresponding apse, in
front of which was a short presbytery, separated from the nave by three arches
supported by two columns which are now set up on the north side of Canterbury
Cathedral (Fig. 553).

Its importance consists in these columns on account of the capitals which
surmount them ; for though the shafts and bases go back to late Roman times, the
capitals are the work of Anglo-Saxon hands. They, too, were originally Roman, but
were afterwards re-worked by keeping the base and reducing the body of the capital
to the form of three graduated abaci, chamfered at the angles. The barbarous way
in which this transformation has been carried out is an indication of the abyss of
decadence into which the British carvers had fallen, and also explains why Benedict
Biscop and Wilfrid had recourse to foreign chisels.

SS2. Corbriilge. St. Andrew's.
(Vllth Century).

Interior of west end

THE ABBEY CHURCH OF ST. MARY AT ABINGDON, founded in 675 by its
first abbot Heane with the assistance of a local chieftain Cissa, had the peculiar
feature of two apses at opposite ends of the building: "et erat rotundum tarn in
parte occidentali quam in parte orientali." 2 It was the first church in England to

1 Rolls Series The Anglo-Saxon Chronicle.

8 Rolls Series Chronicon monasterii de Abingdon Appendix 11. De abbatibus Abbendoniae.

I 5 2

LOMBARDIC ARCHITECTURE

exhibit this arrangement. For though the plan of the first cathedral of Canterbury
given by Willis shows a western apse, it is a gratuitous addition of his. Eadmer
makes no mention of it : " Finis ecclesiae ornabatur oratorio. Ad quod, quia
structura eius talis erat, non nisi per gradus cuiusvis patebat accessus," 1 and his
words simply mean that the church ended at the west in a chapel, which could

only be reached by several
steps.

The plan is of very
ancient origin. For in-
stance, the magnificent
Basilica Ulpia in Trajan's
Forum at Rome (112-114)
had a hemicycle at either
end. Montano 2 gives the
plan of a Pagan Roman
sepulchral edifice ending in
three apses, the principal
one of which is faced by a
corresponding one at the
opposite end of the struc-
ture (Fig. 554) ; and his
work contains other ex-
amples of buildings with
aisleless naves and two
apses facing one another.
Again, recent excavations
at Silchester have brought
to light the remains of a
civil basilica with a hemi-
cycle at either end. 3 And
so the architect of St.
Mary's at Abingdon would
not have had to go to dis-
tant lands in search of the
arrangement, for Silchester
is only just outside the con-
fines of Berkshire in which

553- Canterbury. Columns from Reculver in the Cathedral
Close (669).

Abingdon is situated.

In Christian buildings
it had been used as far back as the Vth century in the basilica of St. Reparatus
near Orl&insville in Algeria (324), when a new counter-apse was erected to contain
the tomb of the bishop of that name (475).* And later, when the vestibule (in
which an altar of St. Andrew had already been introduced) of the Imperial
Mausoleum near St. Peter's at Rome was transformed into the basilica of Sant"
Angelo, 5 the latter presented, as indeed it always had done, two apsidal ends

1 Rolls Series Gervasii Cant, opera historica. 2 Op. cit.

3 Archaeologia, Vol. LIII. Fox and St. John Hope, Excavations on the site of the Roman city of
Silchester, Hants, in 1892. * Gsell, op. cit.

6 De Rossi, Inscriptions christianae urbis Romae.

ANGLO-SAXON ARCHITECTURE

facing one another. It has been suggested 1 that it was s x

St. Boniface (f 755) who carried the plan to Germany, * >J| Y|x

but there is no evidence in existence to support the (

'rUo V A \^

idea.

BKIXWORTH CHURCH was built about 680, in the
time of Cuthbald, second abbot of Peterborough. 2 The
surviving parts of the original structure are the nave (now
forming the body of the church) and portions of the pres-
bytery and western porch (Figs. 555, 556). The internal
face of the west end (Fig. 557) contains high up a three-
light opening (a later addition), the arches of which are \^ ^/|
carried by baluster shafts with Pre-Lombardic cubical

capitals bevelled in the lower part, surmounted by an

- , - -ru u i .run.

abacus of barbarous form. Ine balusters do not follow the

characteristic type of those at Monkwearmouth and Jarrow,
but belong to a Roman type, and are the oldest specimens
of the kind (for those in St. Andrew's, Hexham, are only used for a decorative
purpose) employed in a church that I have discovered in England. The Pre-
Lombardic cubical capitals are also the earliest examples of their kind in the

v y

554. Rome. Plan of
sepulchral edifice. (From
Montana, " Li cinque tibri
tfarchiteltura.")

Fig. 555. Brixworth Church (about 680).

country. Balusters of this type must have made their appearance about the end of the
Xth century, and have come into fashion in the first part of the Xlth, as we find them

1 The Archaeological Journal, Deceml>er, 1906. Micklethwaite, Something about Saxon chunk building.

2 Sparke, op. cit. Htigonis Candidi coenobii Burgensis historia.

LOMBARDIC ARCHITECTURE

represented in English manuscripts just at that period. I may refer to one in the
British Museum (Cottonian MS., Claudius B. IV.), described in the catalogue, and I
think rightly, as rather earlier than the Norman Conquest, exhibiting specimens of
baluster shafts, bulbous capitals formed by truncated inverted pyramids, and arcading
with alternate round and triangular heads, thus showing how the activity in building
the time of Edgar, Canute, and Edward the Confessor, was reflected in the

illuminations of sacred volumes.

The walls of the porch were raised and the tower built some time after 870,

when the church suffered at the hands of the Danes the same fate which befell the

mother-church of Peter-
borough, but before the
Norman Conquest. The
former probably took place
during the peaceful reign
of Edgar (959-975), who
was such a benefactor to
Peterborough ; the latter,
in the course of the new
invasion of the Danes, who
in 1010 had burned the
neighbouring town of
Northampton, and made
it urgently necessary to
fortify Brixworth Church.

* * *

With the organization
of the Church in England
carried out by Archbishop
Theodore (668-690) and
Abbot Hadrian (669-708)
a Church which became
a national institution and
prepared the way for the
political unity of the coun-
try the direct action of
Rome through her mis-
sionaries came to an end, and the stream of Christian culture emanating from the
City was stayed till we come to the days of Lanfranc (1070-1089) and Anselm
(1092-1109). The Primatial Chair of Canterbury, which from the time of Augustine
had been filled, except for a few years under Deodatus (655-664), by Italians and a
Romanized Greek, received no more foreigners till 1070. And the same was the
case with York, occupied till now, with the exception of Ceadda's and Bosa's
tenure of the see, by an Italian, Paulinus (627-633), and an Anglo-Saxon educated on
Italian lines, Wilfrid.

The withdrawal of direct Italian influence involved the cessation of the activity
of the French and Italian builders and carvers, which had passed like a meteor over
Northumbria. There were no longer monks from Italy to superintend the con-

Fig. 556. Brixworth Church. East end (about 680).

ANGLO-SAXON ARCHITECTURE

'55

struction of ecclesiastical buildings. Architecture in England must now pursue its
own course with the local means at its disposal and, apparently, without external aid.
And this lasted till the time of Alfred the Great (871-901), when the relations of
England with the Continent once more became intimate, particularly with France,
whence the learned king drew most of the intellectual influences by which he strove
to raise the culture of his subjects from the low level at which he found it. These
relations were resumed afresh, and with more tangible results, in the time of
Dunstan (943-988), when a new current of foreign artistic influence passed over
England. A similar current was felt in the reign of Ethelred II (978 or 979-1016)
and Canute (1014-1035 or 1036), and becoming stronger and stronger, and finally
carrying all before it, resulted in the
appearance under Edward the Con-
fessor (1041 or 1042-1066) of the
"New style" of Architecture, viz.
the Lombardo-Norman. Its intro-
duction did something to vivify the
inert ecclesiastical architecture of
England with its barbaric orna-
mentation, but at the same time it
sounded its knell ; for the moment
had come for the old, uncomely
forms to make way for a new crea-
tion of youth and vigour, backed up
as it was by the strong hand.

However, we must not antici-
pate our conclusion ; and first, let
us take a rapid historical survey of
three dark centuries barely illumi-
nated as it were by the spasmodic,
flickering light of an aurora borealis,
so that we may take stock of the
conditions existing in these lands,
and set out the evidence relating to
their ecclesiastical monuments which
is to be found in the annalists, and
contains matter of interest for our
purpose.

In the Vlllth century the history of England is a tale of perpetual wars for
conquest and supremacy between the various states into which the nation was
divided. Such was the story of intestine struggles in the kingdom of Wessex, which
compelled Ine (688-728) to abandon his throne and go on a pilgrimage to Rome,
where he ended his days ; and this in spite of his successes in the field, the civil
organization which he devised for his subjects, and the religious awakening brought
about by Aldhelm, bishop of Sherborne (705-709). The result was that Wessex fell
into the power of Ethelbald, king of Mercia (716-755). In the kingdom of North-
umbria, again, Eadbert (737 or 738-757), resigning the sceptre to his son Oswulf
(757). after the example of his predecessor Ceolwulf (731-737), retired to the monas-
tery of Lindisfarne, while the country became the scene of discord, revolt, and
slaughter ; evils accentuated by fire, pestilence, famine, and, to crown all, by the

!""' SS7- Krixworth Church. West end (about 680).

i S 6

LOMBARD1C ARCHITECTURE

Danish inroads described by Simeon of Durham. 1 A state of things like this was not
exactly favourable for the practice of architecture.

Thse disorders were brought to an end by the Mercian supremacy under
Ethelbald, Offa (755-794 or 796), and Cenwulf (796-819 or 822), though, at the same
time, it fell to pieces through the intervention of the Frankish kings in the affairs of
England. Indeed, with the establishment of Eardwulf on the Northumbrian throne
(795-806) and of Egbert on that of Wessex (800-836), and the outbreak of civil war
in Mercia itself after the death of Cenwulf, the supremacy broke up, and the extensive
kingdom which he left at his death was divided.

The supremacy of Mercia was succeeded in the IXth century by that of Wessex
under Egbert, brought about by the submission of Mercia and Northumbria, and his
conquest of Cornwall. But it was not of long duration. Even before his accession,
and that of his successor Ethelwulf (836-855 or 856), the Danes had begun their raids,
in alliance with the Britons of the West. Ethelwulf could not do more than repress
these raids for a time, for in 866 the pirates, after wintering in East Anglia, first
invaded Northumbria, next subdued the eastern kingdom and put to death its king
Edmund (870), then the lands of Mercia, and lastly Wessex, where, however, they
were compelled by the sword of the liberator, Alfred the Great (871-901), to come to
terms at Wedmore (878). The last years of the IXth century were spent in more
fighting between Alfred and the Danes, and the Xth opened with the determined
attempt of his daughter Ethelfleda (910-918 or 922) to subdue the Danish Confedera-
tion, the " Denlag " or " Danelaw," which had taken the place of the old kingdom of
Mercia and been recognized by the Peace of Wedmore. This object was achieved
by Edward the Elder (901-925), who received the voluntary submission of Scotland,
Northumbria, and Wales, and of the Britons in Strathclyde. Then came the wars of
Athelstan (925-940) with the confederate kingdoms and with the Danes, brought
to an end by his victory at Brunanburh (937), which initiated an era of peace in a
reign made illustrious by the courage, firmness, and wisdom of the monarch. There
followed Edmund's (940-946) reconquest of the Danish Confederation which had
taken up arms against him, and Edred's (946-955) subjugation of Northumbria.

After the misgovernment of Ed wy (955-959), a long period of peace succeeded
in the days of Edgar (959-975). Under his rule, and with the co-operation of
Dunstan, Ethelwold, and Oswald (who had been a monk at Fleury), the triad which
was the source of the splendour of the reign, his people saw the restoration of order
and justice, the promotion of trade and agriculture, the multiplication of abbeys,
instituted not only for monastic purposes but also as places of education which, in
spite of the impulse given by Alfred the Great, was well-nigh extinct. The century
ended amidst famine, internal conflicts, wars between the states, and fresh Danish
invasions, with their attendant rapine, fire, and slaughter, which marked the reigns of
Edward the Martyr (975-978 or 979) and Ethelred II (978 or 979-1016).

The Xlth century opened with Sweyn's revenge for the massacre of St. Brice's
day (1002) ; and this was followed by repeated invasions of the Danes and finally by
their conquest of England (1013), when the barbarian chief assumed the crown, and
Ethelred took refuge in Normandy. Recalled in 1014, the latter made various efforts
to recover the sceptre of which he had shown himself so unworthy, and his son
Edmund (1016) took measures to oppose the power of Canute (1014-1035 or 1036) ;
but the Dane, after the battle of Assandun and the death of his rival, obtained a firm
seat on the throne, and in the course of a peaceful reign conferred such benefits on his

1 Rolls Series Hist. Regum.

ANGLO-SAXON ARCHITECTURE 157

subjects that Edgar's work was almost forgotten. He well deserved his title of Great.
The disturbed reigns of Harold I (1035 or 1037-1039 or 1040) and Hardecanute
(1039 or 1040-1041 or 1042) followed, and then we come to the time of Edward the
Confessor (1041 or 1042-1066) and the end of the period.

When we draw up the list of ecclesiastical buildings which we know were
founded, rebuilt, or restored during the epoch which we have just outlined, we find
that the majority of these belong (i) to the long reign of Ine (688-728) ; (2) to that
of Edgar (959-975), the founder or rebuilder of forty-four monasteries, 1 mainly
through the instrumentality of Dunstan (943-988), Ethelwold (963-984), and Oswald
(961-992) ; and (3) to the reigns of Canute and the Confessor, as William of Malmes-
bury testifies. 2 This is a perfectly natural result, the history of art exactly
coinciding with the social and political history, and even more with that of religion.

Of these buildings some were merely wooden structures, like the monastery church
of Doulting in Somersetshire, where Aldhelm ended his life. 3 Others were built of
stone, and in some cases were so notable in that age and country that William of
Malmesbury describes the church of his own abbey, St. Mary's (whether the original
church erected by Aldhelm himself, as we learn from the " Gesta Pontificum," or a
reconstruction by Aelfric, who became abbot in 970), as " vincens decore et magni-
tudine quicquid usquam ecclesiarum antiquitus factum visebatur in Anglia" ; 4 while
Alcuin 6 is lost in admiration of the new cathedral of York built by Archbishop
Albert (766-782) under the direction of Eanbald and Alcuin himself, to replace the
old one burned in 741. 6

With regard to the form of some of these churches and their architects very
instructive information has comedown to us. We know, for instance, that Winchester
Cathedral, founded in 639, at the desire of Cynegils King of Wessex (61 1-642 or 643)
and with the permission of his son Cenwalh (642 or 643-672), by Bishop Birinus,
who had been sent by Honorius I to convert the West-Saxons, 7 8 was in 863 reached
through an atrium which had a tower rising from the middle of the side parallel to
the front of the church.

Turn's erat rostrata t/wlis, quia maxima quaedam,
Illius ante sacri pulcherrima limina tempi i,
Eiusdem sacrata Deo sub honore hierarchi.
Inter quam templique sacram pernobilis aulam
Corpore vir Domini sanctus requievit /tumatus, 9

Whether this tower went back to the time of Birinus or was erected later, we do
not know. In connection with it I would remark that in the Old St. Peter's at
Rome the fagade of the atrium showed an entrance flanked by two towers, the
first built by Stephen II (752-757), who gave it three bells, the second by Hadrian I

(772-795)- 10 u

1 Sparke, op. cit. Chronicon Johannis abb. S. 1'etri de Btirgo.

2 Rolls Series Gesta regiun Angloniw.

3 4 Rolls Series Willelmi Malm, moiiachi gesta pontificnm Anghrum.

5 Kolls Series The historians of the Church of York and its archbishops De pontificibtis et sand is
Ecclesiae Eboracensis carmen.

6 Rolls Series Symeonis monachi Ilistoria Regain.

' Rolls Series Annales monasterii lie Wintonia. * Rolls Series The Anglo-Saxon Chronicle.

a Mabillon, Acta Sanct. Ord. S. Benedict i De S. Swithuno eipiscopo \Vintoniensi, eiusi/ue transtationc et
miraculis. * Duchesne, Le liber pontificalis.

11 Plan by Alfarano in the Archivio Capitolare of St. Peter's.

I 5 8

LOMBARDIC ARCHITECTURE

It is also stated that the little church of the monastery at Athelney, founded
by Alfred the Great (871-901), was built in a new style (" Fecitque ecclesiam,
situ quidem pro angustia spatii modicam, sed novo edificandi modo compactam " : )
and with a plan somewhat like that of the church at Germigny des Pres
(801-806): " quattuor enim postes solo infix! totam suspendunt machinam, quattuor
cancellis opere sperico in circuitu ductis." * Its architect, too, is known to us, for
there can be little doubt that the John, presbyter and monk, a native of Old
Saxony, who was invited from his convent of Corbie by Alfred to become abbot
of Athelney, may reasonably be regarded as responsible for the building, knowing
as we do that he was " in omnibus disciplinis litteratoriae artis eruditissimus et in
multis aliis artibus artificiosus." 3 I believe, in default of proof to the contrary, that
this church marks the introduction of the central cupola plan into England, for we
must imagine that the four isolated supports carried a square tower over the crossing.

Again, we know that Ramsey Abbey Church, founded in 969 by Oswald, bishop
of Worcester (961-992) and archbishop of York (972-992), with the assistance of Earl
Ailwin, and consecrated in 974, was of cruciform plan with two towers, one over the
crossing and the other at the west end. " Duae quoque turres ipsis tectorum
culminibus eminebant, quarum minor versus occidentem in fronte basilicae pulchrum
intrantibus insulam a longe spectaculum praebebat, maior vero in quadrifidae
structurae medio columnas quatuor, porrectis de alia ad aliam arcubus sibi invicem
connexas, ne laxe defluerent, deprimebat." 4 5 Oswald himself was the architect of
the building, the idea of which he may have derived from the church of Germigny
des Pres, situated only a few miles from the convent of Fleury at Saint Benoit sur
Loire, with which Ramsey Abbey was closely connected for several centuries. Fleury
was a centre of instruction in the liberal arts, as the Ramsey Chronicle tells us :
. . . virum nominatissimum Abbonem, qui liberalium artium notitiam imis hauserat
medullis, de coenobio Floriacensi evocatum." Now Theodulf's church, besides the
existing central tower, had another, used for the bells, rising above the porch at the
entrance, thus described in an account printed by Baluzius : " Porro in matherio
(read ' narthecio ' or ' atrio ') turn's de qua signa pendebant, huiuscemodi inseruit
versus argenteo colore expresses."

Haec in onore Dei Tlieodulfus templa sacravi,
Quae dum quisquis ades oro memento met. 6

The church at Saint Riquier (Centula) (793-798) also possessed a tower in front
of the apse, and another between the church and its narthex. Again, St. Remy at
Rheims, as rebuilt by Archbishop Turpin (756-802) and finished by Hincmar in 852,
had a large tower at its western end. The western tower of the church at Blandigny,
consecrated in 979, 7 may have been suggested by the examples at Saint Riquier
and Rheims.

Soon after the central tower of Ramsey was finished it threatened to fall, and
had to be reconstructed. This was carried out by Abbot Eadnoth the younger under
the advice of Oswald, and the whole church was rebuilt in 991. Ramsey had an
aisleless nave. Aisled churches with central towers, as has been pointed out, 8 did not

1 2 Roth Series Willeliui Malm, monachi gesta pontificum Anglorum.

8 Mabillon, Acta Sanct. Ord. S. Benedicti Venerabilis Johannis abbatis Aethelingiensis elogittm historicitm.

4 Rolls Series Chronicon abbatiae Rameseiensis.

6 Rolls Series The historians of the Church of York and its archbishops Vita Oswaldi archiep. Eboracensis.

6 Miscellanea Catalogus abbatum Floriacensiuin. "> Afon. Germ. Hist. Annales Blandinienses.

8 The Archaeological Journal, December, 1906 Micklethwaite, Something about Saxon church building.

ANGLO-SAXON ARCHITECTURE 159

make their appearance in Great Britain before the time of Edward the Confessor.
The example of Winchester brought forward by Prof. Willis (in the volume on
Winchester, published by the Archaeological Institute) and others is imaginary.
The rebuilding of the cathedral was begun by Ethelwold during his tenure of the
see (963-984), and the dedication took place in 980 ; 1 2 but apparently it was
finished by his successor Alphege (984-1005), as may be gathered from a letter
addressed to the latter by the monk Wulstan. 3 All we know about it is that it had
a tower and atrium with chapels on its north and south sides : we hear nothing
about a nave with aisles.

As Athelney is the first recorded cruciform church with a central tower in
Great Britain, so Ramsey is the earliest recorded example of a western tower.
Nevertheless, the arrangement of a tower in the middle of the west front may be
traced back in England to the reign of Edward the Elder (901-925), for on a coin of
his time struck- by one Wlfgar there appears what seems to be an aisleless church with
a frontal tower rather higher than the nave (Fig. 558). And here I may say that

Kig. 558. British Museum. Coin Fig. 559. British Museum. Coin Fig. 560. British Museum. Coin
of Edward the Elder (901-925). of Edward the Elder (901-925). of Edward the Elder (901-925).

towers are frequently represented on Edward's coins, which is to be explained by the
fact that he and his sister Ethelfleda backed up their operations against the Danes
by the construction of strongholds in many places (Figs. 559, 560).

Nor were these the only churches erected in an imported foreign style during the
period we are discussing the style introduced into the Prankish Empire in the days
of Charles the Great. The original church at Abingdon (675) having been seriously
damaged by the Danes in the IXth century, Athelstan gave orders for its recon-
struction, and this was carried out under Edgar (959-975) by Ethelwold, who had
been a monk at Glastonbury, and was now abbot of Abingdon : " Erat namque
Atheluuoldus magnus aedificator" 4 "tot et tanta monasteria fecit quod vix modo
credibile videatur." 5 The new church of St. Mary at Abingdon is thus described :
"Cancellus rotundus erat, ecclesia et rotunda duplicem habens longitudinem quam
cancellus ; turris quoque rotunda erat." 8 This implies that it was a round church,
with an apse and a round central tower.

The abbey church of Exeter (a cathedral after 1050) as rebuilt by Canute in
1019, in place of the one erected by Athelstan and destroyed by the Danes in 1003,
possessed not only a central tower, but also two others which flanked the west front.
The evidence for this is the reverse of a seal of the old Chapter of Exeter attached to
a document of H33- 7 This church of SS. Mary and Peter is the first historical
instance in England of a pair of western towers. I have found another and nearly

1 Wharton, Anglia Sacra T. Rudl>orne, Historia maior Wintoniensis.

2 3 Mabillon, Acta San ft. Ord. S. BencdtctiVita S. Ethelwoldi episcopi.

* Rolls Series Chronicon monasterii de Abingdon Appendix I. Vita S. Aethcltvoldi.
5 Rolls Series Willclmi Malm, monachi gesta regain Anglorum.

* Rolls Series Chronicon monasterii de Abindon Appendix If. De abbatibus Abbendoniae.
7 Oliver, Lives of the Bishops of Exeter.

i6o

LOMBARDIC ARCHITECTURE

contemporary representation of a church front with two towers in the " Poems of
Caedmon," 1 assuming that the manuscript belongs to the first half of the Xlth
century. The adoption of this arrangement was due to the influence of the Lombardo-
Norman style which had by this time made its appearance in William of Volpiano's
church at Bernay (1013). His employment of it goes back to the erection of
Saint Benigne at Dijon (1002-1018), where we know that a pair of staircases were
formed in the wall of the west front (the " pariles scalas " of the Abbey Chronicle)
communicating with the galleries of the church. And William, in his turn, derived
it from the mother church of Cluny, dedicated in 982.

This influence, showing itself in the form selected for the church at Exeter, is
explained, as is the case with all the English buildings erected between 1000 and the
Norman Conquest (1066), by the consequences resulting from the marriages of
Ethelred II (1012) and of Canute (1017) with Emma (1002-1052), the daughter of
Richard the Fearless, Duke of Normandy (943-996). These events opened the way for
the Normans to get a foothold in the island, and made possible its ultimate conquest.
Let us now pass in review the dated churches, wholly or partially preserved, or
at least known to us by descriptions or drawings, which have escaped not so much

the destructive hand of time as the
violence of human passions, the
rage for novelty which came in
with the Norman Conquest and the
invention of the Pointed style, and,
last but not least, all the crimes
committed in the name of " restor-
ation " during the last century.

ST. MICHAEL'S CHURCH, ST.
ALBANS, was built by Abbot Wulsin
about the year 95O. 2 It consisted
of a nave ending in a rectangular
chancel, and a tower at the west
end. All that is left of this, after
successive alterations and the de-
plorable ill-treatment inflicted under
the pretext of restoration, is the
remodelled nave (which has been
lengthened westwards, involving the
demolition of the tower) and the
chancel (Figs. 561, 562). The only
notable feature presented by the
church is the double splay of the
windows, the earliest dated instance
to be found in England. It is evi-
dent that this form of aperture, of Roman origin, as we saw when dealing with
the church at Bagnacavallo, was late in making its appearance in England.

CHURCH OF SS. MARY AND ETHELBURGA AT LYMINGE (KENT). Queen
Ethelburga's church (640), which had been practically destroyed by the

Fig. 561. St. Albans. Church of St. Michael,
(about 950).

Navt

1 Bodleian Library, Oxford.

Rolls Series Tho. WalsinghamGesta ablahini monasterii S. Albani.

ANGLO-SAXON ARCHITECTURE

161

Fig. 562. St. Albans. Church of St. Michael (about 950).

Danes (804), was rebuilt by Archbishop Dunstan in 965.' * What remains
of his work consists of the aisleless nave, altered on the north side by the
addition of an aisle, and with
a rebuilt west front ; and the
rectangular chancel (Fig. 563).
It is the earliest dated instance
in England of an undoubted
rectangular and not apsidal sanc-
tuary.

DURHAM CATHEDRAL. As
the original church (998) ("ho-
nesto nee parvo opere " 3 ) of
Bishop Aldhun (990-1019) has
given place to the existing cathe-
dral, the work of Bishop William
of St. Carilef( 1080- 1096), it would
not concern us further were it
not for the recent discovery in the foundations of the Chapter House, demolished
in 1796, of the heads of four sculptured grave-crosses. These are now to be seen
in the Cathedral Library, and show representations of, among other things, the
Crucifixion, Baptism, and the Agnus Dei. As we can date them with certainty,
they are very valuable for purposes of comparison, and this is why they are deserving

of our attention. We
give illustrations of
two of the fragments
which, in spite of the
rudeness of the orna-
mentation and the
elementary treatment
of the figures, are
nevertheless of great
interest both in their
general outline and in
the details, and pro-
vide us with definite
information about the
style of carving in an
important religious
centre of Northum-
bria (Figs. 564, 565).

These relics of
the primitive cathe-
dral recall another carving preserved in the Library, a portion of a figure cross
(Fig. 566) brought from St. Mary's, Gainford, built by Ecgred, bishop of Lindisfarne
(830-845), for the monastery which existed as early as 801, and given to Durham

1 Jenkins, A sketch of the life of St. Ethelburga the Queen.

1 G. Gilbert Scott, Essay on the History of English Church Architecture.

3 Kolls Series Symconis monachi hist. Eccl. Dunelm.

Fig. 563. Lyminge Church (965).

VOL. II

LOMBARDIC ARCHITECTURE

by Bishop Aldhun. 1 For though it presents analogies with the Durham fragments,
there are differences in the treatment of the drapery ; and this, together with its

Fig. 564. Durham. Cathedral Library.
Head of cross (Xth or Xlth Century).

Fig. 565. Durham. Cathedral Library.
Head of cross (Xth or Xlth Century).

ruder character, makes me think that the Gainford Cross belongs to the IXth

or Xth century, and before the reign of Edgar.

I would remark at this point that we
have reached the age of Dunstan without
having met with any dated English building
showing the characteristic long and short
work which we first noticed at Monkwear-
mouth (675). We may, then, reasonably infer
that this feature did not gain a new lease of
life and become the fashion till after the
erection of St. Michael's at St. Albans, and
of Lyminge Church. But it must not be
supposed that it was confined to the Anglo-
Saxon period, for it is well known that
there are instances of it in Norman times.

THE CATHEDRAL OF OXFORD. Christ
Church, Oxford, the old convent church of
St. Frideswide, originally founded by
Didanus and his daughter Frideswide about
the year 727, and burned in 1002, was re-
built by Ethelred the Unready (978 or 979-
1016) after IOO4. 23 Of the church of Didanus
and Frideswide nothing is left that can be
seen. 4 The remains of a three-apsed east
end which came to light in 1887, are to be
assigned to Ethelred's work (Fig 567). This plan is not found in England before

1 Rolls Series Symeonis monachi opera Historia RegumHistoria Ecd. Dunelm.Historia de S.
Cuthberto.

2 Rolls Series Willelmi Malm, gesla pontificum Anglorum.

3 Dugdale, of. cit.

1 The Archaeological Journal, December, 1906 Micklethwaite, Something about Saxon church building.

Fig. 566. Durham. Cathedral Library. Fragment
of cross from Gainford (IXth or Xth Century).

ANGLO-SAXON ARCHITECTURE

163

the Xlth century ; unless, indeed, the relics of the ancient church at Lindisfarne can
be referred to the time of Athelstan (925-940), a date which has still to be demon-
strated. Ethelred's masonry is coarse and irregular. Until the contrary has
been proved, we may say that his
church was the first to have the
three-apsed plan in England. It
was an arrangement of Pagan
Roman origin, as we saw when
discussing St. Mary's, Abingdon.

Shortly after the erection of
Ethelred's church at Oxford another
English structure came into exist-
ence, interesting for the material of
which it is constructed. This is the
chapel set up near Aungre (Chip-
ping Ongar) on the occasion of
Alwin's translation of the relics of
St. Edmund from London, which
took place in 1013, or perhaps not
until 1020, when Canute installed
Benedictine monks at Bury St. Ed-
munds and erected a stone church
there, consecrated in 1032, to re-
place the original wooden one. 1 2 3

The chapel still exists as the
nave of Greenstead Church (Fig. 568), its walls formed of oak trunks sawn in half

and set upright side
by side with the
plane surface in-
wards. It enables
us to realize what
English timber con-
struction was like in
the Anglo-Saxon
period.

ST. MARY'S
CHURCH, STOW
(LINCOLNSHIRE),
was built by Ead-
noth II, bishop of
Dorchester (1034-
1049), about the year
1040, with the assist-
ance of Earl Leofric
(f 1057) ar| d his wife Godiva. Remigius, bishop of Dorchester and Lincoln (1067-

5 6 7- Oxford. Cathedral. Frontal arch of lateral apse
of old church of St. Frideswide (Xlth Century).

Fig. 568. Greenstead Church (Xlth Century).

1 Baldwin Brown, Ecclesiastical architecture in England from the conversion of the Saxons to the Norman
Conquest.

Dugdale, op. cit. * Rolls Series Memorials of St. Edmund's Abbey.

M 2

164

LOMBARDIC ARCHITECTURE

Fig. 569. Stow Church. South side of transept and choir
(Xlth and Xllth Centuries).

1092), rebuilt it before 1076, that is
to say before his introduction of
Benedictine monks. 1 ~ Though there
is no documentary evidence, many 3
consider the upper limb of the cross
to be work of the Xllth century,
belonging to the time of Bishop
Alexander of Lincoln (1123-1148),
or perhaps erected after the fire
from which Stow is believed to have
suffered in 1 156. 4

The oldest portions of St.
Mary's are the result of three
separate operations. It is a cruci-
form church with central tower
(Fig. 569). The windows and
doors show long and short work
in the jambs. The imposing arch
with multiplied moulded archivolts
(Fig. 570) forming the communica-
tion between the aisleless nave

and the crossing, has bulbous bases like the striking examples in the Bodleian

" Caedmonis Paraphrasis Poetica " (where the capitals are also of the same form),

believed to have been executed later

than the epoch of 1000, but before

the Conquest. And there are other

instances in the " Liber Geneseos "

in the British Museum, to which we

have already referred, dated rather

before 1066.

Stow Church affords the earliest
dated examples in England of pro-
tuberant bulbous bases. The em-
ployment of this characteristic fea-
ture in the form of substantial roll
mouldings in the north and west
doors of the church, and its presence
in the oldest part of the crypt of
St. Servatius at Quedlinburg (936),
show how cautious we should be in
attributing buildings where it occurs
to the Anglo-Saxon period.

Further, it provides the earliest

1 Rolls Series WillelmiAlalm. gesta ponti-
ficum Anqlorutn.

' 2 Rolh Series Symeonis monachi hisloria
regum.

3 Journal R.I. B. A., Third Series, Vol. VI

Bilson, The beginnings of Gothic Architecture. F ; g- 570. Stow Church. Crossing and choir (Xlth and

4 Madox, History of the Exchequer. Xllth Centuries).

ANGLO-SAXON ARCHITECTURE

165

surviving dated instance in England of a compound arch with roll mouldings.
The design was of Norman origin, for in the Lombardo-Norman style ex-
tensive use was made of it. The somewhat uncouth manner in which it is
introduced at Stow is explained by the different artistic conditions of the two
countries.

THE CHAIEL OF THE TRINITY AT DEERHURST was built in 1056 by Duke
Odda, as we are told by an inscription in the Ashmolean Museum at Oxford, whither
it was removed in 1675. It consists of a rectangular space opening at the east into
a chancel of the same form, of which only portions remain. The surviving door on
the north side and the chancel arch
diminish in width towards the top
and have a hood mould over them.
The jambs are constructed with
long and short work (Fig. 571).

ST. GEORGE'S CHURCH,

KlRKDALE, NEAR KlRBY MOOR-

SIDE (YORKSHIRE), was rebuilt by
Orm in the reign of Edward the
Confessor, and when Tosti was
Earl of Northumbria (1055-1065),
as we are informed by an inscrip-
tion on either side of the sundial
inserted in the wall above the
south door.

The original structure con-
sisted of an aisleless nave with
rectangular chancel. On the
outside various carved fragments
from the ancient monastery of St.
Gregory may be observed, which
have been used in the building of
the church. The most important
is a gravestone built into the wall
near the ground on the north side,
with part of a cross surrounded by

scroll work. In spite of its decayed state we are still able to see that it is of high
quality, though the hand of the carver was not very sure. The intestinal treatment
of the interlacing induces us to ascribe the work to an English carver under the
influence of the Continental artists of the time of Benedict Biscop and Wilfrid, and
before the Danish ravages of 867. The evidence is too slight and uncertain to
support the suggestion that it is the gravestone of Oidilwald, King of Deira
{651-660); and, moreover, he was buried at Lastingham.

In the jambs of the west door are inserted two shafts with Pre-Lombardic cubical
capitals hollowed out at the angles and carrying two high, moulded impost blocks from
which spring the multiplied archivolts. It forms the prototype for doorways of this
kind in Great Britain, and its capitals are the earliest dated specimens in the country
of the Pre-Lombardic cubical type with chamfered angles. An older though only

Fig- 571. Deerhurst. Chapel (Xlth Century).

i66

LOMBARDIC ARCHITECTURE

approximately dated example exists at Brixworth in the capitals of the three-light
opening in the inner face of the west end.

THE ABBEY CHURCH OF ST. PETER, WESTMINSTER. The exact date of the
building which the Confessor intended to be the chief monument of his reign is un-
known. According to Mabillon l it was erected between 1060 and 1065. Freeman, 2
on the other hand, puts it between 1051 and 1065. Micklethwaite, 3 in his turn,
relying on the oldest description we possess of the church, contained in a life of the
king 4 written after the battle of Stamford Bridge (1066) and before the death of
Queen Edith (1043-1075), believes that at Edward's death only the eastern part of
the structure begun in 1055 was in existence. It appears to me that this date would
be confirmed by Edward's transfer of some of the property of Pershore to Westminster
between 1054 and 1056. The building will have been finished afterwards, i.e. some
time before 1150 ; so that the later description in another Life of Edward, dedicated

to Eleanor of Provence, wife of
Henry III (1216-1272), and
written about I245, 5 will refer to
the whole period between 1055
and 1150.

It is my belief that the
surest evidence as to the date
of Edward's work is to be found
in the Bayeux Tapestry. This
important relic, which I have
examined on several occasions,
can only belong to the reign of
William I, the chief figure in
the great drama of the Con-
quest of England, whose de-
fence and glorification are the
main purpose of all the scenes
therein unrolled. Moreover, it
must have been made in the time of Odo I, bishop of Bayeux (1050-1097), the
rebuilder of the cathedral to which the tapestry belonged, and in the nave of which it
was exhibited in past times, as we learn from an inventory of 1476. To be precise
then, it was made between the battle of Hastings (1066) and the consecration of the
cathedral. To a period practically contemporary with the battle, and anterior to the
death of the Conqueror (1087) belong the coats of mail worn by the figures, with the
sleeves only reaching to the elbow (Fig. 572), whereas soon after 1087 they were
made longer, so as to come down to the wrist, and at the same time wider than the
sleeves of 1100 to 1120. I derive this statement from a communication made to me
by the eminent authority, Professor Oman of Oxford, and from one of his works. 8
The numerous representations of buildings never show the pointed arch, the great
characteristic of the last third of the Xllth century, to which (contrary to the

1 Annales Ord. S. Benedicti. * Of. cit.

> The Archaeological Journal, 1894 Further notes on the abbey buildings at Westminster.

4 Rolls Series Lives of Edward the Confessor Vita Aeduuardi regis.

5 Rolls Series La estoire de Saint Aedvoard le Rei.

6 A History of the Art of War.

57 2 - Bayeux. Detail from the Tapestry (Xlth Century).

ANGLO-SAXON ARCHITECTURE

167

general opinion placing it between 1066 and 1080) the Tapestry has been assigned l
on the supposition that it was inspired by the " Roman de Rou " of Wace, without
taking account of the possibility of Master Wace having derived his ideas from
the facts recorded on the tapestry in Bayeux Cathedral of which he was for nineteen
years a prebendary. 2

The date thus established is confirmed by the pictorial representation, partly in
section and partly in elevation, of Westminster Abbey, which the tapestry contains
(Fig. 573). Though the figure of a man engaged in fixing a weather-cock as a finial
on the sanctuary roof is an allusion to the consecration, and while the central tower

IH^^^^^H _

Fig. 573. Bayeux Tapestry. Representation of the old church of Westminster (Xlth Century).

is reproduced with elaborate detail, there is no sign of the two western towers
mentioned in the Life of Edward dedicated to Queen Eleanor :

En miliu dresce une fur,
E deus en /runt del Occident.

So that we may infer that the church was unfinished when the tapestry was worked,
and this also explains why the author of the oldest Life makes no mention of the
western towers. The incomplete state in which the building was left seems to me
to be also shown by the liberal endowment, on a larger scale than that of Edward,
given by the Conqueror to the Abbey. 3

Edward's church was of cruciform plan, with a central tower, nave and aisles,
chapels in two stories projecting from the transepts, and an apsidal choir with
ambulatory. We learn the last detail from the fact that in 1220 the old Lady Chapel
was added at the east end of the choir, so that the latter must have been provided
with an ambulatory, if there was to be access to the chapel. All that is left under the
floor of the present presbytery is three bases of the compound piers of the choir,
which with their shallow mouldings recall those at Jumieges. Some idea of its con-
struction is given by the so-called Chapel of the Pyx in Westminster Abbey, with its
rude unraised cross vaulting. In this chapel the foliage capital of the wall pier on
the south side is work of the Xllth century.

The name of the architect has not come down to us, but we learn that the church

1 Marignan, IM tapisserie de Bayeux.

a Taylor, Master Wace, his Chronicle of the Norman Conquest from the Roman de Rou.

3 Rolls Series Willelmi Malm, gtsla tantificum Anglorum.

i68

LOMBARDIC ARCHITECTURE

was in a new style : " Ecclesiam aedificationis genere novo fecit." 1 Now, seeing that
a new architectural style is not born in a day, and that, after the erection of Ramsey
Abbey Church (969), the cruciform plan with aisleless nave was reproduced in
England with monotonous regularity, while the Latin cross plan with central tower
and aisled nave, of which the Normans had made a speciality, did not make its
appearance there until the building of the church at Westminster, it is quite certain
that it was from Normandy and from the Benedictine Order that the Confessor
derived the design of his building.

The ties uniting the last survivor of the race of Cerdic with the country and
the Order which had received him as a fugitive (1013) in the days of his boyhood,

with his mother Emma and
his brother Alfred, were too
strong for him not to in-
dulge his love of monas-
ticism by the erection of a
sanctuary which was to be
the expression of all that
was dearest to his heart,
and of his affection for
Normandy and the Bene-
dictine Order. To carry
out this design who would
be more fitted than some
monk of the School of
Fecamp, whose abbot, John
of Aglie, stood so high in
the favour of the English
king, or else some member
of the abbey of Bee which
the genius of Lanfranc was
at that moment rendering
so conspicuous ?

Westminster Abbey
was not only the first
church in England planned
as a Latin cross, with nave
and aisles and a tower
rising above the crossing ;
it was also the first exam-
ple of an apsidal choir sur-
rounded by an ambulatory.
We will now proceed to describe some well-known churches, about which we
possess incomplete or misleading historical notices, but which are still regarded by
universal consent as belonging to the Anglo-Saxon period. Our object is to endeavour
to date them within limits of greater precision, and, so far as may be, with certainty.

ST. WISTAN'S CHURCH, REPTON. In 850 the body of St. Wistan was
deposited in the church at Repton, but it was afterwards moved by Canute to

1 Rolls Series Willelmi Malm, gesta ponlificnm Anglormn.

I'"ig. 576. Repton. Crypt of the church (Xth or Xlth Century).

ANGLO-SAXON ARCHITECTURE

169

fig- 574- Kepton.

Evesham about IO34. 1 The erection
of the church is ascribed to the
Scotsman Diuma, bishop of Mercia,
one of the missionaries brought by
King Peada (656-657) from Lindis-
farne to help in the conversion of his
subjects. When Repton fell into the
hands of the Danes in 874? it must
have succumbed to the fate which
befell all Christian buildings in the
districts conquered by these bar-
barians, and there is every proba-
bility that the church dates from the
reign of Edgar (959-97S)- 3

Of the building thus assigned to
the Xth century there survive incor-
porated in the present church, though
not in their original condition, the
rectangular chancel with the crypt
beneath it, besides some traces of the
aisleless nave and transept. On the
exterior the original parts of the
chancel have a stringcourse support-
ing slim lesenas ending in curious
capitals consisting of inverted truncated pyramids with a rude necking (Fig. 574).

I remarked some of the same pattern in the
Bodleian " Caedmonis Paraphrasis Poetica."

Here I would note, if only in the interest
of facts, as against the statements of various
writers, that lesenas have nothing to do with
Germany, their origin being Italian, just as
their name is Italian ; a fact already noticed
by Hiibsch. 4 The truth is that, long before
their appearance in German lands, they had
been used, first of all by the Romans in the
manner that may be seen on the exterior of
the so-called Praetorium in Hadrian's Villa at
Tivoli (125-135) (Fig. 575), then by the archi-
tects of Ravenna, and thirdly by the Comacine
or Lombard gilds.

In the south porch of the church are
preserved two shafts which unquestionably
belong to the church. They have rude capi-
tals like those in the crypt. This crypt
(Fig. 576) has rough vaulting sustained by
pillars which swell out as if compressed by

East end of the church (Xth or Xlth
Century).

F '8- 575- Tivoli. Villa of Hadrian. I'raetorium
(125-I35)-

1 Rolls Scries Chronicon alibatiae de Evesham.

2 Rolli Series The Anglo-Saxon Chronicle.

3 Cox, Notes on the churches of Derbyshire.
* Of. fit.

1 7 o LOMBARDIC ARCHITECTURE

the spiral band which encircles them. They have bulbous bases, and rudimentary
capitals chamfered off to fit the shafts. The fact that the walls of the chancel
slightly impinge on the vaulting of the crypt has given rise to a suggestion that they
are not of the same date ; but the capitals mentioned above make this impossible.

Assuming the church and the crypt to be contemporary, let us proceed to the
question of date, which we can fix with approximate certainty by a process of
elimination.

Before the time of Augustine (597-605) there is no record in England of a
crypt with columns under a chancel. The crypt of Christ Church, Canterbury, was
copied from the one in St. Peter's at Rome, 1 that is to say it had an ambulatory
or corridor following the curve of the apse, from the centre of which started a
passage at right angles to the chord of the apse leading to the chamber over the
tomb of St. Peter. 23 It was on the model of the crypts of St. Peter's at Rome
and Christ Church, Canterbury, that those constructed by Wilfrid at Hexham and
Ripon were planned ; for they, too, had underground passages leading to a space
which communicated with the chamber containing the relics. But this does not
make it even remotely possible that the earliest church at Repton, which was
probably of wood, as it was erected under the influence of Lindisfarne, possessed
a crypt with aisles separated by columns. So that we are obliged to date it in the
days of Athelstan, after the battle of Brunanburh (937) ; or more probably in the
peaceful reign of Edgar, after Edmund (943) had broken the back of the fresh Danish
rebellion, but in the last years of the reign, inasmuch as St. Michael's at St. Albans
(about 950) and Dunstan's church at Lyminge (965) still exhibit a system of
absolutely plain wall surfaces.

The introduction into England of wall decoration by lesenas, and later by
arcading, or by a combination of arcading and lesenas, was due, I believe, to the
influence of buildings such as the abbey church of Gernrode (968), in which the
eastern apse is decorated with a range of pilasters and another of engaged columns,
while the western towers are embellished with an arcade of alternate round and
triangular-headed arches. The marriage of the devout Edith () 947), daughter of
Edward the Elder (901-925), with Otto the Great (936-973) must have brought
the Anglo-Saxon clergy into communication with Old Saxony and its monuments ;
and it is thence that they may have derived the idea of such decorative motives
rather than from Italy (though it remains true that it was in Italy that lesena and
arcade decoration was created), seeing that, though triangular-headed arcading appears
in Italian carved representations, it is very rare to find it used there in the decoration
of buildings. Moreover, the capitals at Repton of inverted, truncated, pyramid form,
or with barbarous, rude mouldings, and also the bulbous bases and spiral columns,
suggest German influence derived from the crypts of St. Wipertus (936) and St.
Servatius at Quedlinburg (936). However, it is certain that, so far as arcading is
concerned, if we confine ourselves to existing dated monuments, on the one hand,
England has no architectural decoration of this kind to show before the close of
Edgar's long reign, while, on the other, it made lavish use of it before the Norman
invasion. And so, English buildings which show this treatment are to be dated
between 965 and 1066; and as the disastrous reign of Ethelred II was anything
but favourable to architectural development, we may reasonably suppose that such

1 Rolls Series Gervasii Cantuariensis opera historica.

Rohault de Fleury, La Afesse, Etudes arckeologiques stir ses monuments.

3 De Rossi, Inscriptiones christianae urbis Romae.

ANGLO-SAXON ARCHITECTURE

embellishments began to be used under Edgar, and that they were afterwards
elaborated in the times of Canute and of the Confessor.

We will conclude by remarking that the church and crypt of Repton were
certainly built before 1034, the year in which Canute removed the relics of St.
Wistan, for we know that crypts were constructed on purpose to receive the bodies of
saints. And therefore the date must be fixed some time in the second half of the
Xth, or at latest in the first years of the next century.

THE ABBEY CHURCH OF ST. MARY, DEERHURST. We know that the abbey
was certainly in existence in 804 ; that it was destroyed by the Danes ; that
Alphege, bishop of Win-
chester (984-1005) and
archbishop of Canterbury
(1005-101 1), there received
the monastic habit ; that it
was still an abbey in 1006 ;
that Edward the Confessor
deprived it of its posses-
sions in order to endow the
abbeys of Westminster and
Saint Denis at Paris ; and
that finally it became a cell
of the latter between 1054
and I056. 1234 It is also
stated that in 1016 Canute
and Edmund had a meet-
ing there, though there is
some disagreement among
the annalists about the
locality, the Anglo-Saxon
Chronicle placing it at
Olney, near Deerhurst. 5

The presence of Al-
phege at Deerhurst proves
that the abbey had been
restored after the damage
which it had suffered from
the Danes, and this prob-
ably took place in the reign
of Edgar, to which we may
assign the oldest portions of the church (though no longer in their original con-
dition), that is to say, the aisleless nave with two quasi-transept chapels, and a tower
porch. The suggestion that the whole was rebuilt by the Confessor, and consecrated
in 1056 6 cannot be entertained, for that date belongs to the chapel of the Trinity
which we have described above.

The nave terminated in an apsidal sanctuary, of which traces remain, andthough

1 Dugdale, op. cit. 3 Wharton, op. cit. Osbern, Vitas. Elphegi arch. Cantuariensis.

3 Butterworth, A short account of the ecclesiastical buildings at Deerhurst, Gloucestershire.

* Rolls Series Chronica magistri Rogeri de Houedene.

5 Rolls Series. Dugdale, op. cit.

F 'g- 577- Deerhurst Church (Xth Century).

172

LOMBARDIC ARCHITECTURE

F'g- 578. Deerhurst Church. Chancel arch (Xth Century).

these show a different style of
masonry from that of the rest of
the original structure (Fig. 577)
they are none the less contempo-
rary with it.

The rectangular western tower
had originally a double porch, of
which only the outer is left intact.
Here may be seen a sculptured
panel with two nimbed figures, ap-
parently the Virgin and Child,
under a canopy which seems to
be supported by three pillars with
stepped bases and capitals. There
is no evidence of a central tower
rising between the chapels, and
therefore we are unable to say
whether it was a case of a real or
only of a pseudo-transept. The
whole of the church was roofed
with timber.

The artistic details of the
church point to a single date and the hand of a single carver of low quality.
For instance, the hood mould round the sanctuary arch (Fig. 578), resting on
rude animal heads, is contem-
porary with a similar one in the
wall separating the two halves of
the porch, while the stepped sup-
ports of the canopy described
above, and the piers of the two-
light triangular - headed opening
on the inner face of the west end
(Fig. 579), are treated in precisely
the same way.

The lofty western tower, which
originally must have been still
higher as the bell-chamber has dis-
appeared, tells us that the church,
for reasons already stated, cannot
be earlier than the reign of Ed-
ward the Elder (901-925). But if
it had been erected at that date,
so much of it would not have sur-
vived as is the case, for then
it must have passed unscathed
through the struggles between the
Anglo-Saxons and Danes which
desolated Mercia up to 941 ; so
that we are obliged to come down Fig. 579. Deerhurst Church. West end (Xth Century).

ANGLO-SAXON ARCHITKCTURE

173

to the reign of Edgar (959-975), which saw a revival of religious zeal among the
Anglo-Saxons now that they were freed from the pagan Danish yoke, in order to
find a state of things favourable to its
reconstruction.

On the other hand, it is known that
the abbey was in existence in 1006, and
we cannot suppose that Edward the Con-
fessor rebuilt it before he impoverished
it. Besides, the masonry of his time in
Gloucestershire was very different, as the
walls of the chapel of the Trinity at Deer-
hurst tell us.

Deerhurst Church contains the earliest
English example of the hood mould of
an arch springing from heads, a feature of
which considerable use was made in the
Lombardic style, but of Etruscan origin.
Indeed, the Etruscans used projecting
heads not only on the imposts and key-
stones of arches, but even inserted them
in the spandrels of arches. Thus the " Porta
dell' Arco" at Volterra (Fig. 580), which,
though not in its original state, has not
changed sensibly from its original appear-
ance, 1 exhibits three heads in relief, one on
the keystone of the arch, the others on the
imposts. The gate known as the "Arco di Augusto" at Perugia has the remains
of two heads in the spandrels, while the Porta Marzia (Fig. 581) displays the

remains of heads of the Dioscuri
in the spandrels, and another,
thought to be that of a horse, on
the keystone of the arch. At
Faleri the " Porta di Giove " and
the "Porta di Bove" have key-
stones carved with the heads of
Jupiter and of a bull respective ly.

THE CHURCH OF ST.
LAUREiNCE, BRA DFORD-ON-
AVON. We learn from William
of Malmesbury 2 that in his time
there was standing at Brad ford -
on-Avon a small church said by
tradition to have been built in

honour of St. Laurence by Aldhelm, abbot of Malmesbury (680-705), Frome, and
Bradford, and bishop of Sherborne (705-709). " Et est ad hunc diem eo loci
ecclesiola, quam ad nomen beatissimi Laurentii fecisse predicatur." It is on this

Fig. 580. Volterra. I'orta dell' Arco.

Fig. 581. Perugia. Arch of the Porta Marzia.

1 Martha, L'art ttrusqtu.

2 Rolls Series Gcsta tontificum Anglorum.

174

LOMBARDIC ARCHITECTURE

Fig. 582. Braclford-on-Avon. Church of
St. Laurence. Chancel arch (Xlth
Century).

statement that the idea of so many writers, even some of the most recent, 1 is based,
that the existing structure is the work of the sainted abbot and bishop. In my

opinion, however, the account preserved by the
historian was only a legend. And we are not
the first to have doubts about the remote date
assigned to St. Laurence, for they are shared
by others. 2

The church, in spite of the injuries and
mutilation which it has suffered, is still, as a
whole (excepting the west front), such as it
was designed by its architect. It consists of
a rectangular nave with a chancel of the same
form attached to it (Fig. 582). The exterior is
decorated with lesenas, small clustered shafts,
and blank arcading (Fig. 583). This ornament-
ation is by some thought to be a later addition,
but any one who looks carefully at it will see
at once this is not the case.

The most remarkable feature is the blank
arcading. I have never come across any church
in East or West, of Aldhelm's age or earlier,
with this decorative treatment, continued, more-
over, round the front and the chancel. There-
fore we must conclude that Aldhelm, brought

up at Canterbury in the School of Abbot Hadrian (669-708) (" qui esset fons
litterarum, rivus artium " 3 ), cannot have learned from a foreigner, of African origin
but brought over
from a convent of
Campania, a new
form of architectural
decoration ; and for
the same reason he
cannot have ac-
quired it during his
visit to Rome.
Ranges of arches,
either blank, or open
and forming pas-
sages, were not em-
ployed in this ex-
tended form before
the Xlth century ;
and in England,
judging from dated

buildings, they do Fig. 583. Bradford-on-Avon. Church of St. Laurence. Chancel (Xlth Century).

not make their ap-
pearance till after the Norman Conquest. It is inconceivable that Aldhelm should

1 Hodgkin, The History of England from the earliest times to the Norman Conquest.
- Baldwin Brown, of. cit. * Rolls Series WilMmi Malm, gesia pontijicum Anglorum.

ANGLO-SAXON ARCHITECTURE

'75

have introduced them on his church, and so freely too, at so remote a date, and
that then the model should have remained for centuries a mere isolated phenomenon,
ignored, never copied, and presenting a striking contrast to the appearance of other
English churches. On the other hand, we cannot imagine that William of Malmes-
bury, who lived, roughly speaking, from 1 095 to 1 143, would have recorded the tradition
about the antiquity of the church if it had only just been rebuilt in his time.

Another important point to be noticed is the double splay of the windows, a
feature which, as we have remarked already, did not appear in England before the
time of Dunstan.

I believe that the only period in which we can date St. Laurence is that of
Edward the Confessor, under the influence of the artistic movement of the epoch of
1000, or, perhaps, with greater likelihood, the first years of the reign of the Conqueror,
at a time when Saint Etienne at Caen (1066-1086) was being designed with its double
encircling range of large blank arches and blank arcading of an elaborate nature.
This would explain the mixture of Anglo-Saxon (the plan and the doorways) and
Lombardo-Norman features (the blank arcading) in the church. Nor need we be
surprised that, after an interval of more than half a century, William of Malmesbury
should have recorded, in his " Gesta Pontificum Anglorum," finished in 1125, an
erroneous tradition about its origin ; for, as has been remarked, 1 fifty years do not
pass, even in our own
time, without the in-
vention, even on the
spot, and propagation
of false ideas about the
origin of buildings ; and
my own long and varied
experience in such mat-
ters entirely confirms
this opinion.

ST. JOHN'S CHURCH,
ESCOMB (DURHAM).
Though we have no in-
formation about the ori-
gin of this church, we
are not altogether igno-
rant of its history. We
know, for instance, that
"Ediscum," one of the
possessions of St. Cuthbert, was alienated by force, with other lands and churches,
from Bishop Aldhun (990-1019), but afterwards restored to him. 2 3 4

It consists of a nave with rectangular chancel (Fig. 584). The lofty, narrow
chancel arch, with long and short work in the jambs, should be noticed (Fig. 585).
In the chancel is a slab carved in high relief with a rude cross ornamented with
studs (Fig. 586), which recalls another on a gravestone in St. Andrew's, Auckland

1 Archaeological Journal, 1898 Micklethwaite, Some further notes on Saxon Churches.

2 Rolls Series Symeonis monachi hist, de S. Cuthberto.

3 Rolls Series Symeonis monachi hist. Eccl. Diinelm.

4 The Durham Liber Vitae (British Museum). Publications of the Surlees Society, Vol. XIII.

Fig. 584. Escomb Church (Xlth Century).

i 7 6

LOMBARDIC ARCHITECTURE

p'ig. 585. Escomb Church. Chancel arch (Xlth Century).

(alienated, like Escomb, with other
lands and churches belonging to
St. Cuthbert from Aldhun 1 2 ), more
elaborately treated, but of the same
type, and ornamented with pellets
in the upper corners of the field : a
style of decoration of which there
are but few examples in England.

Escomb Church is generally
regarded as of great antiquity ; for
instance, as belonging to the age
of Benedict Biscop and Bede, or of
about the year 800. We do not
share these views, but believe it to
have been erected after its restora-
tion to Aldhun, and very probably
after Canute had made his generous
donation to St. Cuthbert ; for it
does not seem possible that the
original structure, probably of
wood, like the cathedral at Ches-
' ter-le-Street before 104.2, can have
survived, almost untouched, the
dark days of the Danish invasions.
We must also take account of the
following reasons :
English church with a rectangular

(1) The earliest dated example of an

chancel which can be certainly instanced is SS. Mary and Ethelburga at
Lyminge (965). I believe that it was in the time of Dunstan that the plan

came into fashion. Nor need we wonder that the

square chancel, which was certainly not the plan
favoured by the Roman Church, obtained a firm
footing under Dunstan, if we remember the spirit of
independence which distinguished him, and even
carried him to the length of disobeying a Papal
order. 3 Its adoption may have been due to reasons
of expense. The builders of that age were obliged to
choose straight walls, which were easy and simple of
construction, in preference to curvilinear ones, which
require specially prepared materials and a higher
degree of skill. Or it may have been dictated by
the small dimensions of some churches, which were
too narrow to allow of the throwing out of an apse
large enough to contain the altar and provide room
for the free movement of the celebrant.

(2) The feature of long and short work, which

Fig. 586. Escomb Church.
(Pre-Conquest).

Carving

1 .Kails Series Symeonis monachi hist, de S. Cuthberto. * Rolls Series Symeonis monachihist. Eccl. Dunelm.
'> Rolls Series Memorials of St. Dunstan, arch, of Canterbury Epistola Adelardi ad Elfegiim arch, de
vita S. Dunstani.

ANGLO-SAXON ARCHITECTURE

'77

we first noticed in St. Peter's, Monk-
\\carmouth (675), does not prove, to
judge by existing remains, that its use
was general in England as a construc-
tive or decorative device, or both com-
bined, and introduced, either only in the
windows, or at the angles of the build-
ing as well, before the reign of Edgar
(959-975)- It was a feature which,
together with that of lesenas, appears
to me to have reached its highest ex-
pression as a form of decoration at the
end of the Anglo-Saxon period.

ST. MARY'S CHURCH, NORTON
(DURHAM). The Durham "Book of
Life" tells us that " Northtun " was
given, or rather given back, to St.
Cuthbert about the end of the Xth
century. In 1083 the church was made
collegiate, at the same time as Auck-
land and Darlington, in order to receive
the canons ejected by Bishop William
of St. Carilef from Durham.

Its plan was cruciform, with an aisleless

VOL. II

588. Norton Church (Xlth Century).

F'6- 587. Norton Church (Xlth Century).

nave and central tower (Figs. 587, 588).
The nave has been rebuilt on the old
lines, and the chancel reconstructed
on a larger scale. The eastern and
western arches of the crossing have
been rebuilt, and those on the north
and south have lost the innermost
archivolt. The only portions of the
original structure left are the transept
walls, those on the north being almost
intact, while the south arm has been
altered in quite recent times. The
tower is also original, though the
upper part is no longer in its primitive
state.

This is the only early cruciform
church with a central tower in
Northumbria. This feature shows that
it is later than Alfred the Great's
(871-901) church at Athelney, and
the church at Ramsey (969), that is to
say, the two earliest examples in
England ; while the remaining square-
headed window in the transept re-
minds one of Escomb. Lastly, the

i 7 8

LOMBARDIC ARCHITECTURE

outer archivolt of the tower arches recalls the chancel arch at Stow Church (about
-40), the plan and elevation of which suggest analogies with Norton. 1 We

regard it as a re-
construction, like Es-
comb, following the
donation of Canute ;
carried out, perhaps,
in the days of Bishop
Egelric of Durham
(1042-1056), who re-
built in stone the
primitive wooden
cathedral at Chester-
le-Street. 2

BOSHAM CHURCH
(SUSSEX) is repre-
sented, though only
in a conventional
way, on the Bayeux
Tapestry, with Harold
on the point of enter-

Fig. 589. Bayeux Tapestry. Representation of Bosham Church (Xlth

Century).

ing the door, followed by a travelling companion (Fig. 589). Originally it consisted
of an aisleless nave with a tower at the west end (Fig. 590), and at the east a chancel
which appears to have been of
rectangular shape. The nave walls
were afterwards cut through to
admit of aisles, and the end of the
chancel was pulled down in order
to increase its length.

The most interesting feature in
the church is the chancel arch (Fig.
591), with its half compound piers
surmounted by rude continuous
capitals of bulbous form, and a
common abacus, from which spring
multiplied archivolts. The rude
bases are formed of rolls and hollow
mouldings, and rest on a large roll
moulding common to all the mem-
bers, and the plinth.

Bosham Church must be rather
earlier than Harold's visit to Nor-
mandy (about 1064), and had, per-
haps, been recently erected, either
by Godwin (f 1053), the owner of
Bosham, or by Harold himself, when

he went there to pay his devotions. Kg. 59 a-Bosham Church. Tower (Xlth Century).

I n' * eli '? uar y> January, l8 9 4-Hodges, Norton St. Mary the Virgin's Church.

- The Reliquary, April, i8 94 -Hodges, Chtster-U-Street-St. Alary and St. Cuthberfs Church.

ANGLO-SAXON ARCHITECTURE

'79

The developed art shown in the
chancel arch, with the grouping of
the shafts at the sides, and the way
in which they correspond to the
archivolts above, and also the char-
acter of the capitals and bases as
compared with the original tower
arch at Stow, fix the date of the
church in the years after 1040, and
during the reign of the Confessor.

To his reign may also be as-
signed the original portions of the
nave and the square chancel at-
tached to it of Wittering Church
(Northants).

* *

With the evidence of the
churches which we have just ex-
amined, and with the light shed by
the proper comparisons, let us now
see whether we can find our way
among the obscurities of those which
are quite undated, though they pre-
sent features which have caused

F 'E- 591. Bosham Church. Chancel arch (Xlth Century).

them to be assigned to the Anglo-Saxon period. It will appear, I think, that the
meagre list of buildings of that age, which since the time of Rickman 1 has gone

on growing to excessive dimensions, will have to be re-
duced.

ST. MARY'S CHURCH, BREAMORE (HANTS), is of cruci-
form plan, with aisleless nave and rectangular chancel, but
has lost the north arm of the transept (Fig. 592). The form
of the quasi-transept, the arms of which are not open, but
closed like two chapels, access being given by an arch in
either case (Fig. 593), connects the church with those of
Deerhurst, Worth, St. Mary's Dover, and Repton ; while its
poor and meagre artistic features, confined nowadays to the
cable moulding on two impost courses, and the roughness
of the masonry (Fig. 594), relieved neither by arcading nor
lesenas, lead one to place it about the same date as the
church at Lyminge (965), and before Repton.

Approximately contemporary with it will be the nave
of Britford Church (Wilts), which has two arches in the
side walls, probably giving access originally to two lateral
chapels. The arch on the north side has the intrados
ornamented by a band with plain sunk panels at intervals,

Fig. 592. Breamore Church.
Plan (Xth Century).

Studies of the Stylts of English Architecture.

N 2

i8o

LOMBARDIC ARCHITECTURE

while the jambs are made up with fragments of carving which may belong to
any time between the VHIth and Xth centuries (Fig. 595).

Fig. 593. Breamore Church. Arch
of south arm of transept (Xlh
Century).

Fig. 595. Britford Church. Carving
in north opening (Vlllth - Xth
Centuries).

Fig. 594. Breamore Church (Xth Century).

THE TOWER OF BARNACK CHURCH (NORTHANTS). Of the original western
tower (Fig. 596) only two stages are left. They are of stepped outline, separated by

ANGLO-SAXON ARCHITEC TURK

181

a stringcourse, and striped vertically
by rude narrow lesenas of varying
dimensions. Some of the windows
have round, others triangular heads.
Two of them are filled with transeniiae
of interlacing circles. The west
window on the ground floor had,
apparently, a projecting figure (" pro-
tome ") above it, remains of which
may be seen.

The door, on the south side, with a
hood mould, recalls the original arches
which carry the tower of St. Mary's,
Norton. Above it is a window with
two birds in the spandrels, facing one
another, of barbarous design and exe-
cution ; antf at the top a circle with
rudely represented flowers. Three
carved slabs may be noticed built
into the north, west, and south sides
respectively of the upper stage. On
each is represented a tree with branches
of scroll work. The carving is in fairly
high relief and shows some vigour,
though of rude design. Above one
of the pieces is a cock, and in another

Fig- 597- Barnack Church. Tower arch (Xlth Century).

Fig- 596. Barnack Church. Tower (Xlth Century).

case a bird which is decayed beyond
identification.

The tower of Barnack belongs
to the same family as that of Earl's
Barton, but is an elder sister, as is
proved by its greater rudeness both
in construction and decoration. It
must have been erected after the
Danish ravages in Northamptonshire
in 1010, and very likely in the early
years of Canute's reign (1014-1035
or 1036), and after the building of
Repton Church, which, though it also
shows decorative treatment with plain
lesenas, has ruder masonry. And it
will come before Stow (about 1040),
for the tower arch (Fig. 597) is less
developed than the surviving original
one in that church.

ST. PETER'S CHURCH, BARTON-
ON-HUMHER (LlNCS), originally con-
sisted of a square tower (Figs. 598,
599), the ground floor of which served

182

LOMBARDIC ARCHITECTURE

Fig- 598. Barton-on-Humber. Tower of the church
(Xlth Century).

Fig. 600. Rome. Santa Pudenziana. Campanile (Xllth or
XHIth Century).

as a nave, and east and west of this
two projecting structures, the for-
mer, of rectangular shape, being the
chancel. This has now disappeared,
but remains of the foundations were
discovered recently. In the tower,
to which a stage has been added,
the two-light openings of the ori-
ginal parts have mid-wall baluster
shafts with pulvins. Those in the
added stage which have not been
rebuilt are divided by moulded
shafts a sort of combination of
the shaft and the baluster, like
those in the tower of Glentworth
Church (Lines) and in one case by
an ordinary shaft. The capitals
(two with a crocket at each angle)
are of Pre-Lombardic cubical type,
with each face ornamented by a
semicircle, and carry pulvins.

Apart from the singular plan
of the church, we may notice two
features, the baluster shafts and
pulvins, and the arcaded decoration,
as likely to throw some light on
the date which, contrary to the pre-
vailing opinion, has already been
put in the last part of the Xth cen-
tury. 1

With regard to the moulded
corbel pulvins, if we are to judge
by buildings of certain or approxi-
mately certain date, this impost
member did not make its appear-
ance in England before the reign
of the Confessor. And in their
earliest form they consist simply
of a flat upper face and a straight
chamfer, which is exactly what we
find at Barton-on-Humber. Later,
they were given a hollow chamfer
profile, and curl over at the ends,
as at Sompting and Jarrow, or take
other forms. If they occasionally
appear as mere square-edged ob-
long blocks, as at Worth, this must
be due to the incapacity of the

1 Baldwin Brown, op. cit.

ANGLO-SAXON ARCHITECTURE

183

workman or the economy of his
employer.

\Ye may remark here that,
though pulvins of Ravennate origin
served as models or suggestions for
the builders of other countries, cor-
bel pulvins of elongated form and
considerably flattened at the sides
did not appear for the first time in
the IXth century on the campanile
of Santa Pudenziana at Rome (Fig.
600), as is believed ; 1 for the Lom-
bardo-Roman bell-towers, derived
as they were from those of Lom-
bardy, the prototype being repre-
sented by San Satiro at Milan (876),
did not appear in Rome till about
the end of the Xlth century, and
the archetype is to be found in
the little church of Santa Maria
in Cappella (" Sancta Maria ad
Pineam") (Fig. 601), dedicated in
1090, as may be read in the well-
known inscription built into the

Fig. 601. Rome.

Santa Maria in Cappella.
(1090).

Fig. 599. Barton-on-Humber Church. East side of the
tower (Xlth Century).

inner face of the entrance wall. The
campanile of Santa Pudenziana is an
addition made in the time of Inno-
cent III (1198-1216), who restored
the church of Siricius (384-399).

The triangular and round-headed
arcading on the tower of Barton-on-
Humber, recalls and must be an echo
of that on the two western towers of
the abbey church of Gernrode (Xth
century). We have noticed similar
arcading, alternating with a range of
round arches, represented in one of
the Cottonian MSS. (Claudius, B. IV)
in the British Museum.

Hence, taking everything into
account, and remembering, as we
pointed out when discussing Repton
Church, that the earliest appearance
in England of the decorative use of
lesenas was not before the time of
Dunstan or the year 965 though it
must come before 1034, and consider-
1 Baldwin Brown, op. tit.

Campanile

I8 4

LOMBARDIC ARCHITECTURE

Fig. 602. Uracebridge Ct

Tower (Xllth Century).

its less barbarous and less irregular character
of building activity which marked
the time of Robert, forms a very
appropriate setting for the tower
of St. Michael's, in spite of its
display of Anglo-Saxon details
details which we know were per-
petuated by English builders even
after the Norman Conquest. For
Robert, besides contributing to the
reconstruction of Abingdon Abbey,
erected at Oxford the Castle (1071),
St. George's in the Castle (1074), a
great bridge to the north of the
city (1066-1087), ar >d rebuilt some
parish churches both within and
without the walls. 2 3

To return to Barton-on-H um-
ber, the addition to the height of
the tower must have been made
under Norman influence and after
1066, for it was only then that

ing that the architectural decoration
of the church shows an advance on
that of Repton, we shall not be far
wrong if we date Barton-on-Humber
Church later than Repton, and, to
be as exact as may be, in the first
years of Edward's reign (1041 or
1042-1066).

The baluster shafts and rude
pulvins in the tower carry one's
thoughts to St. Michael's, Oxford,
where the two-light openings in both
stories have balusters surmounted
by rude moulded pulvins. I believe
with Freeman l that the tower was
built by Robert d'Oily the Elder.
And, indeed, the original masonry,
external as well as internal, showing
as it does long and short work at
the angles, differs widely from the
remains of Ethelred II 's work in
Oxford Cathedral (after 1004) by
While, on the other hand, the era

1 Op. dt.

Kolls Series Ckronicon monasterii de
AbingJon.

3 Kolls Series Annales de Oseneia et
chronicon Thomas Wykes.

Fig. 603. Dover. St. Mary in the Castle (Xlth Century).

ANGLO-SAXON ARCHITECTURE

185

Corinthianesque crocket capitals, cubical crocket capitals, and cubical volute capitals
were produced in England.

These capitals require a few words of explanation. The reappearance of the
Corinthianesque capital with crocket leaves at the angles after the decadence of the
Dark Ages is to be ascribed to the revival which took place about 1000, when it was
introduced as a novelty in the crypt of Saint Pierre de la Couture at Le Mans (997).
It was not seen again in England (supposing that the Romans had introduced this
Etrusco-Roman type of capital) till after the Norman Conquest, and then for the
first time in the chapel of Durham Castle (1072), that is to say, if we are to trust
the evidence of existing and
dated English buildings. The
English cubical crocket capital
and cubical volute capital (i.e.
a cubical capital having at
the angles either the simple
head of a crocket leaf or a
double volute) are merely
simplifications and corrup-
tions of the Corinthianesque
crocket capital.

Some comment is also
demanded by the Lombardic
cubico-spherical capital which
made its first appearance in
Sant' Abondioat Como (1013-
1095), and also by the scal-
loped capital. Of the first
there is no trace in England
before the Norman Conquest :
the crypt of St. Augustine's,
Canterbury, affords the oldest
dated examples. The second
appears in its rudimentary
form in St. Albans Cathedral
(1077-1088), as we shall see
presently.

If we make use of these
two touchstones it will not be difficult to ascertain the true age of certain English
churches, regarded as pre-Conquest, but which really, until the contrary is proved,
must be held to belong to the post-Conquest period. Such are the following.

(i) The tower of St. Mary le Wigford, Lincoln, where the two-light windows
have small Pre-Lombardic cubical capitals chamfered at the angles, and others of
Corinthianesque form, recalling those in the smaller niches of the west front
of the cathedral, surmounted by pulvins. Freeman l was right in assigning it, with
the neighbouring church of St. Peter at Gowts, to the years between 1068 and 1086,
and to the direction of the Conqueror's favourite, Colesvegen, who, having laid out
the lower part of Lincoln across the Witham, was obliged to provide the new suburb
with places of worship.

1 Op. dt.

Fig. 604. Worth Church (Xlth Century).

i86

LOMBARDIC ARCHITECTURE

Fig. 605. Worth Church. Chancel (Xlth Century).

of the volute capital, sometimes with
the addition of a row of rude leaves,
betrays their Norman date. In con-
nection with Glentworth, a compari-
son has been made l with a capital
in the Castle church at Quedlinburg,
which must be the one with volutes
formed by the prolongation of the
interlacing bands and with pine
cones at the angles, of which we
shall give an illustration when we
come to deal with it. But the church
at Quedlinburg is not contemporary
with the Anglo-Saxon period : on
the contrary it is the result of re-
building after the great fire of 1070.
It is interesting to find at Glent-
worth the characteristic shaft par-
taking of the characters both of
column and baluster which we
noticed at Barton-on-Humber, for
it shows that this form of support

1 Baldwin Brown, op. cit.

(2) The tower at the west end
of Bracebridge church, near Lincoln
(Fig. 602), having in its highest stage
four tall two-light openings with
mid-wall shafts (one of which is
polygonal) bearing three scalloped
capitals (ornamented with zigzags,
stars, and studs) and corbel pulvins,
and one volute capital. This struc-
ture is certainly later than St.
Albans Cathedral, and not earlier
than the close of the Xlth century,
when the scalloped capital appeared
in its embellished and perfect form.

(3) The tower at the west end
of Branston church (Lines), the
ground floor of which is decorated
with blank arcading having scal-
loped capitals. For the date of
these what we have just said holds
good.

(4) The towers at the west ends
of the churches of Clee and Scartho,
near Grimsby, and that of Glent-
worth, in all of which the presence

Fig. 606. Earl's Barton Church. Tower (Xlth Century).

ANGLO-SAXON ARCHITECTURE

187

Fig. 607. Earl's Barton Church.
Outer door in tower (Xlth
Century).

(which, moreover, is rare) was employed in England
later than the Anglo-Saxon era.

(5) The tower at the west end of Great Hale
Church, near Sleaford, the Norman date of which is
proved by the occurrence of volute and scalloped
capitals.

THE CHURCH OK ST. MARY IN THE CASTLE,

DOVER, consists of a nave (Fig. 603) with a tower

rising above its eastern extremity, from the sides of

which two chapels project and give it a cruciform

shape. To the east of this is a rectangular chancel.

At the west end is a Roman lighthouse tower, which

has suffered considerably.

This church is one of the most important Anglo-
Saxon monuments that we possess. The fact of the

central tower puts it later than Athelney (871-901),

and the adaptation of the lighthouse as a western tower, the two being incorporated, 1

fixes the date as later than the erection of Ramsey Abbey Church (969-974), where

this adjunct first appeared in Eng-
land. Our choice, then, having to
be made in the period between
969-974 and 1066, we decide for
the time of Harold, who built a
new castle at Dover.* In any case,
it is certain that the church be-
longs to the Anglo-Saxon period,
for while it exhibits the charac-
teristic features of that period, e.g.
the aisleless nave with a tower, a
square chancel, the quasi-transept
forming chapels, and a tower at
the west end, it does not present
a single Norman one. The large
windows and plain brick arches
may be due to imitation of open-
ings in the Roman building from
which the bricks were taken.

WORTH CHURCH (SUSSEX)
forms a perfect Latin cross, with
aisleless nave and a deep chancel,
at the side of which rises a later
bell-tower (Figs. 604, 605). The

form of the piers of the chancel
Fig. 608. Wing Church (Xllh Century).

arch connects it with Bosham, and

the moulding of the impost course of the arch reveals Norman influence, which

1 The Archaeological Journal, 1896 Micklethwaite, Something about Saxon church building.
a Freeman, op. cii.

i88

LOMBARDIC ARCHITECTURE

appears again in the broad lesenas on the exterior. The date must therefore be
placed in the first years of the Conqueror's reign, or at earliest at the very end

Fig. 609. Wing Church (Xlth Century).

of the Confessor's, at a time when Westminster Abbey was in existence to
suggest new ideas. Other writers have also suggested the Xlth century. 1

THE TOWKR OF EARL'S BARTON
CHURCH (NORTHANTS) was the tower
at the west end of a church which has
been replaced by the existing one. It
is still in its original condition, except
for the embattled parapet (Figs. 606,
607).

Apart from the good masonry,
everything about it is of rude character,
from the lesenas, unequal both in size
and distribution, and the irregular
blank arch courses and round heads of
the windows, to the roughly worked
and squat balusters and the clumsily
carved crosses. The architectural deco-
ration is, however, so varied, and shows
such an effort after originality, though
of a rather childish character, that it
is effective.

Earl's Barton tower was the highest
expression of an infantile art, doomed
to disappear before the Lombardo-

Fig. 610. Sompting Church. Tower (Xllth Century).

1 Simpson, A History of Architectural Develop-
ment.

ANGLO-SAXON ARCHITECTURE

189

Norman style. It may be called the swan's song of Anglo-Saxon architecture, in
which some architect of the race combined in a sort of ill-ordered epitome all its
leading characteristics, in order to form a kind of record for posterity.

For there can be no doubt that it is to the end of the Anglo-Saxon or the dawn
of the Norman epoch that the tower must be assigned, as it is somewhat later than the
one at Barnack, which must have been its model, and displays an amount of architec-
tural decoration, though of a quite different character, which is only equalled by that
on St. Laurence at Brad-
ford -on -A von, with which
it must be contemporary.

WING CHURCH
(BUCKS) consists of a nave
with aisles, the former ter-
minated by a polygonal
chancel (Fig. 608), below
which is a crypt in the
form of an ambulatory,
added later. The Pre-
Lombardic capital in the
two-light opening above
the chancel arch, sur-
mounting, not a baluster
as at Brixworth, but a
shaft as at Kirkdale ( 1 05 5-
1065), brings it into rela-
tion with the latter. The
refinement of the supports
of the arcading round the
apse (Fig. 609), which take
the form of rolls instead
of lesenas, points to the
post-Saxon period and to
Norman influence, which
is further indicated by the
two tiers of windows in
the apse, and also by the
basilica plan of the church.
We know that, apart from a few of unusual importance, English churches of
the Xth century and of the Xlth before 1066 (and Wing must belong to one or the
other) had only aisleless naves and, possibly, a tower at the west end or over the
crossing.

The only Anglo-Saxon feature of the church is the triangular headed arcading ;
and therefore, on the whole, we are inclined to assign it, not so much to the last
years of the Confessor's reign, as to the time of the Conqueror.

The vertical rolls on the apse recall those on the tower of Sompting Church
(Sussex), the roof of which was altered in 1727 l (Fig. 610). Here each face is
bisected by a roll carried up to the highest point of the gable, and cut horizontally by a

1 Bloxam, The Principles of Gothic Ecclesiastical Architecture.

Kig. 61 1. Great Dunham Church (Xlth Century).

190

LOMBARDIC ARCHITECTURE

Fig. 612. Great Dunham Church. Tower (Xlth Century

ports, and the carving on the capitals, is
time from those at Stow, Bosham, and
Barnack, and by some interval from the
one at Langford. All this suggests that
the tower was built in the early part of
the Xllth century, in spite of the tri-
angular-headed windows, which, after all,
only show that English craftsmen had
not yet got rid of their traditional " im-
pedimenta."

The piece of carving inside the church,
with the Saviour in the act of blessing,
which is generally thought to be pre-
Conquest, must be assigned to a period
not earlier than the XHIth century, on
account of the form of the " vesica piscis "
containing the figure of the Redeemer,
and also the drapery of the figure.

The chief interest of the tower is in
the gable heads which determine the
form of the roof. It was a type of
German origin, which made its appear-
ance with the Lombardo-Rhenish style,
so that the oldest examples are not
earlier than the Xllth century. Its
nationality is proved by the fact that,
while it had a relatively wide vogue in

stringcourse with roughly moulded
billets ; while the windows consist of
single or two-light openings, with
round or triangular heads, and rude
corbel pulvins, crutch-shaped, hollow
chamfered, and curled at the ends.
The prototype of this kind of pulvin
is to be found at Mettlach (987).

Nevertheless, Sompting tower
does not belong to the Anglo-Saxon
age, as is generally believed. Against
that view are the semi-cylindrical
form of the lesenas and the course of
billets a Norman idea. It is further
opposed by the arch inside, springing
from half-piers consisting of square-
edged members and a half-column
corresponding to the similar members
and roll moulding of the archivolts ;
while the continuous capitals show
two coarse turn-over leaves and
cauliculi with berries. This arch, with
its rational disposition of the sup-
separated by a considerable interval of

v .

Fig. 613. Cambridge. Tower of St. Benet's Church

(Xlth Century).

ANGLO-SAXON ARCHITECTURE

191

Germany, it only made rare and
isolated appearances in other coun-
tries. Sompting is an instance of
its sporadic employment, due to
foreign importation. If it were not
so we should not find among
English buildings of the Xllth
and XHIth centuries, like an oasis
in the desert, a solitary specimen
of the " helmed tower," viz. the one
at Sompting.

GREAT DUNHAM CHURCH
(NORFOLK) has an aisleless nave
with a tower at its eastern end
(Figs. 611, 612), beyond which
was the chancel, rebuilt in later
times. The sides of the nave are
decorated internally with rough
blank arcading, the supports being
crowned by an abacus carved with
stars, lozenges, and indentations.
On the impost blocks of the arch
(which has a hood mould) opening
into the tower (original except the

Fig. 615. Langford Church. Tower archts (Xlth
Century).

Fig. 614. Cambridge. St. Benet's Church. Tower arch
(Xlth Century).

battlements) stars are carved in low
relief, a motive of Romano-British
tradition. Instances of the decora-
tive use of stars, either singly or in
groups, in carving have come to light
in excavations on the Roman Wall. 1
The west door has a triangular head
and two jambs ornamented with
billets.

Great Dunham church is a com-
pound of Anglo-Saxon and Norman
work. To the Saxon tradition belong
the disposition of the corner stones,
and the form of the west door ; while
the enrichment of the interior walls
of the nave with blank arcading is
due to Norman influence, for there is
no dated record of such decorative
treatment applied to the inside of an
English building before the Conquest.
To the same source are due the in-
dented and billet mouldings. More-
over, the two-light openings of the

1 The Builder, June II, 1898.

192

LOMBARDIC ARCHITECTURE

tower, the elongated form of the shafts, the Pre-Lombardic cubical capitals each

with its abacus and pulvin, and the bases consisting of two rolls (one of bulbous

outline) separated by a hollow cham-
fer resting on a plinth, all point to
the English craftsmen having been
brought into contact with the work
of their Continental brethren.

For these reasons the church
cannot be assigned to the Anglo-
Saxon period. Its place is in the
first years after the Norman Con-
quest, while the two-light windows
may even be the result of an alter-
ation carried out rather later, in the
last quarter of the Xlth century.
No weight need be given to the
fact that there is a triangular-
headed doorway, for this form con-
tinued to be used after the Con-
quest. An instance occurs in the
tower of Jarrow, erected between
1074 and 1083.

The tower of Great Dunham
with its rounded openings above
the two-light windows recalls that
of St. Benet's, Cambridge (Fig. 613),
which we should assign to a post-
Conquest date (in spite of the long

and short work at the angles), that is to say, to the time when the new town of

Cambridge was rising in the reign of

the Conqueror. The lions, for instance,

from which spring the outer archivolts of

its internal arch (Fig. 614), point to Lom-

bardic influence and a date which is not

Anglo-Saxon. For, though as early as

the first half of the Xlth century the

Lombard gilds made use of animals

flanking doorways to serve as supports,

as we saw in San Flaviano at Monte-

fiascone (1032), the English craftsmen

before the Conquest only employed for

this purpose heads projecting from the

wall, as we learn from the example of

Deerhurst church (Xth century).

Fig. 616. Langford Church. Tower (Xlth Century).

Fig. 617. Langford Church. Crucifix (Xlth Century;.

THE TOWER OF ST. MATTHEW'S
CHURCH, LANGFORD (OxoN), rests on
two arches, that looking towards the nave having piers and two archivolts springing
from an impost course, while the one on the chancel side has half-piers and archivolts

ANGLO-SAXON ARCHITECTURE

193

with a roll moulding (Fig. 615). It is lighted by single, double-splayed windows, and

by pairs of openings outlined by sunk rolls with foliated capitals at the imposts

(Fig. 616). It has been thought to

go back to the Anglo-Saxon age :

perhaps to the early years of the

Xlth century, or more probably

to the reign of the Confessor.

We cannot accept this view.
The arch, with its roll moulding,
while recalling that at Bosham
(which itself is subsequent to the
Anglo-Saxon arch at Stow), is
shown to be later by the more
intelligent way in which the sup-
ports correspond to the archivolts,
and by the more advanced art
displayed in the carving of the
continuous capitals. From another
point of view, the well-executed
foliage on the capitals of the two-
light openings is decisive against
an Anglo-Saxon date. We know
what sort of capitals were pro-
duced by English chisels or axes.
Nor is a pre-Conquest date indi-
cated by the " keyhole " windows,

for there is no dated example of

fig. 618. Romsey Church. Crucifix (Xllth Century),
that period to point to. Moreover,

the external facing, with the returned lesenas at the angles ending in flat imposts,
has an air of finish which is not Anglo-Saxon.

Langford tower must be put in the last quarter of the Xlth century, at a time
when, under Norman influence, English carvers were beginning to produce foliated
capitals. To the same date is to be assigned the headless crucifix now built into the
outer wall of the south porch, a figure of such rude character that, with its drapery, it
might be made of wood (Fig. 617). Some writers have compared it to the
crucifix at Romsey (Fig. 618). But this piece of sculpture, with its figure of the
Redeemer, so much advanced beyond the figure capitals in the choir executed in
the first half of the Xllth century, cannot possibly be earlier than the end of that
century.

VOL. II

CHAPTER IV

THE LOMBARDO-NORMAN STYLE IN ENGLAND

WHEN the works of Saint Etienne at Caen were begun, Lanfranc was
obliged to leave the scene of his achievements at Bee, and resign the
prior's place to his pupil, Anselm of Aosta (1060-1066), in order to
become abbot of the new monastery. The change was brought about
by Duke William (1035-1066), anxious to have near him, in his favourite city, a
trusty counsellor, capable of making clear and easy his way to the conquest of the
crown which on the death of Edward the Confessor had passed to Harold. That
conquest was the common work of these two great men. Normandy was too small
for the lion's grasp of the one, and the eagle's flight of the other.

Although it was the sword of the valiant, fearless warrior, William of Normandy,
which, on the field of Hastings (1066), decided the fate of England, winning for him
the title of Conqueror, and placing the long-coveted diadem on his head, nevertheless
it was the mind of Lanfranc, the soul and spring of the whole enterprise, that
moulded his conception of the invasion and conquest, correcting, supplementing,
bringing it to perfection. In all this he received invaluable aid from the Benedictine
monks and secular clergy of Normandy, and also the support of his pupil, Pope
Alexander II (1061-1073), and of Hildebrand, afterwards Gregory VII (1073-1086).
And after he became archbishop of Canterbury (1070-1089) he performed a
work of the highest importance in organising and consolidating the conquered
country.

It was with Lanfranc, too, that the golden age of Lombardo-Norman architecture
opened in England. And he was scarcely dead when William of St. Carilef began
the building of his great church at Durham, which was to mark the passing
of the Lombardo-Norman forms into those of the Transition, and so lead the way to
the Pointed style.

Lanfranc, uniting in himself the architect, the man of letters, the diplomat, the
statesman, was, it appears to me, the most important figure in the ecclesiastical world
of the Xlth century, with the exception of Hildebrand, who, however, was inferior
to him in the perfect balance of his faculties. Placed at the summit of the hierarchical
edifice, Gregory VII embraced in his view at once the vast horizon of the Catholic
Church and the kingdoms of this world, in which he was the principal factor. But he
did not know how to select the best ways and the most suitable means for carrying
out his boundless designs. His insatiable ambition and ill-concealed restlessness were
serious obstacles to his success ; so that, if Lanfranc had not possessed the will and the
skill to undo the mischief caused by the pontiff's lack of patience in dealing with
William, the cry for separation from Rome would not have had to wait till the XVIth

194

ENGLAND 195

century before it was raised. The triumph of Canossa would not have been repeated in
the case of the Norman lion : the Conqueror was formed in a very different mould from
the Emperor Henry IV. The last words said to have been uttered by Hildcbrand
" Dilexi iustitiam et odivi iniquitatem, propterea morior in exilio " words which are
preserved in the " Officium proprium Sancti Gregorii papae VII " of the Cathedral of
Salerno, sum up the whole of his work. It is the confession of a man who has been
beaten in the struggle against adverse fate.

THE ABBEY CHURCH OF ST. AUGUSTINE, CANTERBURY. Recent excavations
have revealed the scanty remains of the crypt of the church begun by the Norman
abbot, Scotlandus (1070-1087), after demolishing the new work of Wulfric (1047-
1059), and finished in 1091 by his successor Vido (IO87-IO99). 1 23

The body of the church was arcaded, and terminated in a semicircle surrounded
by an ambulatory with three radiating chapels. 4 The arches of the body of the
church were supported by piers, and the nave was separated from the aisles by
columns. The ambulatory had cross vaulting with visible arches.

To judge by what I saw in 1903, the capitals of the supports were of the
Lombardic cubical type ; and among the bases were some of bulbous form in the
Anglo-Saxon style, consisting of a plinth and a thick roll with a smaller roll above
it. Such bases indicate that English craftsmen were employed in the works. When
the first buildings in the Lombardo-Norman style were erected, the same thing must
have happened in England which had taken place previously in Normandy in the
days of William of Volpiano, as we explained in our account of the church
at Bernay ; that is to say, Anglo-Saxon workmen were put under the direction
of Norman workmen, who also undertook the more difficult tasks, such as the cross
vaulting. The artistic parts, e.g. the capitals, bases, stringcourses, cornices,
were generally left to English hands, and this explains the poverty of the
results.

As the plan of the crypt must have been repeated in the choir above,
St. Augustine's affords the earliest certain instance in England of a choir
ambulatory with radiating chapels. For though we know that Westminster Abbey
had an apsidal choir with surrounding aisle in the time of Edward the Confessor,
it does not follow that it had a system of radial chapels.

Related to this crypt is the one at Gloucester, as constructed by Serlo (1089).

Among the remains of the church at Canterbury, the presence of the Lombardic
type of cubical capital should be noticed, for this is the earliest dated English
building in which it occurs.

CANTERBURY CATHEDRAL. The primitive church of Augustine, restored and
increased in height by Odo (942-959), 50 and repaired by Canute after the Danes
in ion had set fire to the roof," was finally burned in the year 1067. Within seven
years (1070-1077) Archbishop Lanfranc rebuilt the cathedral. 8 Under his successor
Anselm (1096-1109), about the year 1096, Prior Ernulf, who had passed with

1 Rolls Series Elmham, Hisloria monaiterii S. Augustini Canltiariensis.

3 Mabillon, Annales Ord. S. Benedicti. 3 Twysden, op. fit. Chronologia Auguslinensis Cant.

* Routledge, Excavations at St. Austin's Abbey, Canterbury The Chunk e/SS. Peter and Paul.

6 Rolls Series The Historians of the Church of York and its Archbishops Vila Oswaldi arch. Eboracensis.
8 Wharton, op. cit. Osborne, Vita Odonis archiep. Cantuariensis.

7 Wharton, op. cit. Vita S. Elphegi archiep. Cant.

8 Rolls Series Eadmeri historia novorum in Anglia.

O 2

196

LOMBARDIC ARCHITECTURE

Lanfranc from Bee to Caen, and had thence been summoned to Christ Church,
Canterbury, demolished the eastern limb of the cathedral and began its recon-
struction. It was completed by Prior Conrad, who gave his name to the choir,
and the dedication followed in 1130.

In 1174 Conrad's choir, which had a painted wooden roof, was destroyed by fire.
It was rebuilt between 1175 and 1184 by the architects William of Sens, and

William " English by nation,"
who extended its length towards
the east, incorporated part of
the outer walls of the preceding
church in the new work (as may
readily be seen from Fig. 619),
and kept untouched only the
crypt with its two lateral chapels.
In the XlVth century the nave
and transept were rebuilt on the
old lines, and in the XVth the
great central tower was erected.
Gervase l tells us that Lan-
franc's church was of cruciform
plan with a central tower. The
arches of the nave were carried
on piers, and there were two
lofty towers at its western end.
Each arm of the transept had a
vaulted gallery or loggia, sup-
ported on three sides by the
outer walls, and on the fourth by
a pier. The choir was raised by
several steps above the level of
the rest of the church, and two
flights of stairs led down to the
crypt. The form of the choir
and crypt is unknown, for Ger-
vase did not see them, and was
not acquainted with any descrip-
tion. But his account of the
church, brief and imperfect as it
is, makes it clear that Lanfranc's
design was taken from his Saint
Etienne at Caen.

Conrad's choir, we learn from Gervase, with Ernulfs new crypt beneath it, was
of considerable length, with an apsidal end. It had arcades at the sides, with
a triforium above them, and was provided with a secondary transept. Christ Church,
Canterbury, is thus the first instance in England of a double transept. At the sides
of the choir were two lofty towers, known as St. Andrew's and St. Anselm's, and at
its extremity was a square chapel. This adjunct, and the similar one at Rochester,
were the first " Lady Chapels " in England.

1 Rolls Series Gervasii Cantiiariensis opera hislorica.

Fig. 619. Canterbury Cathedral. South side (Xlth and Xllth
Centuries).

ENGLAND

197

A drawing made by the monk Edwin gives a view of the exterior of the new
choir and the remains of Lanfranc's church.

Ernulf's crypt has come down to us almost untouched, and merely broken

Fig. 620. Canterbury Cathedral. Crypt. Chapel of St. John or St. Gabriel
(Xlth and Xllth Centuries).

into at its junction with the other part of the crypt belonging to the late Xllth
century structure. Several of its Lombardic cubical capitals, as well as of those in
the two lateral chapels (Figs. 620, 621), were carved and ornamented in various ways
in the course of the works carried
out after the disaster of 1174. This
may be inferred from the fact that
the capitals in Conrad's and Ernulf s
building were left plain "in capi-
tellis veteribus opus erat planum, in
novis sculptura subtilis," as Gervase
says ; and also from the advanced
stage of art displayed by the carv-
ings, which suits the period we have
suggested.

Before leaving the subject I
may remark that Edwin's view of
the church is of the same date as
the other illuminations in the Can-
terbury Psalter, 1 that is to say, of
the Xllth and not the Xth century.
This fact has not been previously
observed, and it reduces the age of

the Psalter, which has hitherto been F 'S- 621. Canterbury Cathedral. Crypt. Capital in Chapel

,. , of St. John or St. Gabriel (Xlth and Xllth Centuries),

studied only from the palaeographical

and not from the architectural point of view, or by anyone who was familiar with

mediaeval buildings. The Psalter contains numerous representations of structures

1 Tripartititm Psalteriuin Eadwmi in the Library of Trinity College, Cambridge.

198

LOMBARDIC ARCHITECTURE

which indicate a date later than the Norman Conquest. We will only cite the
instance of a basilica with a central cupola decorated with two tiers of blank arcading
separated by a band with lesenas at intervals. Now, it is known that, in England,
lantern towers were not embellished with such arcading before Lanfranc's rebuilding
of his cathedral, and that is why the central tower of Westminster Abbey was
devoid of it, as shown by the Bayeux Tapestry. There is also another view of a

Fig. 622. Durham. Chapel in the Castle (1072).

church with arcading of a refined type running below the line of the aisle roof, a
feature which appears on no dated English building earlier than the Conquest.

THE CHAPEL OF DURHAM CASTLE. Durham Castle was begun in 1072,
and the fruitless siege by the murderers of the unfortunate Bishop Walcher
(1071-1080) shows that it had been finished before 1080. The fact that the
canons of Waltham, whose church and lands had been given to the see of Durham
by the Conqueror (1075), contributed in the days of Rannulf Flambard an annual

ENGLAND

199

Fig. 623. Durham. Chapel in the Castle. Capital (1072).

payment towards the building of the
castle, as we learn from the ex-
emption which Queen Matilda ob-
tained for them, only means that
their subsidy went towards the im-
portant works executed by that
energetic and restless prelate to
increase the strength of both castle
and city.

The castle chapel is the one
described by the monk Laurence,
afterwards prior of Durham (1149-
1154) as " non spatiosa nimis, sed
speciosa satis." 143 It is of rect-
angular form, divided into three
aisles with unraised cross vaulting
(Fig. 622). The carving on the capitals of the columns is executed without under-

cutting, and the decorative elements

are sometimes well treated ; but
the representations of living crea-
tures are clumsy in design, and
rude in execution (Figs. 623, 624,
625). The chisels were, generally
speaking, Norman, as is shown by
the introduction of their character-
istic crocket capital. One of the
earliest instances in the North of
England of this type is to be found
in the crypt of Lastingham church,
rebuilt about 1078.*

The Corinthianesque crocket
capital, which has been wrongly

Fig. 624. Dur

apel in the Castle. Capital (1072).

described by some writers as
Byzantine, is of remote origin. In

the tomb of the Volumnii near Perugia is a marble cinerary urn in the form of a

temple, believed to be of the age of

Augustus, in which the capitals of

the pilasters have crocket leaves

at the angles instead of volutes

(Fig. 626). Earlier still, leaves of

this kind were used on the capitals

of the " Tomb of the Reliefs " at

1 Kolls Series Synieonis monachi hist.
Kegum.

- Rolls Series Symeonis monachi his/.
Eccl. Dunelm. continuatio.

3 Publications of the Surtees Society
Dialogi Laiirenlii Duiulmensis monachi ac
prioris.

4 Dugdale, op. cit.

Fig. 625. Durham. Chapel in the Castle. Capital (1072).

2OO

LOMBARDIC ARCHITECTURE

Fig. 626. Perugia. Tomb of the Volumnii.
(B.C. 29-A.D. 14).

Cinerary urn

Cerveteri, and on those of a
terracotta urn in the Museum
at Perugia, dated respectively
about the IVth century, and
the Ilnd or Illrd century
li.C.

English carvers worked
on the chapel at Durham as
well as Norman ones, and
this explains the low grade of
art displayed by some of the
results. I should, for instance,
assign to Northumbrian artists
the capitals with a grotesque
human head, a siren, a stag,
and a horse, all of which pre-
sent various mutual analogies,
while they are quite different
from others of the living
creatures represented, such as

the strange animal head with protruding tongue, and the human figures holding

up the abacus.

Anyone who compares the foliage of the best of these capitals with that in the

cathedral on the capitals of the in-
terior door near the south arm of the

transept, and on the four capitals of

the blank arcading in the nave aisles,

will be at once struck by the inferiority

of the artists who produced the foliated

capitals in the cathedral. Noting this

fact, and considering that the cubico-

spherical is the prevailing type in the

great church, we shall be safe in ex-
cluding any Norman artists from it,

whether monastic or lay.

The masonry of the chapel may

also be set down to Norman builders,

as is shown by the cross vaulting.

English masons were not yet capable

of carrying out such work.

ST. ALBANS CATHEDRAL,
founded as an abbey church in 793
by Offa, king of Mercia (755-794 or
796), was rebuilt between 1077 an d

1088 (With help from the primate Fig . 627 . _ St . Albans Cathedral. Nave(i077-ioS8).

Lanfranc) by Abbot Paul (1077-

1093), who had formerly been a monk of Saint Etienne at Caen. 12 Eadmer

1 Rolls Series Walsingham, Gesta abbatum monasterii S. Albani.
- Rolls Series Matthaei Parisiensis hist. Angl/trum.

ENGLAND

20 1

and Ralph de Diceto say that it was Lanfranc himself who carried out the
rebuilding. 1 -

Abbot John de Cella (1195-1214) pulled down the west front in order to make a
new one. The work on the new front was continued by his successor, William of
Trumpington (1214-1235), who altered the aisle windows in the new style and
remodelled the top of the central tower. Abbot George of Hertford took in hand the
reconstruction of the choir which, with its Lady Chapel, was completed in 1326 by
Abbot Hugh of Eversden (1308-1326). In 1323 two of the piers on the south side
of the nave collapsed, and five arches of the arcade had to be rebuilt. John of
Wheathamstead (1420-1440, 1451-
1464) made extensive changes.
The restorations of recent years
have to a considerable extent
altered the original features of the
church.

Of the Norman structure there
survive, in a more or less altered
state, the transept and central
tower, the junction between this
and the choir, and the body of the
nave except the west front.

The nave (Fig. 627) and tran-
sept (Fig. 628) retain their wooden
ceilings. The aisles, to judge by
what is left of the old choir, were
originally covered with rude un-
raised cross vaulting.

In the transept may be noticed
some baluster shafts which have
been used over again. They have
various mouldings, and are turned
on the lathe. Possibly they were
made in the second half of the
Xth century, in the days of Abbot
Eadmer, who cherished a design of

Fig. 628. St. Albans Cathedral. North arm of transept
(1077-1088).

rebuilding the church. 3 We may also observe some Lombardic cubical capitals with
two half-rounds on each face.

Though the design for St. Albans issued from the School of Lanfranc, English
craftsmen and workmen were employed in its erection. The masonry (Fig. 629),
except for a larger use of bricks, is just like that of the neighbouring St. Michael's
(about 950). The design of the piers, consisting of mere rectangular supports without
engaged shafts, and finished off by simple imposts instead of capitals, was, with its
bare, montonous simplicity, perfectly in keeping with the scanty skill of the Anglo-
Saxon craftsmen and builders employed in the work. The inferior quality of these
workmen is further betrayed by the poor character of the mouldings throughout the
church, and the entire absence of carving. Moreover, the presence of English hands is
indisputably proved by the use made in the transept of the characteristic baluster

1 Rolls Series Historia navorum in Anglia.
3 Rolls Series Walsingham, op. fit.

- Rolls Series Opera historica.

202

LOMBARDIC ARCHITECTURE

shafts which, being of irregular size, were made to fit their places by the addition of
clumsy bases of Anglo-Saxon character ; and their almost total unfamiliarity with the
use of the chisel is shown by the fact that the cubical capitals and the shafts have
been hewn with the axe.

St. Albans Cathedral affords only one feature of importance : the cubical capitals
with two half-rounds on each face ; the earliest specimens of the sort, of certain date,
that I can point to, and the representatives of the new Anglo-Lombardic type of

Fig. 629. St. Albans Cathedral (1077-1088).

cubical capital which, with some further embellishment, became the scalloped capital
so largely employed by the English builders.

LINCOLN CATHEDRAL was founded in honour of the Virgin by Remigius, originally
a monk of Fecamp, and afterwards bishop of Dorchester and Lincoln (IO67-IO92), 1
to replace an earlier church dedicated to St. Mary Magdalen. The foundation must
have taken place when he transferred the see to Lincoln, that is to say, after his return

1 Rolls Series Willelmi Ma'.mesHriensis monaiki gesla pontijicuin Aiiglorwn.

ENGLAND

203

from Rome (where, through the intercession of Lanfranc, Pope Alexander II had
restored to him the pastoral staff and ring), and after the transfer had been decreed
by the council held at Windsor in IO72. 1 We know that the building was finished in
1091, - and in 1092 it was dedicated.

Damaged by fire about 1141, the cathedral was brought up to date, and vaulted
('.... egregie reparando lapideis fideliter voltis primus involvit ") by the Norman

Fig. 630. Lincoln Cathedral (Xllth and XHIth Centuries).

bishop, Alexander (i 123-1 148). 3 The semi-elliptical outline remaining at the west end
of the nave shows the kind of roof constructed. The terrible earthquake of 1 185 rent

1 Kails Series Willelmi Malm, monachi gesla regiun Anglorum.

- Kotls Series Matthaei Parisieitsis monachi S. Albani chronica inaiora.

3 Rolls Series Giralili Cambrensis opera Vita S. Kemigii.

204

LOMBARDIC ARCHITECTURE

Fig. 631. Lincoln Cathedral. Central portal of west
front (Xllth Century).

adding a new facing to that
built by Remigius, in order to
enable the great recesses with
their portals, and the two
apsidal niches, to be formed
within it. It is this which
explains the fact that the
towers, which originally formed
part of the facade, no longer
stand on the front line but
behind it.

Before the Xllth century
there was nowhere to be seen,
not merely in England, but
in no country of Europe, a
portal with such deeply re-
cessed and elaborate mould-
ings as the central one at
Lincoln (Fig. 631), nor did
blank intersecting arcading
appear in England before

the building from top to bottom, 1 and
bishop Hugh of Avalon (1186-1200),
who had been prior of Witham, took
in hand its reconstruction in the Pointed
style with Geoffrey de Noiers for his
architect. 2 Of the two previous struc-
tures were retained : part of the west
front with the two great towers, after-
wards raised by the addition of a
very lofty stage in the new style ; the
westernmost bay of the nave, with some
alteration, repeated on a later occasion ;
and the walls of the aisles reduced to
the form that they present to-day.

In the west front (Fig. 630) the
work earlier than Hugh's time at once
strikes the eye. It consists of the
central portion, with two semicircular
niches, three great recesses, and three
doorways, and reaches as high as the
two ranges of intersecting arcading
above. It is the result of Bishop
Alexander's alterations, which gave an
exceptional thickness to the wall by

1 Rolls Series Chronica magistri
Rogeri de Hotiedene.

- Rolls Series Magna vita s.
Hiigonis episcopi L incolniensis.

Fig. 632. Winchester Cathedral. Part of the Old Crypt (1079-1093).

ENGLAND

205

the end of the Xlth century.
Moreover, it is quite incon-
ceivable that Remigius can have
ornamented the capitals in the
two apse-like niches of the front
with foliage in single or double
ranks, while the most important
abbey churches erected in the
island in the course of the Xlth
century and after the Norman
Conquest contained only perfectly
plain capitals, and merely, by way
of exception, one here and there
embellished with plain leaves at
the angles, and then only where it
occurred, not on the exterior, but
inside the church.

Again, it is clearly to the
Xllth century that we must assign
the finished scalloped capitals in
the portals, for this characteristic
type did not see the light before
that period. Lastly, it is quite

Fig. 634. Winchester Cathedral. North arm of transept
' (1079-1093).

F 'g- 633. Winchester Cathedral. Part of the Old Crypt
(1079-1093).

impossible to ascribe the elaborate
treatment of the doorways to the
Xlth century. The results pro-
duced by English carvers in the
last quarter of the Xlth century
were of a very different character.

The great western towers, at
least in their lower portions, and as
far as the third stage, are also
mainly the work of Bishop Alex-
ander. The elaborate decoration
of some of the arcading and string-
courses is of the same date as that
on the west front, and we know
that the blank intersecting arcading
which embellishes the towers and
their gabling cannot belong to the
time of Remigius. In the much
altered westernmost bay of the
nave, now forming a kind of
narthex, a fragment of his tri-
forium may be observed, and also
part of the wall passage made by
Alexander.

Lincoln Cathedral, though
contributing no new element to

206 LOMBARDIC ARCHITECTURE

the formation of the historical chain which it is our object to construct, is still worthy
of attention as being the first instance in England, after Durham (1093-1131), of
a nave on a great scale with a solid roof. Before taking leave of the church we may
say that we shall discuss the important reliefs of the west front when we deal with
the porch of Malmesbury Abbey.

WINCHESTER CATHEDRAL, rebuilt in 1079 by Bishop Walkelin (1070-1098),
who had previously been a monk at Saint Etienne, Caen, and succeeded the deposed

Fig. 635. Winchester Cathedral. South arm of transept (1079-1093).

Stigand (1047-1069), was consecrated in IO93. 1 The lantern tower, owing to its
defective construction, fell in 1197, but it was soon after rebuilt in the form which it
still retains. In the Xllth, XHIth, XlVth, and XVth centuries various additions
were made, and a general remodelling took place ; so that all that is left of the
original structure that is visible and in a fair state of preservation is the transept
and the crypt.

The crypt consists of a rectangular central space ending in an apse, with
surrounding ambulatory (Figs. 632, 633). The central portion, which is supported
by piers, is divided into two aisles by five short cylindrical pillars, each surmounted by

1 Rolls Series Annales monasterii de IVintonia.

ENGLAND

207

a capital formed by a Doric ovolo and abacus, recalling the central pier of the
Chapel of the Pyx in Westminster Abbey, which belongs to the time of the
Confessor. Like the ambulatory it has unraised cross vaulting. From the eastern
end starts an elongated apse divided down the middle by columns carrying vaulting
of the same kind.

The transept (Figs. 634, 635), partly rebuilt after the fall of the central tower,
was designed with a wooden ceiling for the central space and the triforium, and

Fig. 636. Piacenza. Duomo (Xlllh Century).

unraised cross vaulting in the aisles. The latter has been in part reconstructed with
ribbing, or strengthened by the addition of ribs.

We find here a form of support, new for England, consisting of a cylindrical
pillar with a pilaster and a half-column attached to it from which to start a trans-
verse arch and two springers for the groining. It was copied in Ely and Norwich
Cathedrals. We may also note the blank arcading used as an interior decoration.

Walkelin's church was a production of the School of Lanfranc so far as the
general conception goes, but it was erected by Norman master masons in co-operation
with English workmen. The presence of the former is indicated by the vaulting
the crypt, where the irregular forms of the bays demanded the services of

208

LOMBARDIC ARCHITECTURE

skilled masons. The English element is revealed by the absence of carving, and by
the bulbous bases of the transept piers ; as also by the disaster which ultimately
befell the central tower. The absence of Norman carvers, which we have already
noticed at St. Albans, must have been due to the fact that, during the years

when Abbot Paul's and Bishop
Walkelin's churches were being
built, the artists, perhaps not
very numerous, were employed
on Saint Etienne and the Trinite
at Caen, the two favourite
churches of the Conqueror and
Queen Matilda, and also in the
chapels of the castles which the
Normans were steadily erecting
in England.

Winchester Cathedral, besides
the singular form of pillar with
attached half compound pier
which we have noticed, affords
the earliest English instance of
the use of blank arcading at the
base of the walls, and also of
triplet arches with the middle
one rising higher than the others
(after the Ravennate manner), or
sometimes groups of four, which
appear here in the clerestory : a
feature which afterwards was
very freely used for the arcaded
wall passages in the naves
of Lombardo- Norman churches.
The transept aisles, forming a
continuation of the aisles of the
nave, are another new feature in
England. It was already to be
found in St. Maria im Capitol
at Cologne, and was afterwards
copied in other Continental
churches. In Italy an illustra-
tion is provided by the cathedral

Fig- 637. Tower of London. St. John's Chapel (about 1080).
(From a sketch by Miss P. Bruce. )

of Piacenza (Fig. 636) rebuilt in 1122 after the earthquake of 1117 had destroyed
the earlier church of St. Justina. Here Walkelin's conception is amplified in the
sense that the aisles of the nave not only encircle the transept but are also prolonged
into the choir. We may remark in passing that the form of the nave piers at
Piacenza was derived from the type which we noticed in Saint Philibert at Tournus
(1008-1019), and shall see used again in great English churches.

ST. JOHN'S CHAPEL IN THE TOWER OF LONDON. We do not know the
exact date of the erection of the fortress to which this chapel belongs. Though it is

ENGLAND

209

stated by William of Poitiers l that the Conqueror built a castle at London after his
coronation (1066), it is thought that this means a mere palisade with a ditch, and that
the masonry construction was not begun till 1078 or 1080.'- We hear of the new
structure as the " Tower of London " as early as 1097 . 3

The architect was Bishop Gundulf, 4 the capable builder of the castle at
Rochester. 5 It is probable that it was not begun immediately after 1077 when
he became bishop (having previously been Lanfranc's proctor), as he must have been
deeply occupied in remedying the miserable state in which he found his bishopric,
installing the Benedictines, and providing them with a
monastery. 6 A likely date is 1080.

The chapel (Fig. 637), which is on the first floor of
the Tower, consists of a nave and aisles, with an apsidal
end. The nave is covered by barrel vaulting with a
half-dome at the end, and the aisles by unraised cross
vaulting with a continuous barrel vault for the triforium.

Figs. 638, 639, 640, 641, 642, 643. Tower of London. Capitals in St. John's Chapel (about 1080).

(From sketches by Miss P. Bruce.)

I believe it to have been the work of Norman builders and craftsmen, as, indeed,
was the whole of the Tower. The Normans being, as has been observed, 7 the
importers of this type of fortress into England, we cannot suppose that they employed
inexperienced English masons on work that was quite unfamiliar to them. The
presence of Norman craftsmen is attested, apart from the form of the vaulting, which
is unusual in English Lombardo-Norman churches, by the artistic details (Figs. 638,
639, 640, 641, 642, 643), especially the foliated capitals, quite alien to English
taste at the time.

1 Duchesne, Hist. norm, script, ant. Gesta Guillclmi diuis Normannormii et regis Aiiglorum.

2 Clark, Mediaeval military architecture in England.

3 Sparke, op. fit. Chronicon Angliae : perjohannem abbalcm BnrgiS. Petri. * Hearne, Textus Koffensis.

5 Wharton, op. cit. Ernulfi episcopi Koffensis collectanea de rebus EccUsiae Roffensis.

6 Wharton, op. cit. Vita Gundulfi episc. Koffensis. 7 Clark, op. cit.

VOL. II P

2IO

LOMBARDIC ARCHITECTURE

It may be noticed that the form of the chapel, with a central arcaded space
encircled by an ambulatory, reproduces Walkelin's crypt at Winchester.

ROCHESTER CATHEDRAL was rebuilt either by Lanfranc x - 3 or by his favourite
pupil, Bishop Gundulf (1077-1108) with assistance from the archbishop. 4 This

rebuilding is believed
by some and I
agree with them to
have taken place
about 1080 after the
more pressing needs
of the see and con-
vent had been at-
tended to, during
which interval the
old church was left
standing. It was not,
however, completed
till the time of Arch-
bishop William I
(1123-1136), when
the consecration took
place in 1 130 or 1 133,
and also the transla-
tion into the new
church of the body
of the bishop St.
I t ham ar (655),
whence one may rea-
sonably suppose that
it was about then
that the old church
was demolished.

After suffering
serious injuries from
fire in 1 137 or 1 138
and 1177 or H79, 5B
and being repaired,
altered, and partly
rebuilt (Fig. 644), the

(From

Fig. 644. Rochester Cathedral. West front in XVIIth Century.
Ditgdale, " Monaslicon Anglicanum.")

only portions that I

have found surviving from the age of Gundulf, and therefore of interest for our subject,
are the two western bays of the extensive existing crypt, and some remains of rude

1 Wharton, of. cil. Excerpla ex chronico Cantuariensi, de Koberti Wincheisey archiepiscopi rebus gestis.

3 Rolls Series Radulfi de Dice/o decani Lundoniensis opera hislorica.
' Rolls Series Eadmeri historia novorum in Anglia.

4 Wharton, op. cit. Vita Gundu/fi episcopi Koffensis authore monacho Roffensi coactaneo Ermtlfi episcopi
Rcffensis collectanea de rebus Ecclesiae Koffensis

5 Rolls Series Gervasii monachi Cantuariensis opera historica.

8 Wharton, op. cit. Annales Ecclesiae Roffensis, ex historia ecclesiastica Edmundi de Hadenham.

ENGLAND

211

masonry. The bays of the crypt have short cylindrical piers, surmounted by funnel-
shaped cubical capitals (imitations of Lombardic cubical capitals), hewn with the axe,
and covered by a plain abacus. Below these piers are clumsy bases consisting of a
plinth with thick ovolo, or else plinth, large half-roll, and necking. From them
springs rough unraised cross vaulting without visible arches. Evidence of this kind
both in construction and decoration clearly points to Anglo-Saxon handiwork. To
the same source is
due, as at St. Albans,
the fact that the axis
of the church points
south-east, whereas
the primitive church
of the time of Justus,
first bishop of Ro-
chester, was correctly
orientated.

We know that
Gundulf's church was
terminated by a rect-
angular space instead
of an apse. Some
think that this was
the germ of the " Lady
Chapel " possessed by
so many English
cathedrals, whereas,
as far back as 938,
Saint Be"nigne at
Dijon had a square
chapel at its eastern
extremity, dedicated
to the Mother of God.
It was a feature per-
haps derived from
the old St. Peter's at
Rome, where behind
the apse stood the
mausoleum of the
Anicii, or church

erected by Sextus Anicius Petronius Probus, prefect of Rome in the second half of
the IVth century.

ELY CATHEDRAL. The abbey church founded in 673 by Etheldreda, wife of
Ecgfrid, king of Northumbria (670-685), and consecrated as abbess by Wilfrid, was
burned by the Danes in 870 and rebuilt in 970 by Ethelwold, bishop of Winchester
(963-984). Abbot Simeon (1081-1093), who had previously been a monk at Saint
Ouen, Rouen, and prior of Winchester (where his brother Walkelin was bishop), took
in hand its reconstruction in 1083. The work was completed in 1 106 by Abbot
Richard (1100-1107), formerly a monk of Bee (" ecclesiam suam a praedecessore

P 2

Fig. 645. Ely Cathedral. North arm of main transept (Xlth and Xllth
Centuries).

212

LOMBARDIC ARCHITECTURE

suo incoeptam aedificavit ") ; and it was he who translated into it the body of
St. Etheldreda. 1

In 1109 the church became a cathedral. Bishop Geoffrey Ridel (1174-1189)
erected the western transept, and a new west front with its tower nearly up to the
roof ; for this is surely the meaning of the words " novum opus usque occidentem
cum turre usque ad cumulum fere perfecit," 2 and not that he built the body of the
church from the third bay west of the eastern crossing, or that he completed the west

Fig. 646. Ely Cathedral. North arm of main transept (Xlth and Xllth Centuries).

end of the old church, as has been suggested, for there is nothing in the building
answering to such an interpretation.

Bishop Eustace (1198-1215) next added the narthex or Galilee to Ridel's work.
In the Xlllth century, under Bishop Northwold (1229-1254), the choir was
lengthened ; and the central tower, after its fall in 1322, was rebuilt as an octagon
instead of a square, involving the demolition of two of the transept bays and the
reconstruction of three of those of the choir. In this way all that was left of the
work of the Xlth and Xllth centuries was the arms of the choir transept (Figs. 645,

1 Wharton, op. cit. Thoniae monachi Eliensis historia Elicnsis.

' 2 Wharton, op. cit. Monachi Eliensis continuatio historiae Eliensis.

ENGLAND

213

646), and the nave (Fig. 647) as far as the western transept, the northern arm of which
is wanting. The eastern transept is aisled. The central portion and the trifi>rium
had from the beginning wooden roofs, the aisles having unraiscd cross vaulting. The
roofs of the nave and its aisles were treated on the same principle.

The general effect of Ely Cathedral recalls its elder sister, Winchester, and the
hemispherical cubical capital predominates.

We do not share the theory that the nave [of the church was completed by
Bishop Ridel, or in the reign of Henry II (1154-1189). Besides being opposed to

Kig. 647. Ely Cathedral. Nave (Xllth Century).

the statements of the Ely Chronicle, it is also at variance with the aspect of the
building. The more recent parts of the eastern transept, which must be assigned to
Abbot Richard for the older, that is to say the plainer parts, belong, of course, to
Simeon show too strong analogies of organic structure, masonry, and decoration,
with the rest of the church for us not to recognize the single idea which informs it,
and the impossibility of keeping that idea unchanged through the long series of years
between uo6and the reign of Henry II. The spirit, moreover, which animates
Geoffrey Ridel's work is too distinct to allow us to regard it as a mere continuation of
an unfinished structure left by his predecessors. On the contrary, the church must
have been perfect and complete, as is indicated by the alterations required in order to
fit the new work on to the old.

2I 4

LOMBARDIC ARCHITECTURE

I think that the small Lombardic portal known as the " Prior's doorway," a later
insertion in the south aisle wall, should be ascribed to Ridel. Its design is more in
keeping with his time than with that of Eustace, in whose Galilee the pointed arch
predominates.

Ely Cathedral was the first English church which exhibited the Romano-
Ravennate decorative feature of an arched corbel course. But apart from this we
have not discovered any other new element.

WORCESTER CATHEDRAL was rebuilt in 1084 by Bishop Wulstan II
(1062-1095) in place of the church erected by Oswald, bishop of Worcester

(961-992) and archbishop of
York (972-992) which in its turn
was the successor of the original
structure (680) of the time of
King Ethelred and Archbishop
Theodore. It was seriously
damaged by fire in 1113, and
injured by the fall of the central
tower in 1175 ; and the dedica-
tion did not take place till 12 18. 12
The only parts of Wulstan's work
spared by later reconstructions,
which are of interest to us, are a
considerable portion of the crypt
underthe choir,and a portion of the
south-east pier of the central tower.
The crypt (Fig. 648) consists
of a rectangular central space,
terminated by a semicircle with a
surrounding ambulatory, flanked
by two apsidal chapels. The
capitals of the columns are of the
Lombardic cubico-spheiical type,
occasionally chamfered off in the
lower part in order to fit the
shaft. Some of the bases are of
Anglo-Saxon type, with two thick
rolls and an ovolo.

What is left of the pier is interesting as illustrating the types of capital and
base used in the church proper, which were the same as those in the crypt.

The remains of Wulstan's church contain no specially noteworthy feature. It
was evidently the work of both Norman and Anglo-Saxon hands, the presence of the
latter being betrayed by the fall of the central tower less than a century after its
erection. The towers built by Norman masons did not share the fate which, as we
have seen, befell the similar towers of Winchester and Ely. Norman hands are seen
in the cross vaulting of the crypt, a task demanding skill only acquired by long

Fig. 648. Worcester Cathedral. Crypt (1084).

1 Rolls Series Annales prioratus de Wigoniia.

" Rolls Series Willelmi Afalm. tnoiiachi gesta. pontijicum Anglomw.

ENGLAND

215

practice in this difficult craft. The carvers employed were Anglo-Saxon, as can be
seen from the clumsy attempts to decorate two of the plain capitals in the crypt, and
also from the base mouldings. And this is confirmed by the decorative features of
the arched passage or Slype (Fig. 649) leading out of the west side of the cloister,
where the bulbous capitals formed by a very large roll between two smaller ones
(which inverted form the base as well), or by
a curiously moulded bulbous roll, are surely
of Anglo-Saxon character.

THE PRIORY CHURCH, MALVERN, was
founded by the monk Alvius or Aldwine in
1085.* The principal remains of the original
church consist of the six arches on either
side of the nave, supported by stout cylin-
drical piers (Fig. 650) and their responds,
with a portion of the wall above them ;
remains of the west front and south aisle ;
and the connection between the latter and
the transept. In the south wall there is also
an original portal, the jamb shafts of which
have bulbous capitals and bases of the same
shape, only inverted. Both nave and aisles
had wooden ceilings.

The construction of the church must be
set down to Anglo-Saxon masons, whose
presence is shown by the almost entire
absence of ornament, and also by the deco-
rative details of the portal.

Malvern Priory Church is noteworthy
for an innovation in the form of the supports,
that is to say, the employment of heavy
cylindrical piers to carry the nave arches. It
is true that cylindrical piers had previously
been used in the Chapel of the Pyx at West-
minster Abbey, in the crypt at Winchester,
and the chapels of Durham Castle and the
Tower of London ; but all the same it was at
Malvern that substantial supports of this form
made their first appearance in an English
monastic church of large size. From Malvern they spread over England, sometimes
short and squat, in other cases elongated like their prototypes in St. Philibert at
Tournus (10x38-1019) ; in one place quite plain, in another decorated in various ways,
and surmounted by the usual capital derived from the type we first found in the crypt
at Winchester. They were either plain, or else embellished with inverted truncated
semi-cones, foliage, flowers, arcading, interlacing, lozenges, discs, and other ornamental
motives. They occur in churches from London, where they appear in St. Bartholo-
mew's, Smithfield, founded in U23, 2 to Carlisle, where we find them in the Norman

1 Rolls Series Annales pnoralits di Wigornia.

2 Moore, The Church of St. Bartholomew the Great.

Fig. 649. Worcester Cathedral. Arcading in
the Slype (1084).

2l6

LOMBARDIC ARCHITECTURE

Fig. 650. Malvern. Priory Church (1085).

part of the cathedral (built
when William Rufus re-
stored the city in 1092) l
with spurs added at the
corners of the base plinth,
the earliest instance of this
Lombardic feature known
in England.

These cylindrical piers
had a long career, so that
in the second half of the
XI Ith century we find them
still employed, for instance,
in the nave of Hereford
Cathedral, belonging to the
time of bishop William de
Vere ( 1 1 86 - 1 1 99). Or,
again, in Waltham Abbey
(Fig. 651), founded as the
Church of the Holy Cross
and St. Laurence in 1062,
and rebuilt by Henry II
when he installed in it
canons regular. 23 The

original carving on the capitals of the south door exactly fits the second half of the
XI Ith century. Of the church consecrated in 1066 no trace is left.

GLOUCESTER CATHEDRAL. St. Peter's Church, Gloucester, was originally
founded in 68 1 by Osric, under a grant from Ethelred, king of Mercia (675-704).
Another king of Mercia, Beornwulf (who' according to Henry of Huntingdon 4 only
reigned one year),
rebuilt it in 823.
Aldred, bishop of
Worcester (1046-
1062) and archbishop
of York (1060-1069),
rebuilt and dedicated
it in 1058. Finally,
Abbot Serlo (1072-
1103), who had been
a canon of Avranches
and a monk at Mont

1 King Eley, The cathe-
dral church of Carlisle.

2 Rolls Series Aniiales
monaslerii de Waverleia.

3 Leland, Collectanea
de rebus Britannicis.

4 Jfolls Series Historia

Anglorum. Fig. 65 1. Waltham Abbey Church (Xllth Century).

ENGLAND

217

Saint Michel, began to build it anew in 1089, not leaving one stone upon another of
Aldred's work. The dedication took place in I loo. 1

Extensive damage was done by the fire of 1102 ("ecclesia S. Petri Gloucestriae
cum civitate igne cremata est"), and the church must
have suffered again in 1122 when the monastery was
burned. At later dates it was again injured in parts
by fire, suffered from an earthquake, lost one of its
western towers, and was altered and partly rebuilt in the
Pointed style.

The plan of Serlo's Church is the usual one derived
from William of Volpiano's design for Bernay (1013),
that is to say a basilica of Latin cross form with nave
and aisles, and two apses projecting from the eastern
walls of the transept. The only difference is in the two-
storied ambulatory surrounding the choir, with three
radiating chapels opening out of it.

The original crypt consisted of a central rectangular
space with a semicircular end, enclosed by piers with arches. Columns divide it into
three aisles, and it is encircled by an ambulatory containing three radiating apsidal
chapels. One of the capitals (Fig. 652) has on one face a rude human head, as it
were flattened, in very low relief without any undercutting. In the eastern radiating
chapel the two columns which support the arch of the apse have capitals with four
truncated inverted half-cones on the face, which form the archetype of the scalloped
capital (Fig. 653). At the entrance to the crypt are two more small chapels

corresponding to those in the transept
above.

The nave of the church is separated
from the aisles by cylindrical piers (Fig.
654). The first two bays, together with the
west front, are work of the XVth century.
The walls of the nave above the arches
are pierced by pairs of two-light openings
for the triforium (which has a wooden
roof), divided by shafts with scalloped
capitals. Above this runs the clerestory
passage, altered when the wooden roof of
the nave was replaced by vaulting.

The north aisle still retains its original
unraised cross vaulting with moulded ribs.
The awkward manner in which they meet
the piers is due, not to the fact that the
piers were originally intended only for
groined cross vaulting, but to the unfitness
of cylindrical piers for receiving multiplied arches. For though such piers have
over the Lombardic form the advantages of taking up less room and therefore
facilitating circulation, and of permitting the passage of more light, they have at
the same time the drawback of being ill adapted for carrying cross vaulting, which
requires supports specially planned to receive it.

1 Kolls Series Historia et carlularium monaiterii S. Petri Glouceslriac. 3 Dugdale, ot. fit.

F 'g- 653. Gloucester Cathedral. Scalloped
Capital in the Crypt (1089-1100).

218

LOMBARDIC ARCHITECTURE

Fig. 654. Gloucester_Cathedral. Nave (Xllth Century).

Considerable
alterations were
made in the tran-
sept in the course
of the XlVth and
XVth centuries.

The choir has
now lost its apse.
The ambulatory is
covered by unraised
cross vaulting (Fig.
655). Of the three
radiating chapels
which originally
opened out of it,
that on the south is
well preserved. In
it two rude wall-
arches of pointed
form should be
noticed.

The interior of

the choir now presents to the eye a facing of Perpendicular work, with a vaulted roof

of the same style. No doubt it originally had a wooden roof. Wide spans were

not vaulted in England till a much later date, and, moreover, if there had been

vaulting it would have been preserved. The theory that the ramping barrel vaulting

of the triforium was constructed to resist the thrust of the choir vault is readily

disproved by the fact

that, under exactly the

same conditions, the

nave of Saint Etienne

at Caen had a wooden

roof.

The ambulatory has
an upper story, from
which again three'
chapels originally
opened. It has a half-
barrel vault divided into
sections by ramping
transverse arches (Fig.
656).

The Lombardo-
Norman portions of
Gloucester Cathedral

Fig- 655. Gloucester Cathedral. Ambulatory of Choir (1089-1100).

belong to two different

periods, and are the

result of two distinct events, viz. Serlo's reconstruction, and the catastrophe of 1 102.

To the first belong the two-storied ambulatory of the choir and the crypt below it ;

ENGLAND

219

to the second, the nave. Each period is revealed by characteristic features, which
at the same time differentiate it from the other.

Thus the crypt and the choir have Lombardic cubical capitals, or else scalloped
capitals, but not as yet perfected as in the body of the church. The choir arches,
too, are in two orders, but not moulded. The unraised cross vaulting of the
ambulatory is simply groined, and the transverse sustaining arches are clumsily
fashioned. Moreover, ornament is in every case banished from the arches, and
the wall supports consist of plain engaged shafts. Lastly, the choir has its ambulatory
in two stories of the same width, and both vaulted. The choir and its junction
with the transept are therefore to be regarded as undoubtedly the work of Serlo,
for they are similar in style, and show a less advanced stage of art than the nave
and aisles.

The body of the church, on the other hand, is the embodiment of constructive,
statical, and decorative
ideas which are diametri-
cally opposed to all this.
The squat piers of the
choir are replaced by lofty
ones, at the expense of the
triforium stage, reduced
to the smallest possible
dimensions, and covered
only by a sloping roof.
The nave arches have
mouldings and ornaments.
In the aisle which remains
the cross vaulting has
moulded ribs, and is sus-
tained, like the transverse
arches, by half wall piers
with capitals, sometimes
elaborately decorated,
exhibiting carving which
is superior both in design and execution to that of the ornamented capitals in the
crypt.

It is easy to see that the choir and the nave of the cathedral are distinct from one
another both in system and in date, a fact which has been remarked by others. 1
Accordingly, the nave is the result either of a rebuilding of Serlo's work which had
been ruined by the fire of 1 102, or of a completion of it after a new design. So sharp
and marked a change in the work is inconceivable if the church had been begun and
finished by the original architect. Moreover, that the nave was built after 1 102 and
not after 1122, is proved by the traces of fire still visible, which must be referred to
the conflagration of 1 1 22. Further, the transept turrets with their intersecting arcading
are, in all probability, later than the days of Serlo, as this form of decoration did not
appear in England before the erection of Durham Cathedral (1093).

Gloucester affords the earliest perfect specimens of the characteristic scalloped
capital. We saw it in its elementary form at St Albans, and now find it in its complete

Fig. 656. Gloucester Cathedral. Upper ambulatory of choir ( 1089-1 too).

F. Bond, English cathedrals illustrated.

22O

LOMBARDIC ARCHITECTURE

shape in the eastern chapel of the crypt, and in an embellished form in the body
of the church.

It is also the earliest case in England of a choir and transept triforium entirely
vaulted ; while the two rude pointed arches in the south choir chapel are the earliest
instance of the use of that form in construction. It was not long before it was
employed in one of the monastic buildings connected with the church. The difference
between the construction of the cross vaulting in the crypt and that in the lower choir
ambulatory, and also the occurrence in the crypt of foliated capitals and of one with a
human head, show that both Norman and English carvers and masons were
employed. The Normans, for instance, are responsible for the best of the vaulting in
the crypt. The foliated capitals and the carved head must also be ascribed to them.

Buildings erected in
England in the
Xlth century, after
the Conquest, are
almost devoid of
artistic features of
this kind, so com-
mon in Normandy ;
and this means that,
for the time imme-
diately following
the Conquest,
wherever we do not
find Anglo - Lom-
bardic or scalloped
capitals and bulbous
bases, there, in all
probability, we may
recognize Norman
hands.

Before leaving

Gloucester Cathedral we may observe that the Chapter House, altered at the east
end, embellished internally with blank arcading, covered by a pointed barrel vault
provided with transverse moulded ribs, and showing traces of fire, seems to show
the same decorative motives as the nave (zigzags and billets), and should be dated
after the fire of 1 102. To the same period belong the arched passage or Slype on the
south side of the Chapter House, and some parts of the old Abbot's Lodging, the
present Deanery, for instance, the rectangular room with a barrel vault and scalloped
capitals. These buildings, having solid roofs, may have escaped the disaster which
destroyed the monastery in 1122.

NORWICH CATHEDRAL was founded with the dedication of the Holy Trinity
in 1096 by the Norman bishop, Herbert Losinga (1094-11 ig), 1 originally prior of
Fecamp, then abbot of Ramsey, and afterwards bishop of Thetford, whence he
transferred the see to Norwich in IO94. 2 The greater part of it is his work. 3 The

1 Dugdale, op. cit.

- Wharton, op. cit. Bartoloinei de Cotton inonachi Norwicensis annales Ecclesiae Norwicensis.

3 Wharton, op. cit. Historia de episcopis Norwicensibus.

Fig. 657. Norwich Cathedral. Choir (Xlth and Xllth Centuries).

ENGLAND

221

" Registrum primum," preserved in the Cathedral Treasury, states how far the
work of Herbert reached, viz. to the altar of the Cross or St. William on the north of
the choir enclosure, though the writer's information was only derived from the
tradition current in his time.

In the " Anglia Sacra" 1 Bishop John of Oxford (1175-1200) is credited with the
completion of the church, but the work of this prelate was really confined to the

Fig. 658. Norwich Cathedral. Nave (Xlth and Xllth Centuries).

repairs necessitated by the fire of n/i. 2 I believe that the cathedral was finished by
Herbert Losinga's immediate successor, Everard (1121-1145), a v ' ew which is con-
firmed by the evidence of the building itself.

Norwich Cathedral, though it has lost two of the five choir chapels, while a third
has been altered, and though the choir has been tampered with, together with the
transept (the chapel in the northern arm has disappeared), aisles, and lantern tower,
while nave, choir, and transept have been vaulted in the Pointed style, nevertheless, on
the whole, still retains its original form. The plan reproduces that of Gloucester,

1 - Wharton, op. cit. Historia de tpiscopis Noriaicensibus.

222

LOMBARDIC ARCHITECTURE

with the variation that the choir chapels are lobed instead of being rectangular
with apsidal end.

The original piers of the choir (Fig. 657) consist of cylinders with engaged
shafts and pilasters. The ambulatory has unraised cross vaulting of a superior
character to that in a similar position at Gloucester. The triforium above has a
wooden roof. Originally it must have contained ramping arches designed to give

Fig. 659. Norwich Cathedral (Xlth to XVIth Century).

support to the lofty walls of the choir, which had a wooden roof before the construc-
tion in the XVth century of the present vaulting. The clerestory has been altered.
The two-lobed chapels opening out of the ambulatory, and the outer walls of the
choir, are decorated with blank arcading.

In the transept, which, again, had at first a wooden roof, may be seen ranges of
blank arcading either simple or intersecting, with the arches in some cases sur-
mounted by gables. A quasi-triforium is formed in the thickness of the wall, at the

ENGLAND

223

level of that in the choir. The openings for the clerestory passage are either single,
in pairs, or in triplets with the centre one higher than the others.

The interior of the tower has on each side an arcaded wall passage, above
which is a range of pairs of blank arches, with a round opening at either end. The
upper story, with its own wall passage, and elaborately ornamented, is an addition
to Herbert's tower, as is shown by the difference of the original masonry.

The nave and aisles are divided by supports, alternately larger and smaller, after
the Lombardic fashion (Fig. 658). Where they have not been refaced they consist
alternately of compound piers and cylinders (either plain or with spiral grooving) for
the first six arches : the rest are compound piers. The aisles have unraised cross
vaulting, and are decorated with blank arcading.

We said that the erection of the cathedral, interrupted by the death of Bishop
Herbert (1119), was completed by his suc-
cessor Everard, and not by Bishop John of
Oxford (1175-1200). If an interval of more
than half a century had elapsed before the
work was resumed, we should never have
found that unity of conception and execution
which pervades the building, a unity mani-
fested by the masonry, the architectural
decoration, and, with one exception, the
arrangement of the church, and of so obvious
a character that it takes a skilled eye to
discover the line which divides the work of
Herbert from that of Everard. That line is
to be found in the sixth bay of the nave, for
it is there that the cylindrical piers on the
ground floor come to an end. Our dates
for the two sets of building operations bring
the statement of the Norwich Chronicle
as to Herbert's share in the work into agree-
ment with the account in the " Registrum
primum," and with the evidence of the
structure itself.

As we look at Norwich Cathedral we cannot but be impressed by the feeling that
the exterior view of the choir and transept presents a whole of an imposing effect
which is not equalled by that of any other church in the Lombardo-Norman style
(Fig. 659).

The most notable features are the following :

(i) The circular openings, either closed or forming windows.

The combination and disposition of such round openings with a raised edge on
the tower of Norwich is curiously reminiscent of the well-known tomb of Eurysaces
and his wife Atistia, just outside the Porta Maggiore (Praenestina) at Rome (Fig.
660), believed to belong to the end of the Republic or at latest to the first years of the
Empire. 1 Or it may have been derived from the round openings of a campanile such
as the early Xlth century one belonging to San Francesco at Ravenna. It was only
about the middle of that century that such circular openings, either forming windows

. 660. Kome. Tomb of Eurysaces (
Century B.C. or 1st Century A.D.).

1 Caetani Lovatelli, rasseggiale nella Kama anlica // sefolcro di Eurisace fuori delta for/a Maggiore in
Korna.

224

LOMBARDIC ARCHITECTURE

or with a decorative purpose, began to appear in English towers, as the examples at
Earl's Barton, Cambridge, and Great Dunham inform us. From Norwich it soon
passed to Exeter, where the south tower of the cathedral, begun by Bishop William
Warelwast (1107-1136) and finished by Bishop Henry Marshall (1194-1206), displays
rows of these openings between smaller ones of the same form in groups of two, four,
or five. From Exeter it travelled to Ely, where the great west tower has rows of
circular openings each containing a quatrefoil. These smaller apertures had
occurred previously in connection with the five-light openings on the tower at Earl's
Barton.

Round openings had already been used by the Roman builders to provide air and
light for the wooden roofs of buildings and for the buildings themselves, as may be
seen in the Temple of Ceres and Faustina, now Sant' Urbano, in the Valle Caffarella
near Rome, erected by Herodes Atticus in the Ilnd century A.D., where the end of
the building is pierced by a large round opening flanked by two windows. Their use
was also extended to cupolas. Plenty of instances can be found, either in surviving
remains or drawings of buildings which have disappeared. Such are the so-called
" Tempio di Siepe " at Rome ; the circular mausoleum in the Villa of the Gordians on
the Via Praenestina near Rome ; a presumable bath chamber shown in one of
Bramantino's sketches reproduced by Mongeri ; a structure of uncertain character
which appears as an illustration in our account of the Holy Sepulchre, and has its
cupola lighted by a row of round openings ; and, lastly, the so-called Tower of
Boethius at Pavia (which fell in 1584), the appearance of which has been preserved
for us by Sangallo and Spelta. 1

(2) The double-cone moulding, which reappears for the first time after the
Romano-British age in this cathedral.

As we look at the exterior and seek for points of comparison, we are struck by
the fact that, in spite of the lavish use of blank arcading, there is no suggestion of
the external open galleries which had just at this time come into fashion in Italy.
The explanation of their absence is to be found in the fact that when Herbert
Losinga, before the transfer of the see to Norwich and the foundation of the cathedral
there, visited Rome about the year 1093 in order to surrender the ring and pastoral
staff which he had obtained by simoniacal means, 2 this architectural motive was only
just making its appearance in its simplest form on the chapel of St. Aquilinus
at Milan.

TEWKESBURY ABBEY CHURCH is believed to have been founded in 715 with a
dedication to the Virgin, and must have been in existence in 800, for in that year (or
in 802) Brihtric, king of Wessex, is said to have been buried in the chapel of St. Faith.
It was refounded by Robert Fitz-Hamon, and put under Giraldus, abbot of Cranbourne,
who now became the first abbot of Tewkesbury (i 102-1 109). The precise date of this
new foundation is not known, but there are grounds for believing that it took place in
1 102 when Giraldus and his monks at Cranbourne moved to Tewkesbury. The new
church appears to have been dedicated in ii23. 345 It suffered from fire in 1178, and
the new work visible at the top of the nave walls is to be explained by this disaster.
It was not, however, " redacta in pulverem " as the Winchester annalist says, 6 for
traces of the fire are still to be seen on the fabric. In the course of the restoration

1 Historia de' fatti notabili occorsi nelt universe e in parlicolare nel Regno df Got At, ecc.

^ Rolls Series Radulf. de Dicelo opera hislorica.

3 Rolls Series Chronica magistri Rogeri de Hoiiedene. 4 Dugdale, op. cit.

6 Rolls Series Annales monasterii de Theokesberia. 6 Rolls Series Annales monasterii de Wintonia.

ENGLAND

225

the choir and transept were retouched, vaulting substituted for the wooden roof,
and the form of nearly all the windows changed. Nevertheless, with the exception of
the choir, the church remains as it was in the first quarter of the Xllth century
(Fig. 661). The plan is copied from that of Gloucester.

Tewkesbury Abbey, built under the direction of a monk, whose name Alfred
points to his being an Englishman, contains the following distinctive features.

(l) The spurred bases in the arcaded wall-passages of the nave (Fig. 662) ;

Fig. 66 1. Tewkesbury Abbey Church (Xllth Century).

the second instance of this feature that we have met with in England, the first being
at Carlisle.

(2) The great recess of the west front (Fig. 663), apparently suggested by the one
in a similar position in the Palatine Chapel at Aachen (796-804), copied again by
Bishop Alexander (1123-1148) in his cathedral at Lincoln.

(3) The unraised cross vaulting with square-edged ribs in the chapel on the
ground floor of the transept, undoubtedly belonging to the early years of the Xllth
century, and probably to the time of Abbot Giraldus ; consequently one of the first
examples to be found in England.

On taking leave of Tewkesbury we may observe that the triforium-like passage
VOL. II Q

226

LOMBARD1C ARCHITECTURE

with pairs of openings was derived from the one in Chester Abbey Church (Fig. 664)
(which only became a cathedral in 1541), begun in 7093 by Hugh Lupus, Earl of
Chester, who imported monks from Bee, and made Anselm's chaplain, Richard, their
abbot (1093-1117), and finished by Abbot William (1121-1140). The foundation of
Hugh Lupus replaced the original one of Wulfhere, king of Mercia (657-675),
restored by Athelstan (925-940), and dedicated to St. Werburgh. 1

The quasi-triforium of Tewkesbury also recalls the one in the transept of

Fig. 662. Tewkesbury Abbey Church. Nave (Xllth Century).

Pershore Abbey (the history of which is given by William of Malmesbury - and
Dugdale 3 ), which I believe does not represent the rebuilding of the wooden church
burned in 1000, but belongs to the years immediately before the refounding of
Tewkesbury.

SOUTHWELL CATHEDRAL. We know on the authority of the chronicler
Thomas Stubbs * that the church of St. Mary, Southwell, was as old as the time of
Kynsige, archbishop of York (1050-1060), who gave it two large bells. The same

2 Rolls Series IVillelmi Malm, monachi gesta pontifictim Anglorum. 3 Op. cit.

4 Rolls Series The historians of the Church of York Chronica ponlif cum Ecclesiae Eboracensis : pars prima,
anctore anonymc.

ENGLAND

227

writer mentions it again in his
account of Archbishop Aldred
(1060-1069). The new founda-
tion is generally admitted to
have taken place under Thomas
II, archbishop of York (1108-
1 1 14), for the " Registrum
Album " of the church contains
a letter from an archbishop
Thomas to the people of his
diocese in the County of Not-
tingham asking for a contribu-
tion of alms towards the build-
ing. The writer cannot be
either Thomas I (1070-1100) or
Thomas III (1300-1304), for
there is nothing in the structure
answering to the dates of these
prelates.

The new church (Fig. 665)
had its choir rebuilt in the
X 1 1 1 th century : the rest retains
its original form, though it has
lost the two transept chapels.

Fig. 664. Chester Cathedral. North arm of transept
(Xlth Century).

Fig. 663. Tewkesbury Abbey Church. West front (Xllth
Century).

The nave was designed with a wooden
roof, while the aisles had unraised, ribbed
cross vaulting (Fig. 666).

Southwell Cathedral appears to me
to be the result of two distinct sets of
operations, during the first of which the
building was finished as far as the nave,
while the second saw the completion of
the church. This would explain the
change in the mouldings of the arches in
the aisle vaulting. But there can have
been only a short interval between them,
for the structure is evidently the result
of a single conception. The interruption
may have been connected with the dis-
turbed beginning of Thurstan's primacy
(1114-1140). The two operations fall
within the period between 1 108 and about
1125, and in any case they cannot belong
to the time of Thomas I, on account of
the great cylindrical piers with orna-
mented capitals, and of the ribbed cross
vaulting. Still less can they fall in the

Q 2

228

LOMBARDIC ARCHITECTURE

years of Thomas III, for in that case Southwell would have been a church in the
Pointed style.

Southwell, with Gloucester and Peterborough, proves that by the first quarter of
the Xllth century Lombardic ribbed cross vaulting was diffused in England. Its
other noteworthy features are these.

First come the round windows which light the nave and transept (Fig. 667).
This form of aperture, which we shall notice presently in the church of Steinbach

Fig. 665. Southwell Cathedral. Nave (Xllth Century).

near Michelstadt (815-819), and in the IXth century cathedral 01 Cologne, had
at this time come into fashion at Rome under Paschal II (1099-1118), as
SS. Giovanni e Paolo, restored after the fire of Guiscard in IO84, 1 and San Clemente
tell us. The form in which it occurs at Southwell was afterwards introduced in
Waltham Abbey (1177).

Secondly, there is the pointed-arched arcading which decorates the south-
western tower, the earliest example of the sort in England (Fig. 668).

Inside the church we should also notice the Corinthianesque "storied" capitals,
with crockets at the angles embellished with roses and other ornaments, belonging to

1 I'. Germane di S. Stanislao, op. fit.

ENGLAND

229

the great eastern arch of the crossing (Fig. 669). They exhibit subjects from sacred
story, with figures of infantile design ; and also display scroll work, lilies, crosses
leaves, flowers, the symbolical lamb and cross, a dove, and a chaotic group of fanciful
buildings with arches and round windows which seems to represent Jerusalem, and
has been described, without the smallest foundation, as a church in the Byzantine
style. 1

It has been thought that these reliefs, now a good deal hidden by the organ, are
Anglo-Saxon, because they have
analogies with illuminations of the
epoch about IOOO (forgetting that
illumination is one thing and
carving another), and that they
were incorporated in the Xllth
century church, 2 without consider-
ing that these capitals exactly fit
the shafts below them and were
obviously made on purpose for
them, and that the continuous
abacus is of the regular Norman
pattern, and, thirdly, that the
crockets at the angles betray a
similar origin.

Moreover, it is inconceivable
that, at the end of the Xth cen-
tury or the beginning of the Xlth,
the Anglo-Saxon artists should
have been capable of producing
figure subjects, of rude character
it is true, but at the same time
of such broad artistic conception,
while their successors, who had
so many more opportunities for
the practice of their chisels after
the Conquest, avoided, for a con-
siderable period and as much as
they could, the representation in
carving, not only of figures, but
even of foliage, and confined
themselves of set purpose to
geometrical forms in which they
could use the axe instead of the chisel, to such an extent that in a structure of the
celebrity of Ernulfs and Conrad's choir at Canterbury (1096-1130) they used the
former exclusively : " Ibi arcus et caetera omnia plana, utpote sculpta secure et non
scisello." 8

Fig. 666. Southwell Cathedral. North aisle of nave
(Xllth Century).

OXFORD CATHEDRAL. Ethelred II's monastery was refounded after the
appointment by Roger, bishop of Salisbury (1107-1139), in mi of Prior Guimund,

1 5 Archacologia Oxoniensis, 1893 Notices of archaeological publications The illustratea archaeologist.
3 Rolls Series Gervasii monachi Canluariensis of era historica.

2 3

LOMBARDIC ARCHITECTURE

chaplain to Henry I
(1100-1135), who pro-
vided a large endow-
ment. Prior Robert of
Crickladc (1141-1180)
obtained from Pope
Hadrian IV(i 154-1159)
a confirmation of the
new privileges* of his
priory (1158). In 1180
the translation of the
body of St. Frideswide
took place, and in 1190
the church suffered from
a fire : " Combusta est
ecclesia Sanctae Frides-
wide." J 2 3 With the last
event we conclude the
history of the church, so
far as it has reached us,
as being all that is of
immediate interest for
our purpose.

Oxford Cathedral
contains no new element
which can contribute to

an exposition of the development of Lombardo-Norman architecture in Great Britain.

But we may well spend a short time over a church situated in so ancient and

celebrated a seat of culture and learning, especially

as more has been written about it to lead students

astray than to guide them along the path of truth.
The oldest parts of the building show three

separate constructions, and it is with reference to

their respective dates that the erroneous opinions

have been formulated which we shall endeavour as

briefly as possible to refute.

We have already spoken of the scanty remains

of Ethelred's church, incorporated in the new one

erected by Prior Guimund about im, the walls of

which are constructed of rubble faced with courses

of dressed stone. The similarity between the con-
struction of these walls and those of the church of

Bernay (1013) (we say nothing of that of Fecamp, of

the year 990, which is also adduced in evidence,

because we showed that there is no existing trace of

it above ground) has given rise to the suggestion

that they belong to the days of Ethelred. But this does not take into account

Fig. 667. Southwell Cathedral (Xllth Century).

Fig. 669. Southwell Cathedral. Capital
in the crossing (Xllth Century).

1 Dugdale, op. dt. - Rolls Series Annales de Osentia et chronicon Thomae Wykes.

3 Rolls Series Willelmi Malm, monachi gesta tontificnm.Anglorum.

ENGLAND

231

the difference between the condition of Normandy at the beginning of the Xlth
century a prosperous, well-governed country, under the influence of the artistic
and intellectual movement created by William of Volpiano and his fellow
labourers, and that of England bleeding from the wounds inflicted by wars
and invasions, not to speak of the vengeance, accentuated by famine, that the
massacre of St. Brice's day
had brought on the unhappy
country. Nor does the theory
consider that, from the time
of Wilfrid (634-709) to that
of Edward the Confessor
(1041 or 1042-1066), no
English dated church ex-
hibits masonry of this kind.

That it really belongs to
Guimund's work is clearly
proved by two things. The
first is the presence of the
much discussed pair of open-
ings from the old quasi-
triforium in the south arm
of the transept, with their
bulbous based shafts, their
capitals with two scallops
on each face, and their plain
arches (Fig. 670). St. Albans
Cathedral (1077-1088), was
the first building of certain
date to show capitals of the
kind. There is no trace of
a quasi-triforium of this sort
before that in Chester Cathe-
dral, founded in 1093. It is
impossible, therefore, that
Ethelred's church can have
exhibited these features
which had not been thought
of in his time.

The second is the
presence of thick cylindrical
piers in Ethelred's reputed

work, but really that of Guimund. These piers have obviously been altered, but we
are not without information as to the kind of capital they possessed, for in the south
choir aisle two of Guimund's time remain, with their capitals showing stiff, plain
leaves at the angles. But English churches of large size did not contain stout
cylindrical piers before Aldwine's application of them in Malvern Priory Church
(1085), though the School of Lanfranc had used them in the crypt of Winchester
Cathedral and the chapel of the Tower of London, but of smaller dimensions.

Guimund's church did not remain long intact, perhaps on account of defects

Fig. 668. Southwell Cathedral. West end (Xllth Century)-

232

LOMBARDIC ARCHITECTURE

Fig. 670. Oxford Cathedral. Opening in quasi
triforium (Xllth Century).

the date of this restoration,
if blown by the wind, pal-
metto leaves, and interlaced
stalks completely undercut.
All are worked with a sure
and vigorous hand, and
have a suggestion of the
Pointed style. The treat-
ment of the foliage indicates
a date contemporary with
that of similar work in
Canterbury Cathedral, of
1175 to 1184. The partial
damage shown by their
surface may be explained
by some injury suffered,
perhaps in the fire of 1190,
involvingtheir being scraped
over.

PETERBOROUGH
CATHEDRAL was destroyed
by fire in 1115, and two
years later Abbot John of
S&ez (1114-1125) took in

1 Essay on Gothic Architecture
in Observations on the Faerie Qiieene
of Spenser (and ed.).

They are

in construction, which would explain the
new facing with which it was invested ;
or, it may be, owing to injury received
when Stephen burned Oxford in 1152.
Certain it is that the structure was
remodelled, partly by rebuilding, partly
by alterations, partly by facing the old
work with new. All this was carried
out by Prior Robert of Crickdale, after
the confirmation of 1158 had put him
in a position to do so, and very probably
about 1 170 ; so that the work was quite
finished in 1180 when the translation
of St. Frideswide took place, an event
which must have been subsequent to the
renewal of the church. As long ago
as 1762 Thomas Warton l had fixed on
1 1 80 as the date of Christ Church.

The no less discussed capitals in
the choir (Fig. 671), though much
decayed, are additional evidence as to
decorated with crocket leaves, others bent as

Fig. 671. Oxford Cathedral. Choir (Xllth Century.)

ENGLAND

233

hand its rebuilding. After his death the work made slow progress, his successor,
Henry of Anjou (1128-1133), if we are to believe the account of him in the Anglo-
Saxon Chronicle, not being the person best fitted for the task. It was not till 1 140 or
1143 that the monks took possession of the new choir which had been completed
by Abbot Martin of Bee (1133-1155). The works were continued under Abbots
William of Waterville (1155-1175), who erected the transept and the three lower
stages of the lantern tower, and Benedict (1177-1193), who constructed the nave as far

Kig. 672. Peterborough Cathedral. South arm of transept (Xllth Century).

as the west end. The last additions were the western transept and west front, and the
church was dedicated in 1237 under Abbot Walter of Bury St. Edmunds (1233-1245).'
In later times, from the XHIth century onwards, the bell-tower was erected, the
windows altered, the central tower rebuilt in the Pointed style keeping only two of
the supporting arches and the piers, recently reconstructed on the old lines. Further,
a porch was added to the west front, and the eastern end of the choir was concealed
by a new structure. In 1541 the abbey church became a cathedral.

1 Sparke, op. n't. Chronicon Angliat per Johanndii abhalein Burg i S. Pelri Hu^onis Candidi cottubii
Burgcnsis historia.

234

LOiMBARDIC ARCHITECTURE

Choir, transept (Figs. 672, 673) and nave (Fig. 674) were designed with wooden
ceilings, but the aisles had from the beginning unraised ribbed cross vaulting (Fig.
675). The date of the painted ceiling of the nave has been the subject of much
controversy. I have examined it, especially from under the roof, and have come to
the conclusion that it belongs to the XlVth century, when the lantern of the central

| :

F'g- 673. Peterborough Cathedral. North arm of transept (Xllth Century).

tower and its eastern and western arches were rebuilt, and the original flat ceiling was
replaced by a new one which fitted the new pointed western arch of the tower.
Moreover, we cannot imagine that there would have been any question of so rich a
ceiling at a time when the nave which it covered showed such poverty of decoration.

Peterborough Cathedral tells the same story as the one preserved by the
Chronicle andjHistory of the monastery to which we have referred. The work of John
of Seez includes the apsidal end of the choir with its perfectly plain arches, and the

ENGLAND

235

choir aisles west of the apse, where again the capitals are absolutely plain. To Abbot
Martin are to be assigned the upper story of the choir aisles, which shows a larger
amount of ornament and embellished capitals, and also the eastern side of the transept.
William of Waterville is responsible for the completion of the transept and the
erection of the first two bays of the nave and aisles, where the tympana of the
triforium arches repeat the decoration found in the choir and transept. Benedict
extended the nave to the west end.

The unraised ribbed cross vaulting in the choir aisles should be noted. It is

Fig. 674. Peterborough Cathedral. Nave (Xllth Century).

original and not the result of alterations in the course of which the diagonal ribs
were added, as has been mistakenly suggested. This I was able to verify for myself
when the central tower was being rebuilt. This vaulting was unquestionably
constructed between 1117 and 1125, and it provides a fixed point for the history of
Lombardic ribbed cross vaulting in England.

DURHAM CATHEDRAL, erected and dedicated in 998 by Bishop Aldhun
(990-1019), was rebuilt by Bishop William of St. Carilef (1080-1096), first a
secular priest of the church at Bayeux, and then monk and prior of the convent of
Saint Carilef (Saint Calais). The first stone was laid in 1093. How much of the

236

LOMBARDIC ARCHITECTURE

Fig. 675. Peterborough Cathedral. South aisle
(Xllth Century).

building he erected we do not know
precisely. All we are told is that, after
a vacancy of three years in the see,
William's successor, Rannulf Flambard
(1099-1129), found that the work ot
the late bishop and the monks reached
as far as the nave. These works,
which must have included the two
easternmost bays of the nave aisles, as
the cross vaulting is of the same char-
acter as that in the choir aisles and
transept, were not completed ; for we
hear that at the translation of the
relics of St. Cuthbert in 1104 the choir
was still blocked up with the timber
used for the construction of the recently
finished vaulting.

Between 1 129 and 1133, when there
was another vacancy in the see, the
monks completed the nave, that is to
say, they roofed it with vaulting, as we
are told in a well-known passage relat-
ing to Rannulf's work : " His namque
sumptibus navem ecclesiae circumductis
parietibus ad sui usque testudinem erexerat " " muros navis novae ecclesiae
Dunelmensis fabricavit." I23i

Durham Cathedral, though deprived of its original east end, replaced by the
secondary transept in the Xlllth century, with the central tower almost entirely
rebuilt, the west front altered and partially concealed by the present Galilee, the
western towers raised by four arcaded stories, and changes effected in other ways,
still remains on the
whole such as it was
when erected be-
tween 1093 and 1133
(Fig. 676). The plan
is that of a Latin
cross with choir, nave,
and aisles, terminat-
ing originally in an
apse with minor

1 Rolls Series Symeonis
nionachi historia Eccl.
Dimelm.

- Rolls Series Symeonis
nionachi hist. Eccl. Dunelm.
continuatio.

3 Kolls Series Willelni
Malm, monachi gesta ton-
tificum Anglorum.

4 Leland, De rebus

Britannicis collectanea. Fig. 676. Durham Cathedral (Xlth and XI]

ENGLAND

237

apses at the sides. The choir aisles
(Fig. 677) have ribbed cross vaulting of
parabolic form springing from alter-
nately larger and smaller supports. The
choir itself (Fig. 678) has XI I Ith century
vaulting replacing the earlier vaulting
which threatened to collapse. In the
transept (Fig. 679), both the aisles and
the central space are also covered with
ribbed cross vaulting. We observe here
corbels with semi-human heads recalling
those we noticed in Saint Etienne at
Caen. Later they will be used to orna-
ment the apex of hood-mouldings above
pointed arches in naves. These heads
sometimes have the tongue protruding,
the source of the equally characteristic
monster heads with the tongue curled
over the roll moulding, which occur in
English doorways of the XI Ith century,
for instance at Lincoln and Southwell
Cathedrals. They are rude work, but
vigorously treated ; and an experienced
eye can see that the original heads in
the south arm of the transept are
artistically somewhat superior to those
in the north.

The nave is separated from the
aisles by arches resting on compound
piers alternating with cylindrical piers
(Fig. 680). Its cross vaulting is ribbed,
and sustained by transverse arches of
pointed form springing from corbel
heads like those just described, but
artistically more elementary than the
similar ones which ornament the apex
of the hood-mouldings above the
pointed arches in the nave of Malmes-
bury Abbey, erected some time after
1 153. The fact that these corbel heads
are a later addition shows that the
architect of the nave intended merely
to throw transverse arches across it,
and cover it with a wooden roof. The
ribbed cross vaulting in the two eastern-
most bays of the nave aisles has plain
ribs, as in the choir and transept aisles
and in the north arm of the transept.
In the case of the other bays the

Fig. 677. Durham Cathedral. North choir aisle
(Xlth Century).

Fig. 678. Durham Cathedral. Choir (Xlth and
Xllth Centuries).

2 3 8

LOMBARDIC ARCHITECTURE

diagonal ribs have the zigzag moulding, which shows that Rannulf Flambard took up
the work at the second bay.

Intersecting blank arcading is freely used on the ground floor of the church.
There is a triforium, with a clerestory passage above. The west front is also
decorated with intersecting arcading. The archivolts of the central portal are
ornamented with zigzags, foliage, medallions with human and animal figures. The
jambs are left plain. The simple treatment of this doorway as compared with the

two elaborately decorated ones near
the west end of the aisles (Fig. 68 1),
the shallow relief, and flat, clumsy
character of its carving, show that
there is an interval of several years
between them. The aisle doorways
may very well belong to the time of
that great builder, Bishop Hugh
Pudsey (1153-1195).

In the original parts of the
cathedral the Lombardic cubico-
spherical capital and the scalloped
capital predominate. The only ex-
ceptions are the Corinthianesque
specimens on the internal doorway
near the south arm of the transept,
and four others of the same type
which occur in the blank arcading
of the nave aisles. For bases the
prevailing forms are (i) the Norman
type found in Saint Etienne and
Saint Nicholas at Caen, consisting
of two hollow chamfers, so shallow
that they appear almost like a single
straight chamfer ; (2) that formed by
one or two slight hollow chamfers
and an ovolo ; and (3) that moulded
into a hollow between two very flat
ovolos.

Durham Cathedral presents
three peculiar features worth atten-
tion.

First and foremost is the intersecting blank arcading in the choir aisles. It
is the earliest specimen outside Italy : previously there had been nothing but small
intersecting arches such as are carved on a capital in the crypt of Lastingham Church
(about 1078). It was an Anglo-Norman invention, but whether the originator derived
his idea from the large arches intersecting two smaller ones in the great Mosque at
Cordova (785-900), or from the intersecting arches used in carving by the Romans
and the Lombard gilds, it would be impossible to say. Certain it is, however, that it
did not make its appearance before the last quarter of the Xlth century.

Then there is the decorative treatment of various forms applied to the
cylindrical piers with the object of relieving their ponderous appearance, and

Fig. 679. Durham Cathedral. South arm of transept
(Xlth and Xllth Centuries).

ENGLAND

239

The feature appears frequently elsewhere
enclosed by a single arch, in the

removing the monotony of the effect,
in England.

Thirdly, there are the pairs of openings
Ravennate fashion.

But the great importance of the building consists in the ribbed cross vaulting
with transverse arches of round or pointed form, which it contains. Much has been
written about this vaulting, and the discussion is still going on. It is argued on the
one side that it is a later addition, because it is to France that belongs the credit of
the invention of this
form of vaulting, the
principal factor of the
Pointed style ; while,
on the other side, the
view is maintained that
it is of the same date
as the rest of the
structure, in which case
the credit of the dis-
covery will fall to
England.

Unfortunately the
champions of the two
parties have hitherto
confined their re-
searches to too re-
stricted a field ; and it
has never struck them
that, while they were
arguing whether ribbed
cross vaulting appeared
about the end of the
Xlth century or at the
beginning of the Xllth
(although, as Dehio 1
rightly observes, the
Lombards had already
made use of the ribbed
cross vault with but-
tresses effecting the
same object as those developed later in France), this very same form had already
attained a respectable antiquity in Italy. The proof of this statement depends on my
discovery of the facts in San Flavianoat Montefiascone(iO32). Those facts make the
existence of cross vaulting with diagonal arches in Durham Cathedral, as early as the
time of its rebuilder, William of St. Carilef, easy to understand. And we must not
forget that when he went to Rome as an envoy from the Conqueror, he may well have
had the opportunity of inspecting the new form in San Flaviano itself, for Montc-
fiascone was one of the regular halting places on the road to the Eternal City.
Moreover, the Normans were brought into direct contact with South Italy after their

1 Reperlorium fur JCtmstwissensckaft, 1896 Die Anfiinge ties gothischen Baiulils.

Fig. 680. Durha

Centuries).

240

LOMBARDIC ARCHITECTURE

countrymen, by the victory of Civitella (1053), an d the investiture conferred by Pope
Leo IX (1049-1055), had obtained a firm footing there ; and, as v/e have seen, vaulting
of this form existed in Campania. Nor is it conceivable that cross vaulting of this
type was invented straight off at Durham, far away as it was from the very rare
examples of intersecting vaults with visible ribs left by the Romans. In architecture,
when it comes to essential elements, the idea is one thing and its execution another. 1

The Durham ribbed cross vaults are not ordinary intersecting vaulting to which
ribs have been added afterwards, for the latter are quite independent of the vaulting
cells, and are thus constructed on the Lombardic principle. Nor is it a case of
ribbed replacing groined cross vaulting, the idea being disproved by the vaulting

shafts at the angles where
the choir and transept aisles
meet. The clumsy way in
which the vaulting is set
must be ascribed, over and
above the fact that it was a
first attempt on the part of
the English builders, to the
absence of the logical Lom-
bardic arrangement of com-
pound supports and plain
arches.

The ribbed cross vault-
ing in the transept aisles and
the first two bays of the
nave aisles is of the same
period, and belongs to the
years between 1093 and
1099. The whole of it may
be credited to English work-
men, now emancipated from
Norman tuition. The Nor-
mans had no previous ex-
perience of this type of
roofing in their own country.
The choir itself as well
as its aisles had cross vault-
ing. In the clerestory walls
may still be seen traces of its springing marked by the junction between the original
work and that of the XII Ith century. Further, the existence in the set back face of the
triforium of groups of three wall shafts above the main arch piers suggests that each
bay contained two quadripartite vaults with transverse visible arches springing only
from the half-piers. What form this vaulting took we cannot say. It need not have
been ribbed merely because that in the aisles is so, for the church of Rivolta d'Adda
and Sant' Ambrogio at Milan have simple cross vaulting in the aisles, and the ribbed
form or barrel vaults in the nave. Semicircular arches, intended to receive the
thrust of the vaulting, cross the triforium at intervals corresponding to the half-piers and
the groups of vaulting shafts. In 1235 the original vaulting of the choir threatened to

1 Rivista if Italia, October, 1908 U. Gnoli, Le origini delT architetttira Loinbarda.

Fig. 68 1. Durham Cal

Portal in North aisle (Xllth Century).

ENGLAND 241

fall in, 1 and its reconstruction was carried out in 1242 by Bishop Nicholas of Farnham
(1242-1248), for such must be the meaning of Lcland's 2 words "fecit testudinem
templi " ; the choir vaulting with the corbels for the diagonal ribs being in the
Pointed style, which is not the case with the vaulting of the nave.

The builders of the church had a further intention, as has been observed, 3 of
vaulting the transept, for the triforium on the east side contains the regular buttress
arches set in relation to the half-piers and pairs of wall-shafts. But in the course of
the work they changed their minds. On the western side they omitted the pairs of
wall-shafts on the face of the quasi-triforium from which the ribs of the cross vault-
ing were to start, and in the clerestory, on the west side of the south arm of the
transept, a range of continuous arcading was constructed, evidently implying a flat
ceiling. The explanation of this change is perhaps to be found in fears for the
stability of the western sides, devoid of the support of the cross vaulting in the aisles
and of the buttress-arches of the triforium, in view of the weight of the vaulting.
However, they soon gained fresh courage and decided on the construction of vaulting,
beginning with the north arm of the transept. The absence of the pairs of vaulting
shafts in the triforium stage was made up for by the insertion of corbels. The plain-
ness of the ribs suggests that they are of the same date as the cross vaulting of the
choir aisles, of the transept, and of the two first bays of the nave aisles ; but the carved
corbels from which the ribs spring, besides being a later insertion, show an artistic
advance beyond the capitals of the ojd door near the south arm of the transept,
inserted not later than 1099, and are certainly subsequent to 1104: therefore the
cross vaulting must be of the same date. The vaulting in the south arm of the
transept shows the same characteristics as that in the northern, but the ribs are
ornamented with zigzag mouldings, and it is therefore later.

The nave was designed for a wooden ceiling, and crossed by transverse arches to
give it stability. Some interval must have elapsed before it was begun, for the
" Ravennate " openings of the triforium cease in this part. The construction of the
vaulting involved an alteration of the clerestory, as some arches will be noticed
which have been blocked up to give room for the vaulting. It is of quadripartite
form, with round diagonal arches springing from corbels, single, or in pairs, and
carried on transverse arches of pointed form. Its construction shows a marked
advance beyond that of the transept, and this progress is also exemplified by the form
of the arched buttresses in the triforium (taking the place of the simple pilasters
backing the arch piers), which are not round as before, but ramping, and counteract
the thrust of the vaulting. But in spite of this difference there is a close analogy
between the ribbing here and that in the transept cross vaults, so that we may very
well ascribe all of them to the work of the same builders.

It has been suggested 4 that the nave vaulting, and indeed all the ribbed cross
vaulting in the church with the possible exception of the choir aisles, is not earlier
than 1133. I do not know at what date to fix this except in the years 1129-1133.
For, on the one hand, we find in the vault construction, beginning with the choir
aisles and ending with the nave, that undoubted sequence in order of time which has
been noticed, 5 corresponding too, as it does, with the historical account left by
Simeon and his continuator ; a sequence both in construction and decoration
confirmed in addition by the artistic progress shown by the corbel heads. While, on

1 Raine, Saint Cuthbert. * De rebus Britannicis collectanea,

* Journal of the R.I.B.A., Third series, vol. vi, Bilson, The Beginnings of Gothic Architecture.
4 De Lasteyrie, Discoiirs stir les origines de r architecture sotkii/ue (Socittt aes Antii/uaires tie Normandic}.
B Bilson, op. cit.
VOL. II R

242

LOMBARDIC ARCHITECTURE

the other hand, in the series of works carried out between 1134 and 1 195 under

Bishops Geoffrey Rufus (1133-1140) and Hugh Pudsey (1153-1195), we meet with a

more advanced style, which differentiates them from the nave.

These conclusions I had reached in 1896. Later study of the building has

only strengthened them, and I have found that, on the whole, the story of the

vaulting as told above, and foreshadowed as early as I8/9, 1 may be accepted as the

truth.

Accordingly, in default of further discoveries, Durham Cathedral must have the

credit of forming the connecting link between Lombardic and Pointed architecture, in

which every constituent element
of the former was to gain fresh
vitality and increased oppor-
tunity, while the style itself
having reached a perfect develop-
ment was destined by a natural
process to give place to a new
form.

Before concluding the pre-
sent chapter a short space may
be devoted to a church which
contains carving of importance
for purposes of comparison : I
mean

MALMESBURY ABBEY
CHURCH. The contradictory
statements of William of
Malmesbury leave it uncertain
whether the church of St. Mary
erected by Aldhelm, abbot of
Malmesbury (680-705) and
bishop of Sherborne (705-709),
was rebuilt or at any rate restored
by Elfric, who became abbot
after 970, and bishop of Crediton
in 977. Thus while in his life of Athelstan (925-940) he speaks of it as " postea regis
Edgari diebus sub abbate Elfrico aedificata " 2 this Elfric being a well-known
builder in his " Gesta pontificum Anglorum " he ascribes it instead to the efforts of
Aldhelm. Richard of Cirencester 3 says it was built by Elfric, and repeats William's
words. In any case, the church was still standing in William's days (he seems to have
died about 1143): "Nam tota maioris ecclesiae fabrica Celebris et illibata nostro
quoque perstitit aevo." 4

Further, there is no documentary evidence as to the precise date of its rebuilding

1 Greenwell, Durham Cathedral. 2 Rolls Series Gesta regain Anglorum.

3 Rolls Series Speculum historiae De gestis regum Angliae.

4 Rolls Series Gesta pontificitm Anglorum.

Fig. 682. Malmesbury Abbey Church. Nave (Xllth Century).

ENGLAND

243

I^HHIHI

Fig. 683. Malmesbury Abbey Church. Carving from outer portal (Xllth Century).

in the Xllth cen-
tury. Every prob-
ability, however,
is in favour of the
years which fol-
lowed the death
of William the
historian ; and the
new construction
may be connected
with the indepen-
dence recovered
by the monastery
in 1 140, and the
treaty which put
an end to the civil
war in 1153, for no

small part of the struggles which disturbed the reign of Stephen (1135-1154) took
place in the neighbourhood of Malmesbury, making it unlikely that while they lasted
the monks would venture on an enterprise of this scale. As a matter of fact, all the

characteristic features of the
surviving parts of the build-
ing which have not been
subjected to later alteration
point to the third quarter
of the Xllth century ; and
I do not think that there
is any one with experience
in such matters who can
possibly fix upon any date
but, approximately, the one
suggested above (Fig. 682).
After this preface let
us pass to the carving on
the porch, as being the sub-
ject in which we are imme-
diately interested.

The porch projects from
the south aisle of the church,
and its sides are decorated
internally with arcading
which has lost some of its
members, and two sculp-
tured lunettes. It is entered

Fig. 684. Malmesbury Abbey Church.
Century).

Lunette in the porch (Xllth

from the outside through a
portal cased in a construc-
tion of the XlVth century.
This portal is decorated with scroll work, interlaced lozenges, foliage, and other
ornamental forms. But more important are the figure subjects (Fig. 683) on the

R 2

244

LOMBARDIC ARCHITECTURE

Fig. 685. Malmesbury Abbey Church. Inner door
in porch (Xllth Century).

second, fourth, and sixth orders. From the description of these sculptures by an

anonymous writer who visited the church in 1634, we learn that all the subjects

(except a very few which had disappeared) were taken from the Old and New

Testaments. 1 Some of those that are

least decayed can still be identified.

Each of the two lunettes (Fig. 684)
on either side of the porch contains a
group of six figures seated in a row
on a bench, and undoubtedly repre-
senting the Apostolic College, with
angels floating in the air above them.
The Apostles, mostly bearded figures,
are seated in various attitudes, looking
some to the right and some to the left,
or with the head resting on the shoulder,
and have about them a certain air of
solemnity. The mantles in which they
are wrapped are draped in numerous
folds which fall in various arrangements.
The shapeless feet are the only part
treated without distinction in a uniform,
monotonous manner.
The doorway (Fig. 685) which leads into the church has the orders decorated

with lozenges and branching scroll work. At the top they enclose a tympanum

containing a figure of Christ in the act of blessing, seated in a vesica supported

by two angels.

It appears to me that all these sculptures, treated without undercutting, are the

production of one school and one

date, though the work of three dif-
ferent hands. Thus, the decorative

treatment of the two doorways is the

creation of a single mind, but it is

noticeable that on the outer one both

ornaments and figures are modelled

with higher artistic skill than those of

the arch within. The carving of the

two doorways must, then, be assigned

to two distinct artists. Again, the

drapery of the figures in the lunettes

and tympanum is closely related, but,
on the other hand, the figures in the
latter are more artistic and better pro-
portioned than those in the lunettes
at the sides. These must accordingly
be assigned to a different chisel.

We may remark at this point

that all this carving, whether it consists of ornament or figures, is to be distin-
guished from that on the capitals in Archbishop Roger's (1154-1181) crypt in

1 Brayley, Graphic and historical illustrator Topographical excursion.

Fig. 686. Lincoln Cathedral. Carvings in the west
front (Xllth Century).

ENGLAND

2 45

Fig. 687. Lincoln Cathedral. Font (\ll\h Century).

York Minster, belonging to the first years of his episcopate. The carving at
York, consisting of scrolls, plants with pellets, palm leaves, intersecting arches,
bunches of grapes, and eight men in
tunics grasping a cable which runs be-
low the abacus, are correct in drawing
so far as regards the decorative motives,
but the figures show no sense of pro-
portion.

As we stand before the carvings
of the porch at Malmesbury, forming a
collection which, with the exception of
the important reliefs on the west front
of Lincoln Cathedral, is unrivalled
among English Xllth century monu-
ments, we cannot but regret that the
relentless hand of time, the mischief
wrought by man, and the quality of the
stone employed, have combined to re-
duce so much of it to the mutilated
and decayed condition in which we see
it to-day. But in spite of this, it is
of assistance in fixing the date of other

works which have given rise to differences of opinion, and of these we will select a
few of the best known.

Let us begin with the reliefs which decorate the west front of Lincoln Minster.
Various views have been put forward by those who have studied the subject, and it
has even been suggested, owing to the way in which they are inserted in the wall,
that they belong to the Anglo-Saxon age and are not in their original place

ignoring the fact that this un-
symmetrical decorative treat-
ment of church fronts was not,
during the Middle Ages, effected
by the adaptation of sculptures
brought from elsewhere, but was
a regular form of decoration,
reduced to a system by the
Lombard gilds, svho made their
principal display of it in San
Michele Maggiore at Pavia
(Xllth century).

The Lincoln sculptures,
^jjj^ft ht-'ri-' ;uul tlicrc renewed, are
treated in bas-relief, and have
sacred subjects (Fig. 686). The
human figures are fairly pro-
portioned and moulded, but
generally stiff, awkward in their
movements, and expressionless in feature, while the drapery is very rigid. The
scene of the damned being carried off by devils, where the carving is untouched,

Fig. 688. Winchester Cathedral. Font (Xllth Century).

246

LOMBARDIC ARCHITECTURE

should be specially noticed.
While these reliefs display
an art considerably more ad-
vanced than that of the
" storied " capitals at South-
well (1108-1114), it is less
developed than that exhibited
by the Apostles at Malmes-
bury, and the last years of
Alexander's episcopate (i 123-
1 148) suit them perfectly.

The black Tournai marble
font in the nave (Fig. 687),
recalling the one at Winchester
(Fig. 688), is the product of
another school of artists.

Let us pass to the so-
called " Prior's Door " in Ely
Cathedral. The jambs and
the archivolts are enriched
with scroll work, foliage,
flowers, pellets, human beings,
quadrupeds, birds, fishes, mon-
sters, all treated with freedom,
and in deep relief, though

Fig. 689. -Worcester Cathedral ^Chapel of St. John. Capitals ^^ undercuttmg Qn the

whole it shows a more skilful

arrangement and better distribution of parts than any other specimen, with the
exception of the porch at Malmesbury. The tympanum contains the figure of
Christ blessing within a vesica held up by two angels. These figures are rather flat
and not at all undercut, but the drapery is richer than that at Malmesbury, and the
hair is treated more artistically. I think we shall not be far wrong if we place them
in the first years of Bishop Geoffrey Ridel

(1174-1189).

To the same period, that is to say, to
the years subsequent to the fall of the central
tower (i 175), may be assigned the well-known
carvings (where untouched) on the capitals
in St. John's Chapel, opening out of the
south arm of the main transept of Worcester
Cathedral (Fig. 689). These carvings, con-
sisting of scroll work, a winged dragon with
a serpent's tail, and the winged head of an
angel, exhibit an art obviously of the same
period as that of the Ely doorway.

Next comes the door in the west front
of Rochester Cathedral, reputed to be of
the time of Henry I (1100-1135), but really

I of 01- fV... *u c. c IA. Fig. 690. Norwich Cathedral. Sculptured figure

later than the fire of 1179. Its decoration * 7 north arm of transept (X 1 1 th Century).

ENGLAND

247

comprises foliage (sometimes treated with a flavour of the Pointed style), scroll
work, the figure of Christ in a vesica held by two angels and. surrounded by the
emblems of the Evangelists, the figures of a king and a woman, human heads,
and realistic or imaginary animal heads. All is well designed and modelled,
treated with vigour, and sometimes completely undercut. The king in particular,
though damaged, shows fine drapery and treatment. One would look in vain for
carving of this character in England, not merely in the time of Henry I, but even
at the beginning of the second half of the Xllth century.

Compared with that at Malmesbury the carving on the portal at Rochester
shows advance both in the
decorative parts and in the
figures ; the animals, too,
are more successfully treated
than those in the crypt at
Canterbury, not to say those
on the door at Ely. More-
over, the composition which
fills the tympanum shows
progress beyond the similar
feature of the doors at
Malmesbury and Ely, the
latter being derived from the
former. We shall be correct
in placing it in the first
years of the Xlllth century
in the time of Bishop Gilbert
de Glanville (1185-1214),
when a good deal of work
was carried out in the monas-
tery at Rochester. We can-
not accept the view of those
who would see a foreign hand
in the doorway on account
of the absence of billets,
for this detail is equally
wanting on the doors at Ely
and"Malmesbury. The Anglo-
Saxon artists had been roused
from their traditional inertia by the Norman Conquest. The carvers, whom prefer-
ence for the axe and the rarity of employment had rendered almost incapable of
executing the finer work, were reinvigorated by the new artistic influence, and
slowly but surely began to advance along the road of the revival which was stirring
Western Europe. The figure in the act of blessing, to be seen on the outside of
the north transept of Norwich Cathedral (Fig. 690), the " storied " capitals at South-
well, and the figure capitals at Tewkesbury, are so many stages on the road
leading to the Lincoln reliefs, the capitals in Roger's crypt at York, and the
sculptures at Malmesbury.

Earlier than the carving of the Rochester door is the less advanced work on the
capitals of St. Peter's, Northampton. It consists of foliage, real or imaginary

_;. 691. Chichester Cathedral (Xllth Century).

2 4 8

LOMBARDIC ARCHITECTURE

Fig. 692. Chichester Cathedral. Sculpture (Xllth
Century).

Fig- 693. Chichester Cathedral. Sculpture (Xllth
Century).

creatures, pearls, birds, grapes, complicated interlacing, &c.
the church was about II6O: 1 the date ugo 2 must be that
the building.

Two works not
much later than the
Rochester door are
the marble panels
which were found
concealed behind
the choir stalls in
the last century, and
are now built into
the wall of the
south choir aisle of
Chichester Cathe-
dral (Fig. 691).
This church was
consecrated in
1108, severely
damaged by fire in
1114, repaired by
Bishop Ralph de
Luffa (1091-1123),
reconsecrated in
1184, injured again

The foundation of
of the completion of

1 Parker, An Intro-
anction to the Study of
Gothic Architecture.

2 Serjeantson,^ His-
tory of the Church of St.
Peter, Northampton.

Fig. 694. Baii. Crypt of San Nicola (Xlth Century).

ENGLAND

249

and still more seriously by the fire of 1187, and dedicated in 1199 under Bishop
Seffride II (1180-1204) though the works of the second restoration were still in
progress. The panels represent Jesus meeting Mary and Martha at Bethany, and
the Raising of Lazarus (Figs. 692, 693). The treatment of the figures is superior
to that of any which we have seen hitherto. The peculiar care and variety with
which the hair and beards are treated should be noticed. The hands and faces,
though not yet of the right proportions, are none the less executed with a
certain naturalness. A peculiar feature are the cavities representing the eyeballs.

Kig. 695. Canosa.

Sepulchral chapel of Bohemond
Century).

Fie 6fl6. Santa Maria Capua Vet ere. Tomb called " La

" *-* - Jl

Conocchia.

These reliefs, together with the contemporary fragments discovered during the
restoration of the Cathedral, in spite of the merits to which we have called attention,
are immature both in composition and details, and therefore less advanced than the
figures at Rochester. Accordingly, they will find their place in the last decade of
the Xllth century during the episcopate of Seffride II. Many persons, it is true,
believe that they came from Selsey, whence the see was moved to Chichester in
1075 ; and the latest dates allowed by them are the Xlth century (and before 1075),
or some time in the Xllth. The formless figure sculpture of that date in England
refutes any such attribution.

250

LOMBARDIC ARCHITECTURE

All the reliefs which we have examined may be ascribed to English hands. The
art of carving in the countries most likely to provide artists for such a purpose, viz.
France and Italy, was at that time far more advanced than that exhibited by these
sculptures.

A date later than that of the Rochester doorway must be assigned to the two
well-known figures of angels built into the walls of St. Laurence's church, Bradford-

*ig' 697. Athens. Choragic monument of Lysicrates (335 B.C.).

on-Avon. Their greater purity of line and freedom of movement show that they are
later than the works which we have previously examined, though the execution is still
rough. They might well belong to the early years of the Xlllth century if we found
them in one of the more progressive artistic centres of England. But seeing that
they are in an out-of-the-way locality, even though the monastery was connected
with a rich and powerful abbey like Shaftesbury, it is impossible to date them within
any but very uncertain limits, but in any case later than the Xllth century.

ENGLAND

251

It has been supposed by some writers on these subjects that the Lombardo-
Norman basilica was the model from which that found in the Norman conquests
in South Italy and Sicily was derived. They even describe as " Norman " the style
of the numerous churches erected in those lands in the Xlth and Xllth centuries.

But, as a matter of fact, an examination of these edifices reveals, at the most,
two occasional characteristics inspired by the Lombardo-Norman church, viz. the
plan, and the intersecting
arcading. None of the
following features can be
regarded as Norman im-
portations.

(1) The somewhat
sharply pointed roofs
which occasionally occur,
e.g. in San Nicola at Bari
(1087-1098-1105). The
form was not demanded by
local climatic conditions,
but depended on the
aesthetic consideration of
giving greater elegance to
the outline of the gable by
making it less depressed.

(2) The indication on
the facade of the internal
arrangement of the church.
This idea, which originated
at Ravenna, had been
embodied as early as the
Vlth century in the church
of Bagnacavallo.

(3) The Lombardic
arrangement of supports
alternately larger and
smaller, exemplified in the
nave of San Nicola. It
had appeared as far back
as 985 in SS. Felice e
Fortunate, near Vicenza,
and was of Roman origin.

Fig. 698. Bari. Cathedral. Window in the apse (Xllth Century).

(4) Bell-towers flanking the choir. We saw the origin of these in our account
of the Duomo of Ivrea, and we shall learn more when we come to deal with the old
cathedral of Cologne.

(5) The capitals, revealing influences of all kinds (Fig. 694) Roman, Ravennate,
Byzantine, Pre-Lombardic, Lombardic, Apulian, Calabrian, Campanian, Sicilian
everything except Norman. Unless, indeed, they are Roman, and brought from
elsewhere ; or of Byzantine origin, like the basket capitals with figures, probably of
the Vlth century, in the crypt of the Xllth century cathedral of Otranto.

(6) Cupolas having the drum of polygonal form externally, with engaged shafts

2 5 2

LOMBARDIC ARCHITECTURE

at the angles, as shown in the sepulchral chapel of Bohcmond, Prince of Antioch
(f 1 1 1 1), at Canosa (Fig. 695). This is a motive of Campanian origin. For instance,
on the Via Appia outside Santa Maria Capua Vetere stands a Roman tomb
popularly known as " La Conocchia " (Fig. 696), where the drum of the cupola has
blank arches like large round-headed windows, separated by columns. Drums encircled
on the outside by columns had been seen before this, e.g. the Choragic monument of
Lysicrates (the " Lantern of Demosthenes ") at Athens (335 B.C.) (Fig. 697), or the

Fig. 699. Cefalii. Cathedral (Xllth Century).

tomb of the Gallo-Roman period at Saint Remy, where the open drum has a conical
covering but has not got sham windows between the columns. The Byzantines did
not introduce this feature before the Xlth century. The drum of the cupola
belonging to the convent church of Myrelaion at Constantinople (919-945) has still
got only ordinary buttresses.

(7) The portal, which was a Lombardic creation, the prototype being that of
Sant' Andrea at Montefiascone (about 1032).

(8) Windows made in the form of the Lombardic portal, of which such a fine
example occurs in the cathedral of Bari (Fig. 698), erected after the destruction
of the city (1156) by the troops of William I the Bad (1154-1166). The claim
of the Lombardo-Apulian School to the authorship of this design cannot be
denied.

KN(,I,\NI)

253

(9) Rose windows, an Italian creation of the Xllth century. We noticed this
point when describing the church at Ve"zelay and San Pietro at Toscanella.

(10) External arcaded galleries, which are of Lombardic origin, as we showed in
our account of Sant" Ambrogio at Milan.

(11) The motive of several blank arches enclosed by a single arch, to be seen
e.g. on the exterior of San Nicola at Bari. The idea of a large arch containing
smaller ones had been applied by the builders of Ravenna as long ago as the
Vth century in the interior of the baptistery of Neon.

Fig. 700. Palermo. Cathedral (Xllth Century).

(12) Arched corbel courses. This form was known to the Romans, as we stated
in our account of the chapel of San Pier Crisologo at Ravenna, and from them it
descended to the Ravennate and Lombard gilds.

With regard to the plan of the Lombardo-Norman basilica, it was certainly
taken as their model (with modifications) by the builders of the Duomo of Acerenza
(1080), the second abbey church of the Trinita at Venosa (Xlth century), the
cathedral of Ccfalu (Fig. 699) founded in H3I, 1 and that of Monreale the first stone
of which was laid in 1174 in the reign of William II (i 166-1 iSg). 2 The radiating
chapels which appear in some of these churches were not of Norman origin, for this
arrangement originated in Touraine, where it was applied for the first time in
Saint Martin at Tours in the first quarter of the Xlth century.

There only remains to consider the characteristic intersecting blank arcading

1 2 Serradifalco, Del duonio di Monreale.

2 54

LOMBARDIC ARCHITECTURE

used for decorative purposes. The oldest dated specimen known to me occurs on
the interior of the choir aisles in Durham Cathedral, begun in 1093. This was
followed by the example in Norwich Cathedral where, before 1119, Herbert Losinga
introduced the motive in the transept. Such churches must have been the source
from which the creators of the cathedrals of Cefalu, Monreale, and Palermo (Fig. 700)
the last founded in 1185 by William the Good derived the motive which they
applied in such an elegant form to the exteriors of their buildings, unless, indeed,
further discoveries give some support to the idea that the Sicilian craftsmen arrived
at it independently under Moorish influence.

CHAPTER V

THE EARLY CHRISTIAN MONUMENTS OF IRELAND

following pages contain a summary of the results of a recent study
of the early Christian architecture and sculpture of Ireland. I set
them out here in the hope
that they may contribute, in

outline at least, towards a more rational

classification of the interesting and

characteristic mediaeval monuments of

the island.

Let us begin with the sculptured

crosses. One of the most celebrated

is the wheel-head cross of Muredach

(Fig. 701), as the name is given on

one of its faces, in the churchyard at

Monasterboicc which contains three

such monuments. Its carvings com-

prise religious subjects, representations

of animals, panels filled with the cable

pattern and intestinal interlacing, scroll

work with birds pecking, studs, &c.

To connect it with Muredach,

abbot of Armagh, who died in 924, 1 is

a complete mistake. The evidence of

the large number of mediaeval carvings

which I have seen and studied con-

vinces me that in the Xth century there

was no artist in existence, even the

most celebrated of Italy, France, or

Germany, capable of producing work

of this kind, far surpassing, as it does,

both in design and execution, the best

results which these countries have to

show right up to the close of the Xlth

century. It is equally impossible to ascribe it to an Eastern hand, lor the Eastern

sculptors of the Middle Ages did not produce, squat figures of this type. On the

* The illustrations to this chapter (except Figs. 704 and 707) are /rom photographs kindly provided by the
Dublin Museum.

1 Annals of the Kingdom of Inland by the Four Masters.

Fig. 701. Monasterlx)ice. Cross of Muredach
(Xllth Century).*

255

2 5 6

LOMBARDIC ARCHITECTURE

other hand, the representations on the 'Cross of Muredach of pairs of animals facing
one another and holding some creature or bird between their paws are undoubtedly
due to Lombardic influence. Now this motive, of Etruscan origin, did not make a
start in Italy before the Xlth century. The date of the- cross must therefore be
put at the beginning of the second half of the Xllth century. To the same period
and school belongs the other and more imposing cross at Monasterboice (Fig 702),
about 27 ft. high, wrongly assigned to the Xth century. 1

Rather later is the remarkable tall and slender cross at Tuam (Fig. 703) ;

for though the figure
of Christ is flattened, the
anatomy and the treat-
ment of the beard show
an advance over the best
figures at Monasterboice.
It was set up by O'Hoisin,
that is to say, the famous
Archbishop Aidan
O'Hoisin (1150-1 i6i), 23
whose period was distin-
guished by activity in
building. It was then
that the celebrated stone-
built castle was erected
at Tuam : " Rodericus
O'Conner rex Conatiae
castrum lapideumTuamae
construxit, quod tanquam
novum et inusitatum apud
Hibernos, castri mirifid
nomine iis temporibus
innotuit. " 4 And to the
same age belongs, if not
the completion, at any
rate the foundation or re-
foundation, of the three
churches consecrated on
the occasion of the Gene-
Fig. 702.-Monasterboice. Cross (Xllth Century). ral S y nod held at Tuam

in II72. 5

Of about the same date as the cross of Tuam is the small cross preserved
in the chapel of St. Kevin at Glendalough. Three of its sides are covered with
intestinal interlacing, and in front is a draped figure of Christ, of similar character
to the one on the cross at Tuam. The less elaborate crosses at Clonmacnoise must
also be put in the Xllth century.

To sum up, all these wheel-head crosses, a characteristic feature of which
is the gabled structure, representing the Holy Sepulchre, by which they are crowned,
and on which the Irish carvers seem to have concentrated all the artistic force

1 Stokes, Early Christian Art in Ireland.
3 4 Ware, Hibernia sacra.

2 O'Hanlon, Lives of the Irish Saints.
6 Annals of the Four Masters.

IRELAND

257

inspired by their religious enthusiasm, are later than the Norman conquest of
England (1066). They are also subsequent to the time of Archbishops Lanfranc
(1070-1089) and Anselm (1093-1109), when the Danish coast-towns of Ireland
acknowledged the spiritual supremacy of Canterbury and Rome. Patrick, who
in 1074 succeeded Donatus (Dunan) in the see of Dublin (1038-1074), was
consecrated by the English primate in St. Paul's at London. 1 And from a passage
in the "Annals of Loch Ce" we know that before 1134 the monastic rule settled
by Rome had been enforced in the primatial see- of Armagh (1134): " Imhar
Ua hAedhagain, by whom the Regies of Paul and Peter at
Ard-Macha was erected, died on his pilgrimage at Rome."*

They were the result of a national artistic revival pro-
duced by the renewal of relations with Western Europe
after the long period of isolation in which Danish invasions
and struggles, and disastrous internal conflicts, had plunged
the unfortunate country. This revival, accordingly, was a
reflex of the potent influence exercised by the art of Italy
and by the Papacy, in the era following the epoch of 1000,
on so many countries of both East and West. Not the
least important agent in this movement was the learned
Flann O'Gormain (1104-1174) who, after twenty-one years of
study in France and Saxony, exercised for twenty consecutive
years the supreme direction of the schools of Ireland. 3

So far as carving is concerned this revival cannot have
become effective till considerably after the beginning of the
Xlth century. The school of artists which in that century
produced the barbarous capitals of the church of St. Flannan
at Killaloe, built by King Brian Borumha (1002-1013), could
never have executed the crosses which we have described.
And, besides, the infantile geometrical incised ornament of
Irish churches before the Xlth century bears witness
to the want of skill which characterized the school.

Let us now turn to the buildings, beginning with
the important ecclesiastical centre of Glendalough, the
town of the Seven Churches. Among its ancient build-
ings the churches of the Rock, of Reefeart, of Our
Lady, Trinity Church, the Cathedral and its tower, and
the oratory and cell of St. Kevin, are believed to belong
wholly or partially to the time of Kevin who died
between 120 and 130 years old in 6i8.* 5 Let us take
them one by one.

I. Of the church of the Rock, said to be the oldest of Kevin's foundations,
nothing is left but the scanty remains (restored) of an aisleless nave.

II. Of the little church of Reefeart, so called as being the burial place of the
kings, or because King O'Toole (f 1010) was interred there, there exist the ruins of
the aisleless nave measuring about 30 ft. by 18 ft., with a rectangular chancel at the
east, built of roughly-hewn stones of various sizes, and of rubble with a greal deal of
mortar. This is thought to be the "clara cella" erected by Kevin and mentioned in

1 Ware, De Hibernia et antiquitatibus etus. Rolls' Series The Annals of Loch &.

* Ada Sanctorum Hiberniae Acta Sancti Caimgini. s Rolls. Series Chronicum Scotorum.

VOL. II S

- 703. Tuam. Cross (Xllth
Century). (From a cast in the
Dublin Museum.)

258 LOMBARDIC ARCHITECTURE

the following passages : " qui claram cellam Domino aedificavit," 1 " clara illic
cresceret cella, que Desertum Caymginii appellatur." '

III. The Lady Church is represented by the remains of the nave and square
chancel. The latter is the result of an alteration carried out when the building was
dedicated to the Virgin, which, according to Petrie, 3 took place not before the Xllth
century, because before that time no church in Ireland was dedicated to the Mother
of God or to non-Irish saints. The round-headed window, decorated on the outside
by a carved band springing from two heads, exactly suits the Xllth century. Its
erection must have taken place after the fire of 1 163.* On the same occasion the
window with a hood mould was inserted in the south wall of the nave. The church
is said to have been built by Kevin's orders, with the direction that he was to be
buried in it ; 5 but this is a mere tradition, and another points out an oratory or small
chapel as his burial place. 6

IV. Of Trinity Church we have the ruins of the nave and rectangular sanctuary,
with a square porch once surmounted by a round tower. The oldest part is the
nave. The chancel and porch are later additions.

V. Of the Cathedral there remain the ruins of the nave, considerably larger than
that of any other of the Glendalough churches and about 30 ft. wide, and those of
the chancel which is the result of an alteration probably carried out when the abbey
of Glendalough was erected into a bishopric under Pope Alexander III (1159-1181).

Close by stands the ancient, picturesque, round tower. Both cathedral and
tower are believed by Petrie 7 to have been erected by Gobhan Saer (about 610), on
account of the analogies between them and the church and tower of Kilmacduagh
which tradition ascribes to him.

VI. The two-storied Oratory of St. Kevin is preserved almost intact. In the
course of time there were added to it the chancel, which has disappeared, and
the round tower and sacristy which still exist.

VII. St. Kevin's Cell, believed to have been the first monastic cell erected by
him, consists of the remains of a structure of bee-hive form.

In all these buildings the windows are narrow, and sometimes square-edged, but
usually splayed on the inside. The doorways in some cases diminish in width
towards the top. Those in the Cathedral and St. Kevin's Oratory have also a
tympanum above them, while the one in the Lady Church has the opening framed,
with a cross carved on the architrave.

In view of the almost uniform masonry of the roughly constructed walls, and
the excessive poverty of the architectural decoration, confined as it is to two rude
lunettes, and considering the almost entire absence of artistic ornament, consisting
merely of a poor moulding round a door and a cross, the problem of fixing the date
of these structures is by no means easy. We will, however, attempt to solve it, and
we may begin by classifying them under three heads. To the first belong the
churches of the Rock, of Reefeart, of Our Lady, the Trinity, and the Cathedral ; all
of them built of masonry, but without solid roofs. A second category is formed by
the vaulted buildings, represented by the tower of the Cathedral and the Oratory of
St. Kevin. Kevin's bee-hive cell forms a class by itself.

1 Ada Sanctorum Tertia dies Junii. " Ada SS. Hiberniae Ada S. Caimgini.

3 The Transactions of the Royal Irish Academy, Vol. XX. An inquiry into the origin and uses of the
round towers of Ireland.

4 Annals of the Four Masters.

5 Antiquarian Handbook Series, No. I Dunsany, Tara, and Glendalough.

6 O'Hanlon, ot. cit. * Op. cit.

IRELAND 259

It appears to me that the monuments of the first class are the oldest, but still
not so old as the time of St. Kevin. We have already seen that the churches
erected in the " Celtic " or " Scottish " manner in England by the missionaries sent
forth from lona and Lindisfarne or its daughter monasteries were constructed of wood
(it is even suggested that they may have been of osiers plastered with mud), and
originally covered with reeds or thatch, and afterwards with lead. It was a style of
building carried even to Italy by Columban, as the church of the Virgin at Bobbio
showed. In Ireland the adoption of masonry in place of timber for ecclesiastical
buildings must have been a consequence of the burnings due to the Danes who
invaded the island from 794 onwards, 1 not to speak of those caused by the Irish
themselves. It is true that the Annals of Ulster mention under the year 788 a
stone chapel, to which Petrie refers. 2 It may have been merely a dry-stone structure
of the bee-hive type. In any case one swallow does not make a summer. Inveter-
ate customs are very reluctantly abandoned, and only under the stress of dire
necessity.

The earliest record of Glendalough suffering from devastating fires at the hands
of the Irish and Danes 3 occurs in the year 770. A similar calamity happened in
835.* From this we may infer that the sacred structures were in the interval hastily
rebuilt of wood. It must have been only after the second destruction that it was
decided to replace the timber construction by masonry, "iuxta Romanorum
morem " as Bede describes it. 5 It was under these circumstances that the churches of
the Rock, of Reefeart, Our Lady, the Trinity, and the Cathedral, may have been
erected with stone walls and roofed with reeds or some kind of thatch. It was a
form of construction which was an improvement on the use of wood only, for if it
happened to be burned it was capable of being repaired, instead of requiring
an entirely new erection. The Cathedral was, perhaps, somewhat different, as
it may have possessed a timbered roof covered with sheets of lead, like the
great church at Armagh which was also constructed of stone with a lead roof,
and was burned with its bell-tower and bells in IO2O. 6 As we have seen,
this type of roof had been adopted for the cathedral of Lindisfarne between 687
and 697.

I have mentioned both systems of roofing because they are both " Celtic,"
and also because the gables of these Irish churches show no traces of a junction
with a stone roof. And, in any case, the width of the cathedral forbade the
construction of a solid covering. The annalists, moreover, confirm the theory that
the Glendalough churches as a rule had neither vaulting nor stone roofs. Thus,
in 835 the Danes burned the " oratorium," which must mean the monastery church.
In 1020 they burned "the oratories." In 1061 the "churches" were accidentally
consumed by fire, and again in 1084. And in 1163 the "House of Kevin" ("Cro-
Chaeimhghin ") was burned together with the " church of the two Sinchells." 7 This
" House of Kevin " must have been the abbot's residence, for it is not conceivable
that an ordinary dwelling, liable to be destroyed by fire, should have been
preserved intact through all the series of disasters which befell Glendalough from
770 onwards.

This system of building in stone with a roof of combustible material remained
in vogue. Thus we hear that in the Xlth century to be precise, in 1058 and 1060

1 Annals of Ike Four Masters. 2 Op. cit.

3 * Annals of the Four Masters. " Vita abbalum.

Rolls Series The Annals of Loch Ci. ' Annals of the Four Masters.

S 2

260

LOMBARDIC ARCHITECTURE

respectively " Imlech-Ibhair was entirely burned, both stone church and steeple,"
and " Cenannus was altogether burned, together with its stone church." l

After the reign of Edward the Elder (901-925), when, as we noticed at the
time, the efforts of the Anglo-Saxons against the Danes in England were accom-
panied by the construction of strongholds with towers as an effective method of
successfully opposing the barbarians, and as an echo of the great building era
in England which distinguished the reign of Edgar (959-975), and, perhaps too, in
consequence of fresh disasters which befell Glendalough in 977, 982, 984, and 985,

^^^^^^ the erection of the cathedral tower must
have taken place.

This tower (Fig. 704), which is some
100 ft. high from base to summit, con-
structed of roughly hewn stones of all
sizes and rubble set in mortar, covered
by a circular vault, and originally divided
internally into floors, the holes for the
beams being still visible all round, served
for the various purposes of a bell-tower,
an outlook, and a stronghold and place
of refuge for the lives and property of
the monastic body. This fact was first
brought out by Petrie, 2 who is only mis-
taken in the date of this and similar Irish
towers which he ascribed to the centuries
between the Vth and XHIth. That it
was intended to hold the bells is shown
by the four openings at the summit. The
purpose of defence and refuge is indi-
cated by the entrance being placed at the
height of some yards above the ground.
It was reached by a ladder. The exist-
ence of wooden floors inside is proved not
only by the holes for the beams, but also
by the fact that another tower, that of
Telachaird, was burned in 1171 with all
the unfortunate people who had taken
refuge in it. 3

The painful experience of the past
was, no doubt, the convincing argument in
favour of the adoption of this form of structure, impervious as it was to any attempt
of an enemy to set it on fire. But I do not believe that it could have taken place
until the Irish builders had obtained considerable practice in masonry construction,
and had erected some vaulted buildings. The vault of the tower and the tower is
so solidly built that the wind was able to tear off the conical roof without damaging
the rest though forming a cupola on a circle of only about 8 ft. in diameter, must
have demanded a Herculean effort of constructive science on their part. The
difficulty arose from the fact that they were not accustomed to vaulting ; and that,

Fig. 704. Glendalough. Cathedral.
(Xth or Xlth Century).

Round lower

1 Rolls Series The Annals of Loch Ce.

3 Annals of the Four Masters.

Op. fit.

IRELAND

261

by making their chancels rectangular, they had avoided the constructive difficulties

inherent in the circular form of apses and half-domes, requiring as they do specially

prepared materials and a higher degree of skill.

The typical form of the Glendalough tower, related

as it is to the bell-towers of Ravenna and also to

the staircase towers with their conical cupolas in San

Vitale, was undoubtedly an importation from Italy. If

we could fix with certainty the date of the foundation of

the Irish Colony at Poggio de' Berni in the district of

Sant' Arcangelo (Forli), which is described as " Podium

Hibernorum " and said to be of ancient origin, 1 some

interesting light might be thrown on this importation.

In any case, Continental influence on Irish architecture

from the Vth to the XVI Ith century has been admitted,

even by recent writers. 2

To about the same date as that at Glendalough be-
long the towers of Monasterboice and Antrim (Fig. 705).

To a later period, but before the erection by Ua Maeleoin

of the tower at Clonmacnoise (finished in 1124: its top

was destroyed by lightning in 1135*), which has a finished

facing and is built with regular courses of carefully laid

oblong blocks of stone, will belong the round towers of

Disert Aengus (Fig. 706), Scattery Island, and Station

Island in Lough Derg, the masonry of which, though

still rough, is more regular
than that of the towers de-
scribed above.

These round towers con-
tinued to be erected in Ire-
land for a long time. Thus,
the one at Ardmore, about

108 ft. in height, built of oblong blocks of stone and with
the exterior marked off into zones by stringcourses, is
not older than the erection of the church, that is to say
the end of the XI Ith century. 4 This need not cause
surprise, seeing that the Danes who had established
themselves in Leinster before 85 1 5 were not finally
annihilated till 1171, when the Anglo-Normans van-
quished the fleet of Asgall and put him to death, 6 and
that disastrous internal struggles had not ceased to rage
in the island.

About the same date as the tower of Glendalough
is the oratory of St. Kevin known as " St. Kevin's House "
or "St. Kevin's Kitchen" (Fig. 707). This is a chamber

1 Calindri, Saggio statistico s/orico del rontificio Stato.
A. S. Green, The making of Ireland and its undoing.
Annals of the Four Masters.
Dunraven, Notes on Irish architecture.
Fig. 7o6.-Pi.sert Aenyus. Round W " e . De Hibernia et antiquitatibus tin,.

tower ( X I th Cent ur y ). Annals of the Four Masters.

Fig. 705. Antrim. Round tower
(Xth or Xlth Century).

262

LOMBARDIC ARCHITECTURE

Fig- 707. Glendalough. Oratory of St. Kevin (Xth or Xlth Century).

(about 30 ft. X 2 1 ft.)
of two stories, one
of which has a barrel
and the other a
pointed-arched vault.
The upper supports
the gabled masonry
roof, covered with
stones which form a
continuous structure
with the vault. The
walls are built of
stones of all sizes
and rubble set in
mortar. To this
chamber there was
added later a rect-
angular chancel, now
destroyed, flanked
by a sacristy which
survives. Over the
Three holes

west end rises a round bell-turret which breaks the vault of the roof,
for the bel! ropes are pierced in the barrel vault of the lower story.

This chamber was not built for the double purpose of an oratory below and a
dwelling room above, as has been suggested. The opening now existing in the
barrel vault, intended to form a communication between the two stories, is the result
of an alteration. The upper vault, too, was not constructed to provide a tiny
dwelling with a water-tight covering,
but to carry the sloping sides of the
heavy roof, and provide something to
intercept the weight and take the
pressure off the barrel vault below.

The constructive and statical know-
ledge here displayed something quite
exceptional among the builders of Ire-
land with the object of making the
structure as safe as possible from the
assaults of time and ths violence of
man, points to workmen about contem-
porary with those who built the cathe-
dral tower. The presence of a lunette
over the west door (indicative of Pre-
Lombardic influence), and the care
taken to secure the building as far as
possible from injury, suggest that it
had a very sacred character, viz. that
of an oratory erected on the site of the
primitive wooden dwelling of the saint,
which had perished by fire, thus con- Fig. 708. Kells. Oratory or St. Columba (Xth Century).

IRELAND

263

firming the popular title of "House" or "Kitchen of St. Kevin" which is believed to
have its source in the ancient tradition that he lived there for the last years of his
life. 1 At a later date the oratory was converted into a church. The saint's dwelling
must have been the one erected in the monastery " of the valley of the two lakes,"
which was Kevin's last foundation : " Post hec venerabilis pater insignissimum
monasterium, quod Vallis duorum stagnorum dicitur, illic construxit." 2

Of the same type as this building is the well-known "St. Colum-Cilles House,"
i.e. the oratory of St. Columba, the apostle of the Scot! in Caledonia, who died in 597,*
near Kells (Fig. 708). It,
too, must have been built
as an oratory, and earlier
than the one at Glenda-
lough, because the door-
way is without a tym-
panum. These two build-
ings must be older than
the equally well-known
church of St. Flannan,
standing close to Killa-
loe Cathedral which was
erected by Donnell More
O'Brien (f 1194), king of
Limerick. 4 & The founda-
tion of this church is
ascribed to the year 1007
and the agency of Brian
Borumha, 8 whose reign
lasted from 1002 to 1013 ; 7
and its western door (Fig.
709), which has multiplied
archivolts springing from
two short jamb shafts sur-
mounted by rude Corinth-
ianesque capitals, shows
an advance beyond the
earlier doors at Kells and
Glendalough. This doorway is, in its turn, a reflex of the Anglo-Saxon form of
opening recessed in several orders.

The type of these three structures was reproduced in the celebrated and singular
Chapel of Corinac on the Rock of Cashel, built in the Lombardo-Norman style with a
groundwork of Irish character (Figs. 710,711). It has square towers flanking the
east end of the aisleless nave, a rectangular chancel, from which projects an altar
recess of the same form, and recessed openings of Lombardic type but Norman
decoration. It has been said that it was the work of Cormac MacCullenan (f 908) ;
but it was really built by Cormac MacCarthy in 1127, and consecrated in 1134* or

1 O'Hanlon, of. fit. * Acta SS. tfioerniacAcla S. Caimgini.

3 Stokes, Ireland and the Celtic Church (revised by H. J. Lawlor). 4 Dunraven, ef. ci.

* * Petrie, of. fit. 7 Annals of the Four Masters.

8 Petrie, op. fit.

Fig. 709. Killaloe. Church of St. Klannan. West door (1007).

264

LOMBARDIC ARCHITECTURE

Fig. 710. Cashel. Chapel of Cormac (Xllth Century).

II3O, 1 or, according to another
account, in II35- 2 The presence
of the Lombardic cubical capital,
which made its earliest appear-
ance in 1013 in Sant' Abondio
at Como, makes a Xth century
date impossible. Moreover, the
Lombardic openings, the en-
riched blank arcading of both
interior and exterior, the figure
corbels, and the quality of the
vaulting of the ground floor,
demonstrate undeniable Lom-
bardo-Norman influence of con-
siderably later date than 1066.

We have still to discuss the
Cell of St. Kevin. Its date is
not an easy problem to solve,
but it must belong to a period
subsequent to the first Danish

invasions, for it is not likely that monastic cells were being built of masonry at

a time when churches were constructed of wood.

In Ireland "bee-hive" structures start with the tomb chambers in tumuli, like

those of New Grange, Dowth, and Knowth, the first being the finest. 3 4 These are of

circular plan and bee-hive section, constructed with rough dry-stone walling, which

gradually converges so as to form the vault above. Others are found, of ancient

date, constructed either with or without mortar, which served as oratories or dwellings.

Bee-hive cells of very early date

exist on Skellig Michael, one of

which has formed a chapel. But

they are not so old as is generally

supposed. The first planting of

this monastic colony on the Great

Skellig was surely subsequent to

St. Aubert's foundation of Mont

Saint Michel (about 708), which

in its turn was derived from San

Michele on the Monte Gargano.

The founder is said to have been

the abbot St. Suibhneus, but we

1 Rolls Series Chronicitm Scotomm.

2 A'oHs Series The Annals of Loch

3 The Transactions of the Royal Irish
Academy, Vol. XXX. Coffey, On the
tumuli and inscribed stones at New Grange,
Dowth, and Knowth.

4 Jotirnal of the Royal Society of Anti-
quaries of Ireland, Vols. IV., V., VI.,
VII. Coffey, The origin of pre-historic

ornament in Ireland. Fig. 711 Cashel. Chapel of Cormac Xllth' Century).

IRELAND

265

Fig. 712. Gallerus. Oratory (IXth or Xth Century).

do not know his date. 1 Destroyed
in 812 by the Danes, who starved
the monks to death, it was rebuilt
in 860. Subsequently, an abbot,
Flann MacCellac (f 885), is re-
corded. The date of the removal
of the house to Ballineskellig is
not known. 2 In any case, the
oldest structures of Skellig Michael
are later than the rebuilding of
860. Their form is to be explained
by the fact that it was easier for
the monks to procure stone than
timber.

Of the same type is the well-
known Oratory of Gallerus (Fig.
712), entirely constructed without mortar, and with the vault almost resting on the
ground. Its date is not known, but the greater skill displayed shows that it is later
than the structures on the Great Skellig. Perhaps it belongs to the end of the
IXth century, or the first half of the Xth ; but it is certainly not of the age pre-
ceding the apostolate of St. Patrick (432~46i), 3 as Petrie* imagined. The method
of construction may well be due to the difficulty of procuring mortar, and reasons of
economy.

Later than the type of building represented by the oratory of Gallerus is the one

exemplified by the chapel on
St. Macdara's Island, in which
the low side walls of the nave,
projecting beyond the line of
the front and end, originally
carried a high-pitched stone
roof, the junction of which with
the gables may still be traced.
For it was one thing to raise a
converging vault from the firm
ground, and quite another
matter to cover a space with a
heavy roof of masonry high up.
From a combination of
these two types was evolved a
third, viz. that with two vaulted
stories, the earliest examples
being the oratories of Kells,
Glendalough, and the church at
Killaloe. This type, together
with that of St. Macdara's

- 7'3- Kilmachedar Church (Xlth or Xllth Century).

oratory, appearing in a perfected form in Kilmachedar Church (Fig. 713), went on
being improved through the Xlth and Xllth centuries.

Before leaving the subject of Ireland I may notice that the crypt of Christ
1 a Archdall, Monasticon Hibernicum. ' Bury, op. cit. * Op. fit.

266 LOMBARDIC ARCHITECTURE

Church (Trinity) Cathedral, Dublin, is thought to represent with its vaulted con-
struction (though its form has been changed) the original church founded by King
Sihtric (f 1041 or 1042) and the first (Danish) bishop, Donatus or Dunan (1038-1074),
about the year 1038, or that, at any rate, it preserves the exact plan of that erection.
This idea is based on a passage of the " Liber niger " of Christ Church : " Sitricus
. . . dedit S. Trinitati et Donato primo episcopo Dublin(ensi) locum ad aedificandum
ecclesiam S. Trinitatis, ubi fornices sive voltae sunt fundatae." 1 2 3

An examination of the structure has convinced me that the remodelled crypt is
the result of the rebuilding of the church carried out about 1170 in the time of
Strongbovv and Archbishop Laurence O'Toole (1162-1180). In the first half of the
Xlth century no church of this size and form could have been erected in Dublin.
Neither the Danes in Ireland, nor the Irish themselves, accustomed as they were to
ecclesiastical buildings of quite another type, and to the erection of structures of
very modest dimensions, would have been capable of performing the task. Nor, for
that matter, would English builders of the time have been in any better position.
And there is no record that Sihtric during his pilgrimages to Rome in 1030 and
1035 4 engaged the services of Italian workmen capable of carrying out his
intentions. Nor could the constructive skill then to be found in Normandy, thanks
to William of Volpiano and his pupils, have been utilized for the occasion, for it is
out of the question to suppose that the Benedictines would have placed their
services at the disposal of Bishop Dunan who had handed over his cathedral to
secular canons.

1 Archdall, op. cit. " Ware, Hibernia sacra.

3 Ware, De Hibernia et antiquitalibus eius. * Archdall, of. cit.

CHAPTER VI

THE ECCLESIASTICAL ARCHITECTURE OF GERMANY
FROM CONSTANTINE TO THE ELEVENTH CENTURY

IN the lands comprised within the German Empire of to-day the list of
surviving churches belonging to the centuries between Constantine's grant of
peace to the Christians (313) and the fatal catastrophe of the Roman Empire,
long tottering under the weight of its own greatness and its inherent vices,
and from that gigantic upheaval down to the reign of Charles the Great (768-814), is
confined to a single building, the cathedral of Trier, and that not in its original
condition. Moreover, it was never a structure erected as a whole for its purpose, being
merely the result, in the first instance of an adaptation, and then of a restoration.

THE CATHEDRAL OF TRIER was formed by Agricius (who, according to
Gams, 1 held the see from 314 to 332) in the hall of a Roman building supported by
four lofty columns
united by arches,
which he dedicated
to St. Peter. This
adaptation of a
building of no
great size, instead
of the erection of
a spacious basilica,
must, considering
the importance of
Trier, have been
due to the fact
stated by Har-
nack 2 that at the
beginning of the
IVth century the
number of mem-
bers of the local
church was still

small. The cathedral was damaged by the Franks, and restored by Bishop
Nicetius (527-535-566).* *

On the north side of the church some remains of the Roman building may be

1 Of. tit. s Op. cit. * Man. Gertn. hist. Gesta Trcverorum.

* Browerus, Anttquitales et Annales Trrvirtnses.

26;

Fig. 714. Trier. Cathedral. Roman work on north side (IVth Century)-

268

LOMBARDIC ARCHITECTURE

seen (Fig. 714). The construction is of stone with courses of brick, and there may-
be noticed a continuous band of tiles which follows the line of the alternate

Fi 8- 7'5- Trier. Remains of Imperial Palace (IVth Century).

triangular and rounded heads of the niches and openings on the ground floor
of the building. It looks as if it were intended to break the plainness of the wall,
which is quite devoid of ornaments in relief. The structure may be assigned to the
times of Diocletian (284-305), Maximian (286-310), and Constantine (306-337),
the period of the city's greatest splendour ; l or, more probably, to the reign of the
last, and not long before its conversion into a church by Agricius. It certainly
is not as late as the year 370, as Dehio and Von Bezold 2 and others imagine ; for

that date conflicts with the account in the " Gesta Treve-
rorum," and its masonry is evidently contemporary with
that of the three-lobed structure belonging to the Imperial
Palace (Fig. 715). This is also faced with bands of stone
alternating with bands of brick, and is ascribed to the age
of Constantine ; rightly, I think, on account of the window
arches which, though they have not the comparative finish
of the time of Diocletian, show no signs of the marked
decadence of the post-Constantinian epoch.

The recent restoration has thrown light on the original
construction of the interior of the cathedral, where the round
arches are outlined by a ring of bricks laid horizontally,
whereas those belonging to the adaptation are copied from
the old ones, but without the ring of bricks.

To the alterations of Nicetius belong two capitals (Fig.

Fig. 716. Trier. Cathedral.
Capital (Vlth Century).

716), now built into the wall, which formerly surmounted two of the four supports
of the central quadrangular space. They are imitations of the antique, of Corinthian
pattern, with plain, stiff leaves, and are rude and poor work. They would be of
interest if they were made on the spot, as providing evidence about the state of

1 Browerus, Antiquitates et Annales Trevirenses.

Op. cit.

GERMANY

269

carving in the Vlth century in an important artistic centre such as Trier was under
Roman rule, and as showing the type of capital in vogue there at the time. But
as we know from a letter of Ruffus, bishop of Turin (560-570), to Nicetius that the
latter invited craftsmen from Italy to repair the damage inflicted on the churches
of Trier by the barbarians (". . . artifices de partibus Italiae accitos ... ad vos,
Domino ducente, transmisi " J ), probably it is they who are responsible for these
capitals.

# *

We have already given a brief sketch of the ecclesiastical architecture of the
Prankish Empire under Charles the Great in Italy, Dalmatia, and France. We will
now extend our sur-
vey to the German
lands, beginning with
the most celebrated
of the Carolingian
buildings, the palace
chapel of Aachen.
The restoration in
progress has made it
possible for me to
examine it stripped
of all accretions, and
to penetrate the in-
most secrets of its
organic structure.

THE PALACE
CHAPEL AT AACHEN
was erected by
Charles the Great
between 796 and
804, and dedicated to
the Virgin by Pope
Leo III early in
805 . 2 s 4 The plan is
that of a polygon with
sixteen sides, four of
which are taken up
by the sanctuary, the
narthex.and the stair-
case towers. On the
ground floor the

arches, strengthened by substantial sub-arches, which open out of the central space
(Fig. 717) are carried on piers of broken outline, recalling (in section) that of the
piers in San Vitale at Ravenna, which, in their turn, present a striking analogy to

1 Man. Germ. hist. Epislolae Mermvitigici et Aarolini aevi.

5 Mon. Germ. hist. Einhardus, I'ila Karoli imperatoris.

3 Man. Germ. hist. Annales Tielenses. 4 Jafle, Retsta fontificiim Romantrrtiw.

Kig. 717. Aachen. Palace Chapel (796-804).

2 7

LOMBARDIC ARCHITECTURE

Fig. 718. Aachen.

Palace Chapel.
(796-804).

Vaulting of aisle

others of the Roman period, e.g. those
of an octagonal building near Pozzuoli,
the plan of which has been preserved
by Montano. 1 These arches, together
with the blank wall-arches resting on
powerful wall-piers (which have no but-
tresses corresponding to them outside,
the outer face of the walls being un-
broken, and the walls themselves over
5 ft. thick) sustain a continuous series
of unraised tripartite and quadripartite
cross vaults, some 2 ft. 4 in. thick at
the crown (Fig. 718).

This system of thrusts met by the
outer walls, strengthened on the inside
by pilasters or even columns (either
engaged or set against the wall) is
sometimes described as " Byzantine."
As a matter of fact the Byzantines
borrowed it from the Rome of the first
three centuries of the Empire. The
city and its environs still contain the

proofs of this for any one who cares to ascertain the facts, in the shape of tombs,

the Thermae, the Basilica Nova or Basilica of Constantine (310-312), not to speak

of the abundant evidence provided by old drawings.

The original presbytery was in two stories, and of rectangular plan. Two spiral

staircases, formed in the towers which flanked the narthex and Imperial tribune, lead

to the latter, the gallery, and the corri-
dor communicating with the Imperial

Palace. These staircases have rude

vaulting, and terminate at the height

of the roof, and below the raised part

of the wall, in a round vault as in San

Vitale at Ravenna. Of the same kind,

but only intended to provide access to

the roof, were the two staircase towers,

circular after the Ravennate type, in the

front of the abbey church of St. Gall.

The tribune is of rectangular shape with

a rounded end, and has a barrel vault

constructed, like the walls, and, indeed,

the whole of the interior facing of the

building, of dressed stone brought from

Verdun (" De quadratis autem lapidibus

dirutae civitatis [Virdunicae] Aquisgrani

capella extructa est" 2 ). The jambs of

1 Op. dt.

2 Bouquet, Kerum Gallicarum et Francicarum

scriptures Ex chronico Virdunemi, auctore Hugone pig. 719. Aachen. Palace Chapel. Vaulting in

abbate Flaviniacensi. the gallery (796-804).

GERMANY

271

the two doors leading into it have long and short work, while the voussoirs of the arch
are of white and grey stone alternately. Two low doors lead into the gallery.

When discussing Saint Benigne at Dijon (1002-1018) we remarked that the
Eastern origin ascribed by so many writers to galleried basilicas is quite arbitrary.

The arches of the gallery are carried by piers of the same form as those below
Each arch contains a
screen of two tiers of
columns, which are not
original. Of all the old
capitals belonging to the
gallery and tribune (de-
rived from earlier build-
ings like the columns
themselves : " Ad cuius
structuram cum colum-
nas et marmora aliunde
habere non posset, Roma
atque Ravenna deve-
henda curavit" 1 ) there
survive in the whole
gallery only three of
Corinthian form and late
Roman date ; and they
have been restored. The
idea of filling up the
arch openings with
screens of isolated
columns is of Roman
origin. It appeared fre-
quently in the Thermae
of Rome under the
Empire.

From the piers and
the pilasters of the outer
walls, which are about
3 ft. 3 in. in thickness,
spring rude visible trans-
verse arches with vous-
soirs of various kinds of
stone brought from else-
where, like all the arches in the building. Upon these arches are turned barrel vaults
alternating with vault cells (Fig. 719). These vaults, like those of the ground
floor, the staircases, the Imperial tribune, and the dome, are roughly and coarsely
constructed of pieces of limestone set radiating, with above them a bed of concrete
composed of lime, sand, gravel, and pounded bricks, of the kind used at Rome and
Ravenna.

The two original windows exposed by the restoration are round-headed and
splayed on the inside, where the jambs have the long and short work which we

1 Man. Germ. hist.Einhardus, Vita A'aro/i impcratoris.

Fig. 720. Aachen. Palace Chapel (796-804).

272

LOMBARDIC ARCHITECTURE

noticed in the entrances to the tribune, and also to be seen in the original open-
ings of the staircase towers. Whence we may reasonably infer that all or most
of the openings of the rotunda were constructed on the inside in just the same
way.

Above the arches of the gallery rises the octagonal drum, and upon that the
cupola of the same shape. It is conical in form, about 3 ft. 3 in. thick at the crown

and originally covered
by a roof which, with
the walls of the drum,
was raised in height
in the XI I Ith century.
The blank arcading of
this addition has
spurred bases to the
shafts. Perhaps these
were the source of the
erroneous statement
(for which the study
of books instead of
the monuments them-
selves is responsible)
that this detail first
appeared on bases in
Roman times, and
next in the rotunda
of Aachen. As a
matter of fact, its
creation is not earlier
than the Xth century.
Unlike the lower
octagon the drum
(Fig. 720) is strength-
ened close to the
salient angles of the
exterior, and almost
up to its summit,
by buttresses sur-
mounted by capitals
rudely carved with foliage. This device must have been chosen in preference to angle
buttresses, with the object of increasing the field of resistance.

The rotunda was approached through a large cloister court or quadriporticus,
remains of the foundations of which have been discovered. 1 On the north side of the
great frontal recess or niche (Fig. 721) there remains one of the original windows of
the barrel vaulted corridor which connected the gallery of the rotunda with the
Imperial residence. The architrave of the lunette of this window is supported by an
ill-formed fluted pier with moulded base and capital, the outer face of which has now
been turned inwards, and the present outer one has been re-worked.

The Minster of Aachen as a whole is not so much an original creation

1 Buchkremer, Zur IViederherstellung des Aachner Miinsters.

Fig. 721. Aachen.

(796-804).

GERMANY

273

as an imitation of San Vitale at Ravenna, an edifice which Charles had had
an opportunity of admiring during his visit to the city in 78;. 1 It belonged
to a style which, though it had obtained recognition in Italy, at Ravenna and Milan,
had encountered an obstacle to its wider acceptance in the shape of the ancient
basilica plan on which the Latin Church had set the seal of its approval. We
therefore cannot imagine that the Emperor, who was surely not unaware of the
reasons which had hitherto prevented the spread of the Byzantino-Ravennate

Fig. 722. Constantinople. St. Irene (VHIth Century).

style in Italy, though still the most civilized country of Western Europe, and
containing the powerful building gild of the Comacini as well as the ancient
and still fairly active school of craftsmen at Ravenna, would have cherished the
vain hope, as some believe, that the erection of a building in that style would
produce throughout the vast Prankish Empire, as if by a touch of a magic wand,
craftsmen with the skill to develop and diffuse a type of construction so contrary
to the artistic traditions both of its inhabitants and of the Latin Church of which,
willing or unwilling, they were the faithful adherents.

The fact is, this style, in which the vault is applied only to ground plans

1 Man. Germ. hist. Agnelli liber tontijicalis.
VOL. II T

274

LOMBARDIC ARCHITECTURE

of a certain form, the principal being the circle, the square, and the polygon,
and of which the primary source is the dome, was not successful, or very rarely so,
in gaining a footing in Italy and Northern Europe. And when in the Xlth century
a new age demanded a new style of architecture, it was neither the Byzantino-
Ravennate, nor the Byzantine style pure and simple, which presented itself
as best suited to the tastes and needs of Western Europe, but the Lombardic,
born in the West, originating in the application of the vault to the Latin basilica,
created by the gilds of Lombardy, and fashioned by the Benedictine Order into
the forms which we find in the countries north of the Alps.

No information has reached us as to the architect of this celebrated church

Fig 723. Constantinople. St. Irene (VHIth Century).

or the builders who worked under him. But there are grounds for believing
that it was erected from the designs and under the superintendence of Byzantine
architects, and carried out by Italian masons assisted by Prankish workmen.

The Byzantine School is revealed in the statical principles exhibited by the
structure, and also by the heaviness of the internal construction, the latter being
a characteristic feature of Byzantine churches of the VII I th century. This
may be seen from St. Irene at Constantinople (Figs. 722, 723), which is not the
church rebuilt by Justinian I (527-565), but a reconstruction begun by Leo III the
Isaurian (717-740), after its destruction by the earthquake of 739 ; 1 - 3 not a mere
restoration as is generally believed. I have formed this conclusion after making
under great difficulties, a thorough examination of the building. The Ionic capitals
with pulvins belonging to the eight columns which support the galleries under the

1 Du Cange, Historia ByzantinaConstantinopolis Christiana.
* Van Millin^en, Constantinople. ' Bury, A History of the later Roman Empire.

GERMANY 275

dome, are poor work and certainly not of the age of Justinian. Further, it is enough
to look at the cupola, not lighted by small windows like St. Sophia and SS. Sergius
and Bacchus at Constantinople, and not springing from a low drum like that
of St. Mary Diaconissa in the same place, built by the patriarch Cyriacus in the
reign of the Emperor Maurice (582-6O2), 1 but rising from a high drum, strengthened
by buttresses outside, and lighted by lofty and wide windows like those in the
Nymphaeum of the Licinian Gardens (253-268) and the Imperial Mausoleum
(Vth century) by St. Peter's at Rome.

With regard to the actual masons, we know that, of all the countries then subject
to the rule of Charles the Great, Italy was the most capable of providing them. The
guard-house of Theodoric's palace at Ravenna (VHIth century), and especially
Santa Maria in Valle at Cividale (762-776), are convincing evidence of the capacity of
the builders of Ravenna in the VHIth century. And again, the structures erected in
the Lombard part of Italy in the course of that century by the hands of the Comacine
masters, as well as the fact that Hadrian I asked Charles the Great to send him
a master mason (" . . . prius nobis unum dirigite magistrum " 2 ) to renew the timbered
roof of the Vatican Basilica, are so many testimonies to the ability in matters ol
construction of the members of the gilds.

We may suppose that the " master " referred to was one of the Comacini, seeing
that, during the period of great constructive activity which comprised the pontificates
of Hadrian I (772-795) and Leo III (795-816), the Lombard gilds left undoubted
traces of their presence both in Rome and in other towns of the Roman Duchy as it
existed in the time of Charles the Great. Another consideration is the fact that the
gilds in question were better known than any others among the Emperor's subjects.

Nevertheless, the Ravennate and Comacine craftsmen, with few exceptions and
those of early date, familiar only with the easy field of the old Roman basilica design,
cannot but have been dismayed when brought face to face with the problems of
scientific construction, and with the practical task of building a vaulted structure of
the type of the Imperial chapel. It is a reasonable inference that the direction of
the work was not entrusted to any of these master masons, though at the same time
it is natural that their services should be engaged for its execution (with the assistance
of Prankish workmen for the simpler parts), whereby the great expense of hiring
Byzantine craftsmen, as some think was the case, would be avoided. That masons
of Ravenna did take part in the work is shown by the use of mortar of the Roman
and Ravennate kind. On the other hand, the long and short work in the openings
reveals the share of Prankish workmen, who, as we saw in our account of St. Peter's,
Monkwearmouth (675), were responsible for its introduction into England. The fact
that the dome was covered by a timbered roof makes the presence of Eastern builders
doubtful, and rather points to those of Ravenna, whose predecessors had treated the
cupola of San Vitale in the same way.

This employment of Italian workmen on the largest and most perfect of Charles's
buildings, though it had not the marked direct effect on the Lombardic and derived
styles that has been often attributed to it, still exercised an influence which, though
indirect, was considerable. In the course of its erection the Comacine masters gained
a familiarity with vaulting construction such as they had never been able to do before.
On their return home, fortified by the lessons they had learned and the experience

1 Du Cange, Historia Byzantina Constantinopolis Christiana.

3 Duchesne, Historiae Francorwn scriptures Epistolae summorum pontificum ad principes et reges
Francorum.

T 2

276 LOMBARDIC ARCHITECTURE

they had acquired, after some further practice, in co-operation with the masters
of Ravenna, in buildings of this type on the coast of Dalmatia, they devoted them-
selves to the researches and experiments which resulted in the creation of the
Lombardic vaulted basilica.

Strzygowski 1 believes that Charles the Great's church followed Eastern models
then to be found in the Gallo-Frankish lands. Having made it my practice to base
my opinions on the evidence of wholly or partially existing buildings, or of those
which have come down to us through drawings or descriptions, I regret that I
am unable to accept this hypothesis. And I can only hope that German scholars,
devoted as they are to facts, will not blame me for so doing. The Gallo-Frankish
countries contain no such types. Unless, indeed, we were to make the mistake of
regarding as one the three-lobed vaulted Roman structure at Trier, the similarity
of whose plan with that of a hall in Hadrian's Villa at Tivoli (125-135) has
been already pointed out. 2 It belongs to a type of vaulted construction absolutely
unconnected with a Hellenic-Oriental origin, and essentially Roman. As a matter of
fact the East, so far as is known, does not contain a single example of this kind
of building earlier than the age of Hadrian. Whereas instances of these three-lobed
structures, sometimes provided with external buttresses, can be found in the works of
Montano, 3 Bramantino, 4 Serlio, 6 and among the drawings in the Uffizi at Florence.
In the same way, the East was not the birthplace of the circular buildings with
annular vaulted aisles, which we discussed in connection with the Holy Sepulchre.

It is usually thought that Einhard designed and carried out the most important
of Charles the Great's buildings, from the palaces of Ingelheim and Aachen to the
wooden bridge at Mainz, reaching their culmination in the rotunda of Aachen. This
idea has been universally accepted, based as it was on the authority of Mabillon
(1632-1707) 6 ; and its truth was investigated only by a few, among whom were Pertz, 7
Springer, 8 Dohme, 9 and Delisle, 10 for almost every writer on Carolingian art has simply
taken it for granted without verification. As it seems to require correction I will
investigate it in my turn, though well aware how difficult it is to get new ideas accepted
when the attempt involves the displacement of old ones.

About Einhard we know that he was brought up in the palace school, that he
held the offices of royal steward or treasurer and of Crown notary, and that Charles
the Great sent him (806) on a mission to Leo III (795-816) in order to obtain the
Pope's assent to the act of partition of his dominions among his sons. 11 There is no
documentary evidence to prove that he was also an architect. It is one thing to be
Minister or Treasurer of the Household (" qui regalium aedificiorum praefectus erat" 12
"operum regalium exactor constitutus " 13 ), and quite another matter to be the
architect of the royal buildings. It requires a strong effort of the imagination to
interpret the words of the epitaph composed by Hrabanus Maurus, 14

Quern Carolus princeps propria nutrivit in aiila,

per quern et confecit mtilta satis opera,
and

ac mult is artefuit utilis

1 Der Dom sit Aachen and seine Entstellung. 2 Dehio and Von Bezold, op. fit.

* Op. fit. 4 Op. fit. 5 O p_ cit _ e Annales Ord. S. Benedict!.

7 Einhardus, Vita Karoli imperatoris. 8 De artificibus monafhis et laicis medii aevt.

9 A'unst and Kiinstler Deutschlands und der Niederlande bis gegen die Afitte des Achtzehnlen Jahrhundtrts.
10 De Eginhardo Caroli Magni notario. " " Mabillon, Annales Ord. S. Benedicti.

13 Man. Germ. hist. Einhardus, Vita Karoli imperatoris.

14 Migne, Pair. Lat., Vol. 112Epitaphium Einhardi.

GERMANY 277

as referring to any duties of Einhard as architect and superintendent of the Imperial
buildings. Nor can such duties be any better inferred from the passage in the
Fontanelle (Saint Wandrille) Chronicle : " Heinhardo abbate viro undecunque
doctissimo," l or from Alcuin's well-known letter to Charles, 2 or Einhard's to his own
son Vussinus. 3 The notice from Fulda of the sending to Einhard by Abbot Ratger of
Brun Candidus " variarum artium doctorem peritissimum," 4 tells us no more, for the
latter was a painter and a man of letters, but not an architect. Again, even if we
make the language of Walahfrid Strabus's flowery eulogy on Einhard

Beseleel fabre priinum qui percipit omne
artificum praecautus opus 5

mean that he superintended the workmen engaged on the Imperial buildings, there is
nothing about his having designed them, and in particular the famous rotunda.

Nor is it any good to say, as Dohme does, that, as the plans and the construction
of the chapel at Aachen demanded exceptional mathematical knowledge on the part
of the architect, Einhard must have been the architect because Alcuin tells us that he
possessed such knowledge. In the West, during the Dark Ages, vaulted buildings
were not designed or erected on the basis of calculations, but on a ground-work of
experience, by means of community of efforts, with the help of traditions of construc-
tion, and of the study of buildings surviving from the ancient world. Such are the
conclusions at which I have arrived, and I have only been confirmed in them during
my laborious researches into the subject of the experiments made for the gradual
evolution of the Lombardic vaulted basilica by the most important of the mediaeval
gilds, I mean the Comacine or Lombard corporations.

Now what traditions and what experience in the art of building did Einhard
possess, when, at the age of twenty-five (Dohme and Springer date his birth
approximately in 770 ; Pertz at the end of Pippin's reign [752-768] or the
beginning of that of Charles the Great [768-814]), he took in hand the design,
and in 796 the execution, of the most celebrated edifice of that age either
in East or West ? Those tasks demanded not only a study of its original,
San Vitale (a filiation noticed long ag6 by Hiibsch 6 ), by one who was familiar
with the problem, but also profound technical and statical knowledge which
is not acquired off-hand. My answer is that he had none. It is true that Adhemar
tells us that after the conquest of Lombardy (774), Charles brought from Italy
singers and organists, as well as accomplished teachers of grammar and arithmetic
or calculation, of whom there was a deficiency in his own country : " Ante ipsum
enim dominum regem Karolum in Gallia nullum studium fuit liberalium artium." '
In this way Einhard, after he had grown up, had an opportunity of devoting
himself to the study of these subjects, and we have testimony that his application
was not without result. But there is a great difference between that and producing
the design for the Imperial rotunda, or even having a predominant share in the
preparation of the plans and the conduct of the works, especially when it comes
to vaulting. His literary productions do not seem to suggest the powerful brain
which gave birth to the Palatine Chapel.

1 D'Achery, op. cit.Chronicon Fontanellense. * Migne, Pair. Lot., VoL lOOEfistouu.

3 Duchesne, Historiac Francorum Scriptorcs Eginhardi abbatis cpistolae.

4 Men. Germ. hist. Catalogus abbatum Fuldcnsium.

* Migne, Pair. Lot., Vol. 114 Carmina De Einharto magna Eginhardo.

Op. fit. i .Van. Germ. hist.Ademarus, fft'sleritu.

278 LOMBARDIC ARCHITECTURE

Moreover, had he been the architect and master of the works, it would
be difficult to explain the silence of the chronicles about the fact (while the names
of several contemporary architects are preserved), and also his own ; whereas
he does not fail to mention the church which he built at Seligenstadt, and insist
on its importance: " non indecori operis." 1 It would be incomprehensible that,
when he laid aside his courtly robes as Minister of the Imperial Household and
Crown Notary in order to assume the humble garb of a presbyter and abbot,
he should have forgotten all the science, unequalled at the time, displayed in
the great rotunda, and have exhibited so limited and mean a substitute for it in the
churches which we know he founded. This consideration has peculiar force in the
case of the one at Seligenstadt, erected to receive the precious relics of SS. Peter and
Marcellinus, so coveted and venerated by the founder, and also to form the resting
place of his own mortal remains. For even without raising an edifice too grand for
the pecuniary resources of himself and his wife Emma and they were not
contemptible he might well have built one proportioned to his means, and yet of a
character to form a monument of the remarkable and precocious architectural attain-
ments which have been ascribed to him.

We conclude, then, that Einhard's claim to be the architect of the chapel at
Aachen cannot survive the test of sound criticism. His name is never connected
with it either as designer or executor. The Monk of St. Gall, who has left us the
least incomplete account of the rotunda, makes no allusion to him in this connection. 2
All that we can say is that, being young and without experience in the difficult art of
vault construction, he may have had the opportunity of initiation into its secrets
owing to the fact that the Emperor had summoned to Aachen for the purpose masons
from Italy and France. The craftsmen of Piedmont and Lombardy were the best
known and most skilful in the Empire. Two centuries before, their services had been
engaged by Nicetius, bishop of Trier, as we learn from the letter of bishop Ruffus of
Turin, who must have been referring to workmen of his own or neighbouring dioceses.
And my belief is that it was to them and to workmen from other parts of Italy and
from Transalpine Gaul, that the monk alluded in the words : " ad cuius fabricam de
omnibus cismarinis regionibus magistros et opifices omnium id genus artium
advocavit." 3 Springer, too, thought that these master masons came from Italy and
Gaul.

Having said so much let us turn to the churches of Steinbach and Seligenstadt,
which are very instructive from the evidence which they afford as to the type of
building adopted by Einhard in his own foundations, and to his capacity as an
architect.

THE CHURCH OF STEINBACH NEAR MICHELSTADT is believed to have been
built by Einhard (f 844) some time after Louis the Pious (814-840) had made him
and his wife Emma a grant of Michelstadt (815) in the Odenwald, where a small
wooden church existed, and before 819, the year in which the husband and wife
transferred the chapel at Michelstadt to the abbey of Lorsch. 4 The dedication took
place in 82 1. 5 In it were deposited the relics of SS. Peter and Marcellinus, stolen
from Rome (826-827), and later enshrined in the apse of the church at Seligenstadt. 6

1 Man. Germ. hist. Translatio et miractila sanctorum Marcellini et Petri.

2 3 Man. Germ. hist.Notkerus Balbulus, De geslis Karoli Magni imperatoris.
* Man. Germ. hist. Chronicon Laureshamense.

5 Man. Germ. hist. Annales Fuldenses antiqui.

6 Man. Germ, hist. Einhardus, Trans/ntio et miracula sanctorum Marcellini et Pet.ri.

GERMANY

279

Of the original structure, of T cross plan with nave and aisles terminated by
apses, there remain the nave (Figs. 724, 725) and the north arm of the transept with
their respective apses. In its present condition the nave measures about 79 ft. x 24 ft.
The arches, barely 4^ ft. wide, and now walled up, formerly opened into the aisles.
They rest on quadrangular piers. The walls carried by the arches contained
originally a corresponding number of narrow round-headed windows, splayed on the
inside. The main apse, starting directly from the transept wall and of semicircular
form, is lighted by
three similar win-
dows. The gable
and wall above the
frontal arch of the
apse is pierced by
three round windows,
two of which were
intended to light the
transept, and the
third to give light
and air to the roof.
It seems that the use
of round windows,
derived as we sug-
gested in our account
of Norwich from a
Roman source, was
widely spread at this
time in Germany, for
we find them even
represented in illu-
minated MSS. I
may refer to the pic-
tures of two aisled
churches which I
have noticed in the
St. Gall " Psalterium
aureum " (IXth cen-
tury), 1 where they ap-
pear in the nave and
aisles.

So far as one

may judge from what is left, the arms of the transept were entered through two
arches, barely 4$ ft. wide, with moulded imposts. The way in which the north arm
of the transept is shut off suggests that it was used as a chapel. The apse which
opens out of it has lost its original appearance.

With the exception of the apses, the whole building is roofed with timber.
Underneath the choir, crossing, and part of the nave, extends a crypt, the whole of
which is underground and consists of barrel vaulted passages.

So far as may be inferred from what mutilations, alterations, and extensions in

1 Library of the former Abbey of St. Gall.

Fig. 724. Steinbach near Michelstadt. South side ol church (815-819).

LOMBARDIC ARCHITECTURE

the past have spared, and also from Einhard's allusions, of the same type though of
larger dimensions was the church of SS. Peter and Marcellinus at Seligenstadt ($_*-
previously known as Mulinheim and already containing a small stone church, which
was given by Louis the Pious to Einhard and Emma in 815. Here he erected his
new church dedicated to the martyrs, and in it he finally enshrined their relics
previously deposited at Steinbach and in St. Martin's at Ostheim. 1 2

The church of Seligenstadt consisted of a nave and aisles, the former bein"

about 33 ft. wide,
and the latter only
half as much, sepa-
rated by nine quad-
rangular piers on
either side, which
an excavation in the
modern facing has
shown to measure
some 28 in. x 32 in.
and to be constructed
of bricks taken from
Roman buildings.
They have moulded
imposts.

It is easy to see
that the plan of
both of Einhard's
churches, with its
T cross form, was
derived from that of
the Vatican Basilica.
But they are rough
work, almost devoid
of architectural deco-
ration, and roofed,
with the exception
of the terminal re-
cess, with wood.
These facts are diffi-
cult to reconcile with
the idea that the man

who designed them was the creator and constructor of the principal buildings erected
by Charles the Great.

Kig. 725. Steinbach near Michelstadt. Church

After Charles the Great's conquest of Lombardy had brought Italy into direct
relations with his northern dominions, architecture made a brilliant appearance in
the German lands with the rotunda of Aachen ; but this appearance was as ephemeral
as the Empire which its founder was unable to endow with permanent vitality. It

' Man. Germ. hist. Chronicon Laurcstawtfasf.

9 A/on. Germ. kist.F.iHhardus, Trait slat in ct miracula sattitorttm Marrtllini tl /Wr.

GERMAN'Y 281

was ephemeral, too, because the architectural awakening brought about by the
Emperor in the lands beyond the Alps was the result of his personal influence, and not
the effect of a long period of preparation and, at the same time, the expression of the
spirit and the needs of the age. In fact, all the buildings of his reign and vast
Empire which are of importance for their vaulted construction were due to his
personal will, and intended to promote his own glory and self-satisfaction. Thus, in
addition to the great rotunda, another royal chapel of similar form was attached to
his palace at Casseneuil l which was destroyed by the Normans in 879. Other
buildings, too, if not erected by his orders, were aided by his contributions, such as
Theodulfs church at Germigny des Pres (801-806).

Moreover, Art is dependent on public prosperity, and this was certainly
not assured by the successors of Charles, whose incapacity is the theme of the
historical records of two centuries. The first was Louis the Pious (814-840), born to
wear the tonsure rather than the crown. In the course of a few years he fatally
undermined the inheritance of the Pippins, Charles Martel, and Charles the Great,
more particularly by the weakness of his conduct towards his wife Judith, and his
youngest son Charles the Bald. His reign was disturbed by domestic and civil strife,
fomented not so much by the indifference of his subjects, who were disgusted by his
weakness, as by the discord between the Latin and the German element ; in addition
to which there came the incursions of Danes or Normans and Saracens. Under his
successors, Lothair (840-855), Louis the German (843-876), Charles the Bald
(843-877), and Pippin the Younger (838-846), the anarchy inherited from their father
and grandfather respectively was intensified by the internal struggles to which we
have referred, and by the abandonment of any attempt to resist the Northern
barbarians and the Saracens ; and at length the partition of Verdun (843) dissolved
the fabric of the Empire.

This partition was succeeded by years of fruitless effort to diminish, if it was not
possible to prevent, the raids within the divided realm of the barbarian hordes which
left a trail of blood and ruin wherever they passed. The war against invaders was
supplemented by the one between Louis II and Charles the Bald. Finally, the
death of Lothair (855) broke the last formal tie which still united the Empire of
Charles the Great. It was followed by new family and civil wars, with the usual
accompaniments of incursions and rebellions, until with Charles the Fat (881-887) the
legitimate branch of the Imperial race became extinct in Germany.

Thereupon the eastern Franks elected as king the brave Arnulf of Carinthia
(887-899), bastard son of Carloman, king of Bavaria (865-880). He succeeded in
breaking the insolence of the Normans at Louvain (891;, in suppressing the revolt of
his vassals, and, finally, in assuming the Imperial diadem at Rome (896). His son,
Louis the Child, was elected as his successor (899-911), but his weak rule was
troubled by civil wars and barbarian raids. With his death even the illegitimate
German line of the descendants of Charles the Great came to an end.

Some idea of the conditions of ecclesiastical architecture in the German lands
during the age of the heirs of Charles the Great is afforded by three churches, of
which two are still in existence while we possess the plans of the third. It may be,
too, that there is a fourth, of which a drawing exists. Let us see what they were like.

THE ROUND CHURCH OF ST. MICHAEL AT FULDA was built by Eigil, fourth
abbot of Fulda, between the years 818 and 822,* as we are told in his Life, written by
1 Cordcro, of. cil. * Man. Germ. hhl. /tnnaki f'uldentei antiqui.

282

LOMBARDIC ARCHITECTURE

Fig. 726. Fulda. Church of St. Michael (818-822).

ing structure. It was a cemetery church, of circular
supported by a circle of eight columns, covered with a
a crypt beneath with its vaulting supported by a
building (Figs. 726,
727) contains eight
arches on the ground-
floor with columns
surmounted by four
capitals of Roman
origin, three of which
are Corinthian and
one Composite, and
by four plain cubical
funnel-shaped capi-
tals with deep abaci.
Anapsidal sanctuary
projects at the east.

The building
was altered in the
Xlth century 4 by
removing the vault-
ing and raising the

1 Mon. Germ. hist.
Vita Eigilis abbatis Ful-

densis.

2 Mon. Germ. hist.
De vita Aegili versibus ex-
plica/a.

3 Browerus, Fuldenses
antiijuitates.

4 LUbke, Geschichte
der deutschen A'ltust.

the monk Brun Can-
clidus ; 1 - and it is
known that it was
not finished in 819,
for Haistulf, arch-
bishop of Mainz
(813-826), did not
dedicate it till 822. 3
Anyone who reads
this Life (in verse
as well as prose) will
find a clear and defi-
nite statement about
the primary inten-
" tion of the building,
and also a descrip-
IHf tion of its original
form, corresponding
to that of the exist-
plan, with an annular aisle,
dome of masonry, and having
central column. The actual

Fig. 728. Trier. Porta Nigra (IVth Century).

GERMANY

283

height of the church as we now see it. A triforium gallery was constructed with
openings, each of which was divided in two by a shaft carrying a cubical funnel-
shaped capital and a corbel pulvin, hollow chamfered (recalling the prototypes
of this kind at Mettlach [987]) and curled over at the ends. Eight windows
were also inserted, and the whole was roofed with wood. At the west a nave was
added, approached through a tower porch, the two-light openings of which have
Lombardic cubico-spherical capitals with pulvins like those just described. This nave,
and two other addi-
tions on the north
and south, have given
the rotunda a cruci-
form appearance.

Beneath is a
crypt. The middle
part has a roughly
constructed concen-
tric barrel vault
springing from a cen-
tral column (a frag-
ment of ancient
origin), provided with
a rude Ionic capital,
and an inverted
funnel-shaped capital
for base. The annu-
lar aisle also has a
barrel vault, which
has been cut by the
insertion of cross
walls. Originally it
was lighted by very
narrow windows.

It is obvious that
Eigil, the architect
of the sepulchral
rotunda of the abbey
of Fulda (" Eigil . . .
aliam aecclesiam in
cymiterio rotundam

mira arte typice composuit " : ), derived his idea from the round tombs of ancient Rome.
The annular rotunda, with its dome and barrel vault, reproduces on a smaller scale the
mausoleum of Santa Costanza (IVth century). The circular vault of the central part
of the crypt is modelled on the crypt of the mausoleum in the Villa of the Gordians on
the Via Praenestina (Illrd century), for there can be no doubt that the round building
popularly known as " Tor de' Schiavi " was the tomb of the Gordian family. Below it is
a beautifully constructed crypt, turning round a central mass which serves as a support
for the whole structure. This tomb provided the suggestion for the architect of the
mausoleum of Romulus (f 309), the son of Maxentius, standing in the midst of a

1 Man. Germ. hist. Catalogtis abbatum Fuldtnsium.

Fig. 727. Fulda. Church of St. Michael (818 822).

284

LOMBARDIC ARCHITECTURE

spacious arcaded court close to the Circus of Maxentius by the old Via Appia, though
he gave the central block a more developed form by taking out of it eight semicircular
recesses corresponding to those in the outer wall. In all these tombs the vaulted
crypt is derived from the typical form found in the " tholos " tombs of Volterra.
Thus the Inghirami Tomb (now in the Archaeological Museum at Florence), thought
to belong to the Ilnd orlllrd century B.C., has a central pier supporting an elementary

' annular vault.

In the rotunda
at Fulda the Roman
cubical funnel-
shaped capitals
should be noticed.
They are derived
directly from those
of the Porta Nigra
at Trier (Fig. 728)
belonging to the
second half of the
IVth century, or
more precisely to
the reign of Valen-
tinian I (364-375),
under whom the city
was much embel-
lished l and put in a
better state of de-
fence against the
Germans. The capi-
tals of the gate at
Trier are the oldest
specimens of the
kind that we have
seen. It was on this
form of capital, to-
gether with the
Ravennate pulvin,
that the Byzantines

afterwards modelled their cubical funnel capitals of quadrangular shape with swelling
sides, which in their simplest form may be seen in the cistern of Binbir-direk at
Constantinople (Vlth century).

In connection with the Gate of Trier we may observe in passing that the Gallo-
Roman peoples gave exceptional importance to their city gates. In addition to this
one, we may mention as proof the Porta Palatina of Turin (Fig. 729), erected under
Augustus (29 B.C.-I4 A.D.), and the Porta dei Borsari at Verona (Illrd century).

Though Eigil's church, like that at Aachen, betrays its Italian origin, the type of
its original capitals indicates the work of northern craftsmen. In Italy at the time
this form was rarely used, the preference being given to the Pre-Lombardic cubical
pattern. The vaulted roof is to be explained by the fact that the church was erected

1 Browerus, Antiyiiilates et annales Trevirenses.

Fig. 729. Turin. Porta Talalina (29 B.C.-I4 A.D.

GERMANY

285

very shortly after the completion of the great works at Aachen, and under their
influence ; and, indeed, some of the builders there employed may have been engaged
on it.

THE ABBEY CHURCH OF ST. GALL was rebuilt by Abbot Gotzpertus (816-837)
from the designs of the two monks Winiharius and Isenricus. It took seven years to
finish (822-829).! 23

In studying the original of the very important plan of the abbey drawn on parch-
ment, which is dated about 82O, 4 we found that the church presents three notable
features, viz. the apses facing one another at the east and west ends, the semicircular
aisle round the western apse, and the towers which flank the latter.

The first of these peculiarities we discussed in our account of Abingdon Abbey
(675) ; and we dealt with the subject of ambulatories, with or without arcades, in con-
nection with the cathedral of Ivrea and Saint Benigne at Dijon. Western towers we
know were designed to contain the staircases belonging to the facade, after the fashion
of Ravenna. Their function here is made clear by the legend on the plan, " Ascensus
per cocleam ad universa super inspicienda."

THE CATHEDRAL OF COLOGNE. We possess a written account of the general
form of the church erected by Archbishop Hildebold (785-819), and restored or com-
pleted by Archbishop Willibert (870-889), who dedicated it in 873. It was of basilica
plan, with a choir apse at either end and crypts underneath them. Two wooden
bell- towers flanked the western apse,
each containing an altar. It was
lighted by round windows in addition
to others of rectangular form, some
larger (of which three were in the
eastern gable) and some smaller.

We are not told who restored or
rebuilt Willibert's church after its
injury in the terrible fire from which
Cologne suffered at the hands of the
Normans in 882. We only know, on
the authority of Gelenius, 5 that in
1080, when Sigewin was archbishop
(1079-1089), the easternmost part of
the cathedral was suddenly burned.

A certain amount of light is
thrown on the subject by the im-
portant Xlth century Evangelistarium
executed by the brothers Burchard
and Conrad " ad altare Sancti Petri

1 Mabillon, Annales Ord. S. Benedicti.

3 Man. Germ. hist.Ratpertus, Casus S.
Galli.

3 4 Keller, Bauriss des Klosters St. Gallen
vomjahr 820.

DC admiranda, sacra ,t civili magniludint Fig 73o ._ Co logne. Cathedral Treasury. Illuminated
Colomae Claudtac Agrippintnsis Augustat, Ubi- leaf of Xlth Century Evangelistarium with representa-

orum urbis. tion of the old Cathedral.

286

LOMBARDIC ARCHITECTURE

infra muros Coloniae" for Illinus, canon of the cathedral, 1 in which may be seen
(fol. 16 v.) a picture of the donor presenting the book to St. Peter seated in a chair,
and above them a representation of the metropolitan church (Fig. 730). Granted,
always, that it is certain that the volume is earlier than 1080. For in that case we
have an illustration (if only approximately accurate) of the cathedral of Cologne
in its restored or new form after 882, showing the transepts belonging to the choifs
at either end of the building. The two towers rising at the east end evidently
bell-towers, as the openings in the highest stage show must be earlier than the bell-
towers of the cathedral at Ivrea (973-1001 or 1002), and are therefore the prototype

of this arrangement.
In the present case
it is very probable
that it was suggested
by the staircases
formed in the outer
angles at the end of
some building of the
Roman period. Such,
for instance, is the
Basilica at Trier
(Fig. 730. thought
to belong to the age
of Constantine (311-
337), but which, con-
sidering its grand
dimensions, and the
character of its brick
facing, together with
the enclosing arches
round the windows,

which are original, may well be dated in the time of those great builders, Diocletian
(284-305) and Maximian (286-310).

I may mention here that a three-lobed building, the plan of which by Fra
Giocondo (?) is preserved among the drawings in the Uffizi, shows two staircases
flanking one of its apses.

THE SEPULCHRAL CHAPEL OF LORSCH is of rectangular shape, and its system
and western sides (Fig. 732) are decorated with a range of arches and blank triangular-
headed arcading. The walls are constructed with polychrome polygonal stone
checkers in imitation of Roman polychrome " opus reticulatum " such as may be seen
in the amphitheatre at Assisi. It always had, as now, a wooden roof of very high pitch.

The interior (Fig. 733) contains a sarcophagus found in the old cloister of the
neighbouring abbey of St. Nazarius. It is ornamented with pilasters and Ionic
capitals exactly like those of the triangular-headed arcading on the chapel itself, and
probably formed the coffin of Emperor Louis III the Saxon.

Many writers, Adamy 2 among them, believe that this structure (now known as

Fig- 731. Trier. Basilica (Illrd or IVth Century).

1891.

1 Cologne. Treasury of the Cathedral.

- Die frdnkische Thorhalle und Klosttrkirche zu Lorsch Historischer Vereinfiirdas Grossherzogthum Hessen,

GERMANY

287

Fig- 732. I-orsch. Sepulchral chapel. West side (876-882).

the Michaclskapelle)

was the old vestibule

of the-atrium of the

abbey church of St.

Nazarius, founded by

King Pippin in 764^

and later rebuilt on

a larger scale by the

monk Adalbert

(1144-1151). I think

there is no doubt

that it really is the

burial chapel erected

by Louis III the

Saxon (876-882)

near the abbey of

Lorsch, as has been

stated by others. 23
The chronicle of

Lorsch says that

Louis 1 1 1 buried his

father Louis II the

German (843-876),

the founder of the

national dynasty, in

the abbey of Lorsch (" patrem in Laureshamensi monasterio tumulavit "), and that

afterwards he himself was buried near his father in a church which he had built,

known in the days of the chronicler as " the variegated church " : " Ludowico

rege Germanico, filio Ludowici, defuncto et iuxta patrem apud Lauresham in

Ecclesia quae dicitur Varia, quam ipse huius rei gratia construxerat, sepulto."

Later, in 1053, this chapel was dedicated to the Virgin, the Apostles, and All Saints. 4

Now the exist-
ing Michaelskapelle
is, as a matter of
fact, " varia " owing
to its polychrome
facing. That it can-
not be the original
porch leading to the
atrium in front of St.
Nazarius (according
to Adamy's imagin-

1 Man. Germ. hist.
Chi-onicon Lattrcshamensc.
Savelsberg, Deulsches
Kunstblatt, hcrausgeg. von
Eggers.

1 Forster, Denkmale
deittscher Baukumt.

4 Mon. Germ. hist.
F 'g- 733- Lorsch. Sepulchral chapel (876-882). Chronicon Laurcshamensc.

288 LOMBARDIC ARCHITECTURE

ary design) is proved by the fact that the terrible fire of 1090 destroyed the whole
church, which had a wooden roof, 1 " and would certainly not have spared the atrium
and its vestibule. It is also proved, and still more convincingly, by the absence
of any trace of the junction between the existing polychrome structure and the
spacious cloister court which is supposed to have existed.

Moreover, in the Prankish realm capitals were of quite a different type and
execution in the time of Pippin, under whom the building of St. Nazarius was begun,
and this we learn from the crypt of the church at Flavigny (755-768). Again, edifices
of the age of Charles the Great, when the works at Lorsch were finished, had no
external architectwral decoration, as we know from the rotunda of Aachen and the
churches of Germigny des Pres (801-806) and Steinbach (815-819). And that this was
still the casein the days of Louis the Pious the round church at Fulda (818-822) bears
witness.

The sepulchral chapel of Lorsch is to be regarded as the earliest instance of
a building decorated with ranges of blank triangular-headed arcading, a design of
German origin. For although at an earlier date the baptistery at Poitiers (Vllth
century) had exhibited the decorative use of pediments and blank arcading alter-
nately triangular and round-headed, the chapel at Lorsch is the first dated building
that displayed this particular form of treatment. This characteristic feature maybe
traced back to the Ravennate and Pre-Lombardic blank arcading with round arches ;
while the substitution of triangular for round heads may have been suggested by some
building such as the baptistery of Poitiers. It may even have been derived from the
pedimented colonnading on sarcophagi of the Early Christian period, of which
examples are to be found in the Lateran Museum. Or, again, it may have its source
in some building of the Roman age. It was afterwards copied by the architect of
the abbey church of Gernrode (968). And it was Lorsch and Gernrode which furnished
the pattern to those who carried triangular-headed arcading to England.

I may mention here that in the Xth century MS. of Boethius " De institutione
arithmetica " at Bamberg 3 I have noticed the front of a building with triangular-
headed arcading on its upper part, the heads forming part of a lozenge-shaped
decoration ; while the " Evangelistarium of Essen," believed to be of the Vlllth or
IXth century, and earlier than 834,* shows triangular-headed arcading formed of
interlacing bands and scrolls.

I think that the capitals at Lorsch are the work of French chisels. We shall see
presently how different were the knowledge and handiwork of the German artists.

# * *

On the death of Louis the Child (899-911), Conrad I of Franconia was
raised to the throne, but his reign (911-918) was disturbed by perpetual civil wars
and barbarian invasions. The elevation, however, of the illustrious Henry I
the Fowler (918-936), the victor of Merseburg (933), saw the restoration of order
and security in Germany.

THE CRYPT OF THE CHURCH OF ST. WIPERTUS NEAR QUEDLINBURG. The
church of St. Wipertus was erected by Henry I and his consort Matilda (f968).
A passage in her life fixes the date as 936. 5 e 7 8

1 Man. Germ. hist.ChronitonLaureshamense. - Helwich, Antiquitates Laurishaimenses.

3 Bamberg, Royal Library. Humann, Die Kunstwerke der Miinsterkirche zu Essen.

5 Man. Germ. hist.Annales Quedlinburgenses. 8 Knackfuss, Deutsche Kunstgeschichte.

7 Dohme, op. cit. e Mon _ Germ hist. Vila Mahthildis reginae.

GERMANY

The crypt (Fig. 734) is all that is left of the original building,
of a small basilica ending in a semicircular apse. Every part
barrel vaulting, and
it is surrbunded by
an ambulatory. The
pillars have roughly
made capitals formed
by an inverted ovolo,
a hollow moulding,
and a roll, with a
rude abacus. The
bases have two bul-
bous rolls separated
by a hollow moulding.

Thiscrypt.above
which must have
stood the apse and
presbytery of the
original church, is im-
portant on account of
its vaulted ambula-

289

It has the form
covered with

734- Quedlinburg. Crypt of St. Wipertus (936).*

tory, peihaps suggested by the one in Constantine's Lateran Basilica as enlarged by

Pope Sergius II (844-845), which I carefully examined before its recent destruction.

We have here, in fact, the oldest example of a crypt of this form to be found either

in Italy or beyond the Alps.

The crypt of St. Wipertus recalls the ancient basilica of the SS. Annunziata

at Praia, near Avellino. Of the primitive church there survives the interesting

and, on account of its archi-
tectural form, important apse,
pierced by round-headed
windows instead of arches (Fig.
735). In it is recessed a niche
for the bishop's seat, a Pagan
and Roman idea, for in the
back of an exedra in the
" Palace" of Hadrian's villa at

MB 41 B Tivoli ' a semicircular niche

r ^ff ,^g B occurs. In the IVth century it

occurs.

was introduced in the basilica
of St. Petronilla between the
Via Ardeatina and the Via
Ostiensis,'- and also in that of

Fig. 735. Prata near Avellino. Apse of the Church of the

SS. Annunziala ^Ilth Century).
VOL. II

* The illustrations belonging to
Quedlinburg and Gernrode are from
photographs by Herr K. Kliche.

1 Keina e liarbieri, Kilicvo plant-
melrico e altime'rico di villa Adriana.

Bull, di arch, cristiana, 1874 De
Rossi, t'iaiila delta basilica di Paula
Petronilla nel cimitero di Doniitilla,

290

LOMBARDIC ARCHITECTURE

the Martyrs Simplicius, Faustinas, and Viatrix (382), on the Via Portuensis near
Rome. 1

The spiral terracotta shafts (made for their places) which support some of the
arched openings in the apse at Prata, have capitals carved with very rude, stiff,
plain leaves, and cauliculi like ram's horns, slightly curved at the top. These capitals
enable us to fix, more precisely than has hitherto been done, the unknown date
of the apse of the Annunziata, which some 2 regard as belonging to the earliest
Christian age, while others 3 put it between the Vllth and Xth century. By
a process of elimination its date will be, approximately, that which followed
the Lombard Conquest and the scourge of pestilence and famine which afflicted
Italy about 566, the period which saw the artistic awakening initiated by
Theodelinda (590-625) ; in other words, the first half of the Vllth century.
Before the descent of Alboin (568) and after the reign of Rotharis (636-652),
Italy never saw such degraded work as these capitals (especially the round one),
even though produced in remote places and by local carvers.

The date which we have suggested explains the arcaded form of the apse,
a plan which was in favour from the end of the IVth century to about the
second half of the Vlth both at Rome and Naples ; so much so that in the latter
century Bishop Vincentius (554-577) was still employing it in San Giovanni
Maggiore at Naples (" Hie fecit praefulgidam basilicam. . . . Quern amplis aedificiis
in gyro distinxit " 4 ), while at Rome Pope Felix IV (526-530) adopted it for
SS. Cosma e Damiano.

Fig. 736. Quedlinbur. Old Crypt of St. Servatius (936).

THE CRYPT AND CHURCH OF ST. SERVATIUS IN THE CASTLE AT
QUEDLIKBURG. The erection of the castle church of Quedlinburg was begun

1 De Rossi, La Rcm.i so'.terranea cristianc La piccola basilica damasiana dedica'.a a Simplicio, Faustina,
Viatrice, martiri storici del cimitero di Generosa.

' 2 Archivio storico jer le provintie napoleta:ie, 1878 Taglialatela, D^'^ cn'.i^a basilica e della catacomba
di Prata in Princ. Ulte.". e di alcuni monumenti avellinesi.

3 t'ertaux, op, fit.

1 Man. Germ. hist. Scriptures rerum langobardicarntn Gesta ejiscojorum mafclitaiiorum.

GERMANY

291

by Matilda at the wish of her husband Henry I, shortly before his death in 936, as we
are told in the Life of the Empress. In that year the crypt must have been finished,
for it received the tomb of the great Emperor. In 997 the church was rebuilt by
the second Matilda (f 999), daughter of Otto I the Great (936-973) ; but it was not
finished till 1021, as in that year a second dedica-
tion took place. 123 In 1070 the town was burned
("Quindelincburg exusta est"*), and with it the
church, which was rebuilt and reconsecrated in 1 I29. 5

The crypt of the original building, which was
discovered in the last century below the floor of the
apse in the present crypt, belonging to the recon-
struction after the fire of 1070, is surrounded by a
range of recesses separated by engaged shafts which
carry a continuous architrave ; the whole being com-
posed of stucco, and of rude workmanship (Figs.
736, 737). That this is the crypt of the primitive
church is proved by the existence at its west end
of the tombs of Henry I, of his wife Matilda
(" sepultaquc est coram altari Christ! presulis Ser-
vacii iuxta seniorem suum " 6 ), and of their grand-
daughter, the Abbess Matilda, who in 999 was in-
terred " iuxta tumulos regum, avi et aviae suae
Heinrici et Mechtildis." 7

Of a later date than this crypt, and probably
forming part of the works carried out between 997
and 1 02 1, is the underground apsidal chapel beneath
the south aisle of the church. One of the side walls
contains three arches with two shafts bearing quasi-
Composite capitals, above which are corbel pulvins
also carved with foliage, like some of those at Mcttlach
(997). The bases are of bulbous form, and rest on
tall moulded plinths. The carving recalls that on
the capitals and pulvins at Mettlach, though it is
not so advanced.

The constructional supports in the crypt of St. Wipertus and the decorative
ones in the crypt of St. Servatius found an echo in England. As a matter
of fact we have already seen how, not earlier than the reign of Edgar (959-975),
spiral shafts and inverted truncated pyramid capitals made their appearance
in the crypt and apse of Repton ; while several buildings contain bases with
disproportionate, clumsy rolls, or of the bulbous form, and show the influence
of the outlandish and barbarous mouldings of the supports in the church at
Quedlinburg. This influence must be connected with the monastic intercourse
which from the days of St. Gall (f about 630) and St. Boniface (f 755) had been
going on between England and Germany with the other Teutonic lands, and was
only intensified by the marriage of the pious Edith (929-946), daughter of Edward
the Elder (901-925), with Otto the Great as his first wife.

Fig. 737. <\>uedlinburg. Shaft in old
Crypt of St. Servatius (936).

1 Man. Germ. hist. Annalts Quedlinburgoises.

3 Mon. Germ, hist. I'ita Mahthildis reginae.

5 Knackfuss, op. cit.

1 Mon. Germ. hist. Annalts Quedlinburgenses.

Afon. Germ. hist. Widukindus, Res gestae Saxonicae.

* Alon. Germ. hist. Annales Corheiensts.

* Alon. Germ. hist. Thielmaru;, Chronicon.

U 2

292

LOMBARDIC ARCHITECTURE

In connection with the two important monuments which we have examined at
Quedlinburg I think it opportune to mention that bulbous bases, sometimes resting
on a plinth formed like a church, are represented in abundance in German MSS., e.g.
the " Isidori Etymologiae " (Xth century), the " Psalterium Folchardi " (IXth century),
and the " Vita S. Columbae " (IXth century), in the Library of St. Gall.

In the next place, these structures are evidently the work of Saxon hands,
no doubt the best that could be procured, considering the importance of the place
where they are found, the rank of the founders, and the royal use for which one
of them was intended. They give an idea of the state of building, carving, and stucco

Fig. 738. Quedlinburg. Crypt of St. Servatius (1070-1129).

work in Germany at a time when the local craftsmen had been forced to rely on their
own attainments without help from outside. And this confirms the opinion we
expressed about the carvers of the capitals in the chapel at Lorsch.

Let us now turn to the existing crypt (Fig. 738) and the cathedral church of
Quedlinburg which rises above it. The crypt under the choir (rebuilt in the Pointed
style) and transept of the present church consists of a central space, divided into
nave and aisles by pillars and terminating in an apse, and two lateral arms with small
apses at their extremities. The unraised cross vaulting springs from capitals
ornamented with stiff, plain leaves, palmetto leaves and other kinds of foliage,
cauliculi, crosses, interlacing bands ending in a sort of Ionic volutes (Fig. 739), pine
cones, demons' heads with serpents coming out of their mouths and biting their ears
(Fig. 740), and eagles. Three are of a curious stepped form.

The church consists of nave and aisles separated by arches with columns, between
every two of which comes a pier. The columns have characteristic bases with two

GERMANY

293

Fig. 739. Quedlinburg. Capital in Crypt
of St. Servatius (1070-1 1 29).

I''g- 74 C - Quedlinburg. Capital in Crypt
of St. Servatius (1070-1129).

rolls and a broad hollow moulding between them, while the capitals arc of cubico-
spherical form, carved, like the deep abaci, with animals (Fig. 741), human figures,

birds, foliage, in-
terlacing, scroll
work, monsters,
animal heads, &c.
From an artistic
point of view,
both in the crypt
and the church
the foliage and
other decorative
elements are fair-
ly well treated
while the repre-
sentations of
living beings are
almost uniformly
of barbarous char-
acter.

With the ex-
ception of the two minor apses in the transept, which have half-domes, the surviving
portions of the church of 1070-1129 arc roofed with timber. At the west end is
the narthex with unraised cross vaulting and visible vaulting arches, above which is a
gallery with pairs of openings, covered by a wooden roof. It is flanked by two
towers, rebuilt like the gallery for the bells which unites them in their upper part.
The artistic features in the crypt are of the same date as those in the nave and
transept, and the continuous cross vaulting in the central part of the crypt is con-
temporary with
that constructed
with wall and
transverse arches
in the lateral por-
tions and in the
narthex. The
view, then, of
those who regard
the crypt as be-
longing to a dif-
ferent date from
the church, falls
to the ground.
The date in ques-
tion belongs to
the years between
the fire of I o/o

and the reconsecration of 1129; for, in spite of what is believed in some quarters
to the contrary, there does not survive one stone upon another of the new
structure of 997 that meets the eye. The quality of the builders and artists of

Fig. 741. Qiiedlinburg.
St. Servatius.

Church ot
Capital (1070-1129).

Fig. 742. Ilsenburg. Capital in the
Church (1087).

294

LOMBARDIC ARCHITECTURE

the Harz district in the second half of the Xth century is revealed by the rough
irregular vaulting in the eastern crypt of St. Cyriacus at Gernrode, and by the rude
capitals and bases of the pillars, as well as by the decoration of the apse at the east
end and of the western towers.

On the other hand, the capitals both in the crypt and the church at Quedlinburg
present forms unknown to the West, and still more to the East, before the epoch
of about 1000. I refer to the Lombardic cubico-spherical capitals which appear for
the first time in Sant' Abondio at Como (1013-1095), and do not show themselves in
Germany till after 1015, in St. Michael's at Hildesheim. Moreover, the Lombardic
figure capitals, no longer showing the merely symbolic figures of Early Christian art,
and going beyond the representations on the capitals of the VHIth to the Xth
century, did not gain much extension before the first half of the Xlth, and only

reached their culmination by the addition of scenes
of writhing and struggling monsters in the second
half of the Xlth and the following century.

I may notice here that the capitals at Qued-
linburg have obvious analogies in style, modelling,
and execution, with those in the church of Ilsen-
burg (Fig. 742), erected in 994 and rebuilt after
the injuries it suffered during the disturbed reign
of Henry IV (1056-1106) by Burchard II, bishop
of Halberstadt (1059-1088), who consecrated it in
loS/. 12 Compared with them the capitals at
Quedlinburg show a more advanced stage of art, a
fact to be explained by the earlier date of the
church of Ilsenburg.

They have similar analogies with the capitals
in the church at Driibeck (Fig. 743) which
when Louis III (876-882) conferred rights of immunity
have no information about it between 1058 and 1130.

The

Fig. 743- Driibeck. Capital in the
Church (Xlth or Xllth Century).

was in existence in 877

on the monastery. 3 We

but it is believed to have been rebuilt in the early years of the Xllth century.

capitals at Quedlinburg are differentiated from these by more artistic arrangement of

the foliage, so that those at Driibeck may very well belong to the end of the Xlth or

the beginning of the Xllth century. 4

Over and above the reasons given for this conclusion there is the fact that
crypts of basilica plan, embracing not only the area of the apse and presbytery, as in
the parish church of San Leo (881-882), but that of the transept as well, with
cross vaulting sustained by pillars, did not make their appearance until the Xtth
century was well advanced. The two earliest dated examples are the one in the
cathedral of Speyer of 1030, and that under the existing cathedral of Parma. The
latter crypt (Fig. 744) belongs to the church rebuilt by Bishop Cadalus (1046-1071),
and consecrated in 1106. For though the church of Steinbach has a crypt which
extends not only to the crossing but also under part of the nave, it consists of mere
underground passages with arcosolia like the Roman Catacombs.

Further, we must remember that portals of the Lombardic type only came into

1 Man. Germ. hist. Annales Hildcsheimensis.

- Jacobs, Urkundenbuch des in der Graftschaft Wernigerode belcgenen Klosters Ilsenburg.

3 Jacobs, Urkundenbuch des in der Graftschaft Wernigerode belegmen Klosters Driibeck.

4 Kugler, Kleine Schriften.

GERMANY

295

existence about 1032 with that of Sant 1 Andrea at Montefiascone : consequently, an
entrance of this kind could not have been used for the abbey church of Quedlinburg
in 997.

Before leaving the church we may notice that the characteristic bases in the nave
have their counterparts in the architectural decoration found in the illuminated MSS.
from the VHIth or IXth to the Xlth century. Instances are the " Concordia
Evangeliorum " (IXth century) and the " Psalterium Aureum " (IXth century) at
St. Gall ; 1 the Evangelistarium written for the Emperor Henry IV (Cat. 78 A. 2),
and the Gospels of the abbey of Abdinghof (Xlth century) at Berlin ; the " Alcuin
Bible" of the VHIth-IXth century at Zurich; 3 the " Sacramentarium S. Gregorii

Fig. 744. Parma. Crypt ol the Duomo (1046-1071).

Papae" written at Freising (1052-1078), now at Bamberg; 4 and the" Evangelistarium
of Illinus " in Cologne Cathedral.

Just as in former days the conquest of the Lombard kingdom by Charles the
Great (774), so now the descent upon Italy (951) of Otto the Great (936-973),
with the double object of comforting the lovely and not inconsolable widow of
Lothair (946-950), Adelaide of Burgundy, and of renewing and strengthening the
Carolingian claims to the Imperial dignity, consummated by his coronation as King of
Italy (961) and Emperor (962), was the opening for Germany of an era of building
activity, though not so brilliant as the first. It derived its sustenance from the
free and direct communications re-established with Italy, an intercourse which
continued through the reigns of Otto II (973-983) and Otto III (983-1002), with the
latter of whom we conclude this section, as we have devoted a separate chapter
to the ecclesiastical architecture of Germany subsequent to the epoch of 1000.

1 St. Gall. Library of the old abbey. " Berlin. Library of the Museum of the Decorative Arts.

3 Zurich. Cantonal Library. * Bamberg. Koyal Library.

296

LOMBARDIC ARCHITECTURE

Fig. 745. Gernrode. Church of St. Cyriacus
Capital (Xllth Century).

THE ABISEV CHURCH OF ST. CYRIACUS AT GERNRODE was built by the
powerful Margrave Gero (f 968), born in Sgo. 1 It was quite complete in 968 (some
say in 961 2 ), so that the founder on his return from Rome, where he had deposited
his armour near the altar of the Prince of the Apostles, and had received from

the pope a relic of St. Cyriacus, was able to install
the first abbess. 3

The plan is that of a T-shaped basilica with
nave and aisles having apsidal endings, the transept
being very short, as in the early Roman basilicas.
It has an apse at cither end. That at the west was
rebuilt in its present form in the Xllth century, 4
but originally, no doubt, was of the modest type
of the western apse at Drubeck (877) which seems
to have been rebuilt between the Xlth and Xllth
centuries. It is flanked by two towers.

The skeleton of the outer walls of the original
church remains, here and there rebuilt, and altered
by the subsequent construction of the triforium,
when the arcades of the nave with their pillars were
remade and the windows altered.
In the interior, the eastern apse has a semi-dome, while the presbytery, like all
the rest of the church except the minor apses, western apse, and both crypts, has a
wooden ceiling.

Under the chancel is the crypt, roofed with a combination of rough unraised cross
vaulting and continuous barrel vaulting springing directly from the outer walls, and
supported in the centre by four piers with clumsy bases and rude moulded capitals.

The floor of the transept was raised by the insertion of an arcaded and vaulted
gallery. The nave is now separated from the aisles by four arches on either side
supported by two columns with a pier between them. The piers have moulded
capitals : the columns, on the other hand, are crowned by Corinthianesque capitals
with stiff, plain foliage with occasionally heads coming out of it (Fig. 745). They
have curious bases made up of rolls and hollow
mouldings.

Under the western choir apse is a crypt with
continuous cross vaulting supported by columns
with bases which in some cases have spur-leaves
at the angles, recalling those in St. Michael and
St. Godehard at Hildesheim ; while the capitals
are of the Lombardic cubical type, ornamented
o.n the plane surfaces with concentric grooving and
pairs of semicircles, or with foliage. The capitals
are sometimes replaced by corbel pulvins, cham-
fered and curled over at the ends.

The triforium has, on the nave side, columns
with piers between them bearing similar corbel
pulvins (Fig. 746), the prototypes of which are to be found at Mettlach ; while at the
two ends it has pairs of openings with plain or foliated funnel-shaped capitals.

Fig. 746. Gernrode. Church of St. Cyriaius.
Tub-in (Xllth Century).

1 Puttrich, Denkmale der Baukunst des Mittelalters in Sachsen.
'' Moil. Germ. hist. Thietmarus, Chronicon.

- Liibke, op. cit.
4 Dohme, op. cit.

GERMAN 7 Y

297

The original windows are round-headed and splayed on both sides. The
exterior of the eastern apse is marked off into two horizontal zones by a rude
stringcourse, and into three vertical compartments by pilasters and engaged shafts.

Whether the western towers (Fig. 747) were originally intended for staircases

747- Gernrode. Church of St. Cyriacus (Xth and Xlllh Centuries).

or for the bells, is impossible to say, as the highest stage with its two-light openings
is the result of an alteration. The lowest part, like the eastern apse, is marked out
by thick lesenas. The next stage, however, is decorated with arcading, both triangular
and round-headed. Both stages are lighted by single rectangular openings with

298

LOMBARDIC ARCHITECTURE

triangular or arched tops. The western apse with its blind gallery and west front
have replaced the original arrangement.

The masonry and artistic details of the church indicate three separate series of
building operations. Of the first of these, characterized by the rude art of some of
the mouldings, and an entire absence of ornament, we have already spoken. To
the second should, in my opinion, be assigned : the triforium galleries with the new
arcades which support them, the galleries in the transept, and the reconstruction of
the apsidal west front. The artistic details presented by these portions are separated
by a considerable interval from those of the age of Gero, and are occasionally
superior to the results at Quedlinburg. This second period may be placed at about
the middle of the Xllth century, and with it we may associate the font. I append an

illustration of one of its panels (Fig. 748).

To the third, that is to say to the second
half of the Xllth and the beginning cf the
following century, will belong the gallery for
the nuns on one of the sides, on account of
the way in which the cross vaulting is con-
structed, and the greater artistic refinement
shown in the capitals of its supports.

Gernrode is the earliest existing example
in Germany of a church with an apse at either
end. A still older one, however, was the abbey
church of Fulda, the rebuilding of which was
begun between 790 and 792 by Abbot Baugolf
(779-802), continued by his successor Ratger
(802-818) ("sapiens architectus ' >r ), and finished
by Eigil (818-822). It was dedicated in 819,
and destroyed by fire in 937- 2345 We learn that
it was formed by two basilicas set end to end
but separated by a transept, 6 and that each apse
had a crypt beneath it: "In eadem vero
ecclcsia duas cryptas magnifico opere conlo-
cavit, unam quae respicit solis ortum, alteram
quae solis occasum intendit." 7 We are even
told the name of the architect of these crypts, the monk Racholfus : " Racholfo
dictante magistro et monacho." 8

Other earlier instances were the abbey church of St. Gall (822-829), the
cathedral of Cologne as built by Hildebold (785-819) and finished or restored by
Willibert (870-889), and the cathedral of Hildesheim, erected by bishop Alfred
(851-874), and dedicated in 872, which had a crypt at either end and therefore two
apses facing one another. 10 In other countries a very early example was to be found
at Abingdon Abbey (675) ; and it may very well be that the architect of Fulda was
influenced by the English Benedictine model, just as later the designer of St. Cyriacus

Fig. 748. Gernrode. Church of St. Cyriacus.
Detail of font (Xllth Century).

- Afott. Germ, hist, Annales Fnldenses antiqm.
iVo. Germ. hist. Annales Sancti Bonifacii.

1 Afon. Germ. hist. Catalogns abbalitm Fuldensiitm.
3 Man. Germ. hist. Lambertns, Annales.

6 Men. Germ. hist. Annales Hildesheimenses. 6 Browerus, f'uldenses antiquitates.

7 Man. Germ. hist. Brim Candidas, Vita Eigilis abbatis Fuldcnsis.

8 Man. Germ. hist. Brim Candidus, De vita Aegili versibus explicata.
a Afon. Germ. hist. Annales Hildesheimenses.

10 Bertram, Geschichte des Bisthums Hildesheim.

GERMANY

299

may have derived his plan from St. Gall, which also suggested the round towers on
cither side of the western apse.

In addition to this it contains the earliest dated double splayed windows in
Germany. Those in the round church at Fulda (818-822) are not original. This
form of opening, the history of which we traced in our account of Bagnacavallo
(Vlth century), and to which a wide extension had been given by the Lombard
gilds, had already made its appearance north of the Alps, in the Prankish Empire
in the apse at Germigny des Pres (801-806) ; and also in England in St. Michael's
Church, St. Albans (about 950).

Gcrnrode further affords the first instance for Germany of towers treated with a
scheme of architectural decoration in the Lombardic manner, the prototype being the
campanile of San Satiro at Milan (876), and
likewise of apses marked off into horizontal
zones and divided vertically into compart-
ments by lesenas and wall-shafts. All of
them were ideas imported from Italy, but
carried out with such taste and ability as the
Teutonic craftsmen possessed. The fact is,
the lesenas and shafts are applied to the
apse not merely as buttresses, as they were
in Roman times, but for decorative reasons.
Italian churches of that age provided nume-
rous instances of apses embellished with
one or two tiers of lesenas and corbel arches.
Some of the earliest, to which we have called
attention, are those at Arliano (712-744),
Toscanella (739), and San Leo (881-882);
which last was, no doubt, seen by Otto the
Great when, after a long and desperate
resistance, he stormed the fastness, and
probably also by his right hand, the valiant
Gero.

No country outside Italy exhibited
towers embellished with lesenas, arched
corbel courses, and blank arcading, older than those at Gernrode. And though the
scheme applied was of foreign origin, a partly Teutonic character was given to it,
suggested by the ranges of arcading on the sepulchal chapel of Lorsch.

THE ABBEY CHURCH OF METTLACH was erected by Lutwinus the first abbot,
afterwards archbishop of Trier (695-713), in honour of the Virgin. It was rebuilt
by Hezzel on the model of the monastery church of St. Maximin, in the time of
Archbishop Egbert (977-993), and soon afterwards reconstructed by Lioffinus (987),
(Fig. 749 and Frontispiece) in imitation of Charles the Great's rotunda :" ct
Aquisgrani palacium mittens et exinde similitudinem sumens, turrim, que adhuc
superest, erexit." 1 2 3 4

In the upper story or triforium may be noticed I're-Lombardic cubical

1 Moil. Germ. hist. De rebus Trevirensibiis saec. 1'III-X libellus,

2 Man. Germ, hist. Ex miraculis S. Liutwini, aiidore moncuho Mediolaitmi.

3 Sammartano, &c. , Gallia cristiana Ecclesia Trcvirensis Mediolacus.

4 Hontheim, His/aria Trrvirensis dipitmatica et pragmatica.

Fig. 749. Mettlach. Abbey Church (987).

LOMBARDIC ARCHITECTURE

capitals hollowed out at the angles, and with a flower or other ornament on a console
projecting from the abacus. Others have the form of a simple cube (Fig. 750), or are
shaped like an inverted truncated cone. These capitals, decorated with con-
ventionalized vine branches, foliage, and interlacing, fairly successfully treated
though of monotonous design and frigid execution, carry very depressed corbel
pulvins carved with foliage, or in some cases plain and curled over at the ends. The
carvings on the pulvins recall those in the underground apsidal chapel of St. Servatius
at Quedlinburg (997-1021).

A noticeable feature in the church are the corbel pulvins with rudely curled e"ds.
They are derived from the crutch-shaped pulvins, a Lombard creation of the Xth
century, which make their first appearance in the towers of the cathedral of Ivrea

(973-1001 or 1002).
These are the oldest
dated examples that
I know, and are im-
portant for purposes
of comparison,
being earlier than
the chamfered speci-
mens at Gernrode or
those of similar form
in the round church
at Fulda.

TRILATERAL
CHOIR AND CRYPT
IN THE ABBEY
CHURCH AT ESSEN.
The abbey church
of Essen was
founded by Alfred,
bishop of Hildes-
heim (851-874), be-
tween 858 and 863, and was finished in 873. 1 2 We may infer from what survives
that it consisted of a basilica with nave and aisles, a very short transept, and central
and lateral apses.

The date at _which the well-known three-sided choir at the west end with its
facade was added is not known. It is generally believed to have been in the time of
the abbess Matilda (973-1011); but my view is that it was constructed when the
convent was rebuilt by the abbess Theophanu (1039-1056). Her memory remained
closely connected with the convent of Essen : " unde ibidem eius memoria semper in
benedictione erit." 3 On the same occasion the crypt was altered and extended east-
wards. It was consecrated in 1051 by Hermann II, archbishop of Cologne
(I036-I055). 4

The three sides of the choir (Fig. 751) have on the ground floor arches springing

1 Humann, Die fCunstwerke der Miinsterkirche zit Essen.

2 Leibnitius, Scriptores Bnmsvicensia illuslrantes Chronica etiscoporiim Hildenshcimensium, necnon
abbatum monasierii Sancli Michaelis.

3 Man. Germ. hist. Brunwilarensis monasteriiftmdatio.

4 Humann, Die Kunstwcrke der Miinsterkirche su Essen.

Fig 75- Mettlach. Gallery (987).

GERMANY

301

from piers copied from those in the rotunda at Aachen. The capitals in the gallery
above are Corinthian (ancient, and taken from Roman structures), Corinthianesque
with stiff, plain leaves and a dentilated course round the top, and freely treated Ionic.
Contemporary specimens of the Corinthianesque type arc to be found on the facade
(Fig. 752), which also contains examples of the Lombardic cubico-spherical pattern :
a fact which connects together the eastern part of the crypt, the west front, and the
atrium through which it is
approached.

The crypt under the
eastern choir is the work
of two distinct periods. To
the first belongs the central
part with its plain piers, to
the second the two ends
which have their supports
embellished with angle
shafts and fluting. The
greater amount of orna-
ment on the eastern sup-
ports is explained by their
proximity to the altar.

The earlier of these
periods corresponds to the
time of the abbess Matilda.
As a matter of fact we
know that, after the fire of
9-14 or 946, which must
have necessitated some re-
storation, a dedication of
the crypt took place. 1 The
second period will be that
of the abbess Theophanu.
Any one who compares the
carving on the capitals of
the piers at the east end

with the Ionic Capitals and Fig. 751. -Essen. Abbey Church. Trilateral choir (1039- 1056).

bead-and-reel moulding in

the gallery of the western choir will at once see such intimate relationship both in
design and execution that they may be regarded as contemporary. This point
established, these carvings cannot belong to the epoch of about 1000 for the three
following reasons.

(1) The form of cubical capital on the piers at the east end of the crypt is
decisive against that date, for the Lombardic cubico-spherical capital did not appear
in Germany till later. The date of its appearance at Essen is given by the capitals
in the western atrium of the church which forms part of the great works of
Theophanu.

(2) The form of the capitals in the rotunda at Mettlach, and the type of their
ornamentation, are evidence of the way in which capitals were treated in Germany

1 Humann, Der Weslbau des Miinstcrs zu Essen.

LOMBARDIC ARCHITECTURE

at the end of the Xth century. This manner is quite different from that of the Ionic,
Corinthianesque, and cubico-spherical specimens, to be found in the crypt, west front,
and atrium of the church at Essen.

(3) The west front of the three-sided choir exhibits an arched corbel course. The
earliest example of such a feature (either in the continuous form or divided into

sections by lesenas) in
Germany is to be found
in the abbey church of
Limburg (1024-1045-
1058). One might sus-
pect that it was due to
Italian builders, who,
according to some, 1
were responsible for
the whole structure.
We cannot, however,
admit that Italian
craftsmen had any
share in the work, for
the methods of con-
struction and the treat-
ment of capitals in
Italy at the end of the
Xth century were not
of this character.

We must accord-
ingly date this carving,
and consequently the
alteration of the crypt
and the erection of the
three-sided choir and
west front, in a period
subsequent to the time
of Abbess Matilda, in
other words in the
days of Theophanu,
about whose works of restoration and enlargement we possess definite information :
" E quibusTheophanu, virum se moribus agens, Asidense monasterium cum universis
eius officiis iam partim vetustate collapsis, ab ipsis fundamentis novo erigens opere,
mirabiliter amplificavit." 2 These works must have included the whole of the conven-
tual buildings, and necessarily, or rather essentially, the church as well ; for it is
notorious that in the Chronicles " monasterium " has the meaning of "church."

1 Ilumann, Die A'unstwerke der Miimterkirche zu Essen.
- MOII. Germ. hist. Bruntoilarensis monasterii fundalto.

75 2 - Essen. Abbey Church. Trilateral choir (1039-1056).

CHAPTER VII

THE LOMBARDO-RHENISH STYLE

WHILE the Lombard gilds in Italywere seeking by repeated experiments
to give an embodiment to their conception of a vaulted church, and
while in France the Benedictine Order was striving after a solution of
the problem how to cover every part of churches of large size with cross
vaulting, the master builders of Germany were concentrating their efforts, one may
say exclusively, on the creation of a peculiar ground plan. That plan, taken together
with the severe character and imposing form of the structure raised upon it, was
intended to endow with an Imperial dignity the Lombardo-Rhenish basilica, the
highest expression of German architecture in the Xlth and Xllth centuries. It was
an outward and visible sign of the Imperial idea, brought back to life among the
Teutonic peoples by Otto the Great (936-973), and not only affirmed but also made
good in a greater or less degree by his successors.

No one can look on the imposing towered piles of the cathedrals of Mainz, Speyer,
and Worms, or the solemn naves of their interiors, without being immediately
impressed by this fact. And so forcible is the result that, had the original conceptions
been carried through, and had their authors been able to combine grandeur of archi-
tectural form with the wealth of ornament exhibited by contemporary buildings in
Italy, and also with the constructive and statical knowledge of the Lombard gilds,
there is no ecclesiastical edifice of the Xlth and Xllth centuries which would have
stood the test of comparison with them ; excepting always the church of Abbot Hugo
at Cluny (1089-1130), standing unrivalled in its consummate majesty and pride.

Satisfied, however, with a single aim, the northern builders of the grandest
German churches, though in direct contact with Italy and considerably influenced by
Italian architecture, and though at times availing themselves of the services of her
craftsmen, took absolutely no interest in a rational and original solution of the problem
how to cover their structures with cross vaulting. By such a solution they might have
contributed towards the development and completion of the architecture which
immediately precedes the Pointed style, and was its source and origin.

Hence it is only at the beginning of the second thirty years of the Xllth century
that we find them making their first attempts to substitute cross vaulting for flat
ceilings over the wider spaces. An exception must be made in the case of the abbey
church at Laach, where the vaulting of the main spans must be explained as an
imitation, which was unique, of the Cluniac abbey of Vezelay (1096-1 104). And this
at a time when the Lombardic style had attained its completion in San Michele
Maggiore at Pavia, erected after the earthquake of ill?, and when Durham had seen
between 1129 and 1133 the combination of the pointed arch with diagonally ribbed

LOMBARDIC ARCHITECTURE

cross vaulting, and while in France the Transition, which opened the way for the
Pointed style, had already made its appearance. The latter was soon moulded by
the German School after its own fashion and to its own glory, reaching its culmination
in the new cathedral of Cologne (1248). All which forms the subject of the present
chapter.

THE ABBEY CHURCH OF ST. MICHAEL AT HILDESHEIM was begun by Bishop
Bernward (993-1022), but the precise year is not known. In 1015 the crypt was ready,

and was dedicated. The
church was consecrated
by the bishop in the
year of his death, but it
was only completed by
his successor, Godehard
(1022-1038), who per-
formed the dedication
in 1033. In 1034 it
was struck by lightning
("m on as t e ri um S.
Michaclis archangeli
fulmine combustum et
miserabiliter est dete-
rioratum"),and restored
by Godehard with a
fresh consecration in the
next year, which shows
that the damage done
was inconsiderable.
Injured by another fire,
and beginning to suffer
from the effects of time,
it was repaired and
altered by Abbot
Diedrich II in the days
of Bishop Adelogus
(1171-1190), who in
1 1 86 consecrated it

,123456

Fig. 753. Hildesheim. Crypt of St. Michael's^about 1010-1015 and
1171-1186).*

once more. 1

The church as

originally designed had

a nave with aisles, separated by a transept from the apse at either end. The transepts
were flanked by staircase turrets, and over either crossing rose a large tower. Beneath
the western choir apse is a crypt where the capitals of the piers which support its
roof are formed of a fillet and hollow moulding or an ovolo (Fig. 753)- The two

* The illustrations of Iliklesheim are from photographs taken by Herr F. H. Bo'deker.

1 Man. Germ. hist. Annales Hildesheimcnses. " Man. Germ. hist. Chronicon Hildesheiinense.

3 Man. Germ. hist. Thangmarus, Vita Bermvardi episcopi Hildesheimensis.

4 Alon. Germ, hist. -Wolferins, Vitae Godehardi episcopi Hildtskeiinensis.

5 Leibnitius, op. fit. Chronica episcopontm Hildcnsheimiiisiinn.

6 Bertram, Geschichte des Bisthums Hildesheim.

THE LOMBARDO-RHENISH STYLE

305

columns with Lombardic cubico-spherical capitals on either side of the present outer
doorway come from the upper church. In its midst lies the limestone sarcophagus
which the founder had ordered for himself during his lifetime. The bas-reliefs which
decorate the coped cover should be noticed, especially those of living creatures, among
which only the lamb with the cross on one of the gable ends is fairly successful. The
angels between tongues of flame or clouds on the sides of the cover are very rude work.

The western choir apse is the result of a reconstruction attributed to Adelogus.
The north arm of the
transept is occupied by
a platform supported
by rude unraised con-
tinuous cross vaulting,
above which are two
galleries one over the
other (Fig. 754). In
them may be seen
cubico-spherical
capitals surmounted by
deep abaci or rude
corbel pulvins.

This cross vaulting
with that in the crypt
brings to mind the
beautiful contemporary
vaulting with visible
arches in the crypt of
San Miniato al Monte
near Florence (1013),
and shows what a far
higher level the art of
vaulting had reached in
Italy in the Xlth cen-
tury than in Germany,
or indeed in any
country north of the
Alps.

This transept com-
municates with the
north aisle by two
arches supported by a
column with a Lombardic cubical capital. The south arm of the transept has lost its
end and staircase turret, but it has kept the two arches separating it from the
corresponding aisle (Fig. 755). The three arches which divide it from the crossing are
not original.

The eastern choir has disappeared, together with the subordinate lateral apses,
but the transept in front of it remains. Four great arches carry the central tower,
which has been altered. In the south arm of the transept the capital of the
column supporting the two arches leading into the aisle has its faces and angles
ornamented with spear heads.

VOL. II. X

F 'g- 754- Hildesheim. St. Michael's. North arm of western transept
(about 1015-1035).

306

LOMBARDIC ARCHITECTURE

Fig. 755. Hildesheim. South aisle of St. Michael's (about 1015-1035 and 1171-1186).

The nave is
separated from the
aisles by columns,
between every two
of which comes a
quadrangular pier.
The piers are origi-
nal. Of the columns
only two still
retain their original
cubico-spherical
capitals and un-
spurred bases. The
others have capitals
of the time of
Adelogus (Figs. 756,
757), which, with
their abaci, are
elaborately orna-
mented with scroll
work, foliage, sacred
and profane figures,
animals, &c. Their
bases are provided
with the charac-
teristic spur leaves
at the angles, which
occur so often in
Germany.

Except for the
vaulted platforms

in the transepts (and no doubt the half-domes of the apses) the whole church
had wocHen ceilings.

St. Michael's (Fig. 758) is not only an important monument for the history of

Fig. 756. Ilildesheim. Capital in thi rave of
St. Michael's (1 171-1186).

Fig. 757. Hildesheim. Capital in the nave of
St Michael's (1171-1186).

THE LOMBARDO-RHENISH STYLE

307

art, but also contains more than one characteristic feature. Some of these certainly
the Lombardic cubico-spherical capital had an important share in the formation
of the Lombardo- Rhenish style ; and this is why we include the church in our list of
Rhenish buildings, though it belongs geographically to Old Saxony.

We notice, in the first place, the plan of a double transept flanked by staircase
turrets, with a central tower over each crossing. Other churches before Bernward's
had been erected with two transepts, and one or even two central towers ; for instance
Saint Riquier (Centula) (793-798), and Saint Remy at Rheims (VHIth and IXth
centuries). But these had not the flanking towers with which St. Michael's was provided.

Fig. 758. Hildesheim. St. Michael's (about 1015-1035 and 1171-1186).

Next, we notice the arrangement of an arcarded platform or portico at the end
of the transept. It had been introduced in the case of the great transept of
Constantine's Vatican Basilica (Fig. 759). In the portico to the right Pope Damasus
(366-384) constructed his baptistery, while the one to the left contained chapels and
the tomb of Urban II (1088-1099). In Bernward's church, however, these transept
porticoes were surmounted by galleries.

The arrangement in St. Peter's influenced at a later date the architect of Cerisy
la Foret (1030-1066), from which it was copied in other Lombardo-Norman churches.
It was also present to the mind of the patriarch Poppo (1017 or 1019-1042 or 1045)
when building his cathedral at Aquileia, for the two arches still existing in either arm of
the transept were evidently intended originally not only to strengthen the lofty transept
walls but also to support two loggias which probably disappeared in the restoration
and alteration of the church by the patriarch Marquard between 1365 and 1381.

Another feature is the alternation of piers with columns, not in this case an

X 2

308

LOMBARDIC ARCHITECTURE

1AI1\ I. A

advance in the direction of the Lombardic church, but merely providing a firmer
support for the lofty and substantial nave walls. This expedient (an early instance
occurs in St. Demetrius at Salonica [Vth century]) we have discussed in our account
of Jumieges. It was introduced at Hildesheim under the influence of the Lombardic
movement, at that time specially active in Italy and France. Shortly before, it had
been employed in SS. Felice e Fortunato near Vicenza (985). And while St.

Michael's was in
course of erection, a
far more advanced
scheme was being
embodied in San
Miniato al Monte
near Florence (1013)
in the form of an
alterna t io n of
columns and com-
pound piers from
which started longi-
tudinal and trans-
verse arches. The
result was a sound
and well thought out
concatenation of the
entire structure.

Then we have
to remark the pre-
sence of the Lom-
bardic cubic o-
spherical capital.
Its introduction here
must be later than
1015, for in that year
only the crypt of
Bernward's church
was finished and
dedicated, and the
capital does not
appear in it.

I ought to
notice here that the

date of 1001 as the beginning of the constructive works at Hildesheim is wrong.
They cannot have been started earlier than about 1010, for it is inconceivable that
the foundations took so long to construct. Certain it is that the choir above the
crypt, the body of the church, and the eastern choir, must have been erected between
1015 and 1022, for it was only then that they were dedicated, and it was the
dedication of an unfinished building : " et ex parte dedicavit." l The church was
not finally completed till 1033, when a fresh consecration took place. 2

Fig- 759- Rome. Han of the Old St. Peter's (IVth Century). (From Bonanni,
" Templi Valicani Historia")

1 Man. Germ. hist. Chronicon Hildesheimense.

2 Mon. Germ. hist. Wolferius, Vita Godehardi episcopi Hildesheimensis.

THE LOMBARDO-RHENISH STYLE

We discussed the origin of the cubico-spherical capital when dealing with Sant'
Abondio at Como. We may refer here to its rapid diffusion in the German
lands, where it kept the carvers busy for two whole centuries,
and was only dispossessed by the Pointed style. And it pre-
served its form unaltered, it being very rare to find in Germany
the scalloped type. In German illuminated MSS. I have
never come across any representation of the cubico-spherical
capital till well on in the Xlth century. I may refer to the
" Sacramentarium S. Gregorii Papae " (1052-1078) in the
Royal Library at Bamberg, where it is depicted in an arcade. ^1:

Before leaving this interesting church, I should like to say
a few words about various important works of art ascribed
to the school founded at Hildesheim by Bernward, and carried """^"'--^ ''. .
on under his direction. The productions of this school appear
to me to be in part the result of arbitrary attributions ; and it
is desirable that they should be subjected to a fresh examina-
tion by some one who has made a special study of the subject,
and would treat them as a whole with the aid of new criteria
and a comparative method based on immediate knowledge of
contemporary work of the same class both Eastern and
Western. Meanwhile I shall confine myself to some observa-
tions on the celebrated bronze doors of the narthex of the
Cathedral, and the equally celebrated portion of a candelabrum
(Fig. 760) of the same metal (with a new top added in the last
century) which is in the church.

It is suggested that Bernward derived his idea for the
doors, with the story of Adam and Eve on the left side and
scenes from the life of Christ on the right, from those of Santa
Sabina at Rome, also having subjects from the Old and New
Testament, which he must have seen and admired when, in
1001, he was the guest of the Emperor Otto III in his castle on
the Aventine. In the same way, the candelabrum, with its
spiral band of scenes, was inspired by the Column of Trajan. 1
Unfortunately there is no mention of these works (which, it
has been recently noticed, 2 show such different treatment of the
figures) by Thangmarus, the bishop's contemporary biographer
and tutor ; nor is it stated that the doors made by Bernward,
and afterwards set up in the cathedral by Godehard, 3 are
identical with those before us. We might just as well assert,
and with more foundation, that the spiral candelabrum, the
supposed " columna aenea" of Bernward, 4 is one of the "duo
candelabra longa " which Adelogus gave to his cathedral. 5

Nor must we put too much reliance on the inscription on
the two middle bands of the doors, put there it is not clear

1 Grisar, Analecta Komana,

2 Humann, Zur Beurtheilung mittelaltcrlicher Kunstwerkc in Bczug auf
ihre zcitliche ttnd iirlliche Entstehung.

3 Moil. Germ. hist. Wolferius, Vila Codehardi.

4 Leibnitius, op. cil. Chronicon coenobii S. Michaelii in Hildesheim. Fig. 760. Hildesheim. Candelabrum
* RIon. Germ. hisl.Chroniion Hildesheiuienst. in the Cathedral (about Xllth Century)

3 io

LOMBARD1C ARCHITECTURE

when, for it is improbable that they would have been cast and fixed in place in 1015
when St. Michael's was not yet built. I believe that this inscription is of a piece with
the one on the great bell at Rivolta d'Adda (Xlth century) stating that the bell-
tower was added to the church in the Xth century, when the latter was not in existence.
Any one who compares the reliefs on the doors and candelabrum with the
carvings on Bernward's tomb in the crypt of St. Michael's will not fail to realize
the enormous interval between the artists responsible for the bronze work and
the sculptor of the sarcophagus, or to be convinced of the impossibility of their
being all of one date and the products of a single school. Not to speak of the
fact, which no one has yet noticed, that the candelabrum was designed with figures
at the angles of the base, though such appendages to the lower torus of a column
base did not reach Germany from Italy, where they originated in the Xth century,
before the second half of the Xlth century in the simple form of claws. And
the monuments show that it was not used by the School of Hildesheim till the
Xllth century. On the other hand, the rude carvings of the sarcophagus ill consort
with the far more advanced reliefs on the bronze cathedral doors ; and it is not
likely that the feeblest member of the school would be selected to execute
the tomb of its founder and master.

If I may hazard an opinion, taking account of the figures on Adelogus's
capitals in St. Michael's, which are less advanced than those of the candelabrum,

and still less so than those on the
cathedral doors, and bearing in
mind the numerous Xlth and
Xllth century carvings and
stucco work which we have
examined in Germany, I am
inclined to think that the can-
delabrum, if executed by German
artists at all, belongs, at the
earliest, to the last years of
Adelogus's episcopate, and that
the doors were not cast before the
Xllth century, in view, among
other things, of the treatment of
the nude which indicates an ad-
vanced stage of art.

STRASSBURG CATHEDRAL,
rebuilt by Bishop Werinherus of
Hapsburg (1001-1029) in 1015,
but only finished after his death,
about the year IO3I, 12 would
have furnished, had it survived,
an important link in the chain of
monuments which produced the
Lombardo-Rhenish style. As it
is, the reconstruction following on the repeated conflagrations between 1130 and
1176 spared next to nothing of that bishop's church; that is to say, merely building

1 Man. Germ. hist.Annalts Argenlinenses. 2 Dacheux, La catUdrale de Strasbourg.

Fig. 761. Strassburg Cathedral. Choir (Xllth Century).

THE LOMBARDO-RHENISH STYLE 3II

materials used, for instance, in the oldest part of the crypt. It is true that it has been

suggested that a portion of the Xlth century structure may be recognized in the

crypt under the choir (Fig. 761), planned like a miniature basilica with nave and aisles

separated by cruciform piers alternating with columns, with a barrel vault over the

central part, and cross vaulting in

the aisles. There is, however, no

truth in this, for not only is the

system of vaulting against it, but

even more, the artistic features of

the supports. In fact the Lom-

bardic capitals with scroll work

and well-rounded figures treated in

high relief are obviously later than

the carving in San Flaviano at

Montefiascone (1032) ; and I feel
sure that no one familiar with the
decorative carving of the first cen-
turies after the epoch of 1000 would
date these figure capitals in the
age of Bishop Werinherus. More-
over, we shall presently see what
was the manner of carving capitals
both in crypts and churches, in
the Rhine lands, in and about the
bishop's time.

THE ABBEY CHURCH OF
LIMBURG was due to the Emperor
Conrad II (1024-1039), who
entrusted Abbot Poppo (1020-
1048) with its erection. ;/There are
conflicting notices as to the date of foundation, with the result that Trithcmius l
puts it in 1024, Wurdtwein 2 in 1030, Bucelinus 3 in about 1031, and Browerus 4
in 1034. The choir altar was dedicated in 1039, and in 1040 the high altar
in front of the choir. In 1041 the crypt was finished and three altars consecrated
in it. 6 The church was completed under Henry III (1039-1056); to be
exact in 1045 7 or 1058,8 when it was dedicated. The architect was the monk
Gumbertus (f about 1036).* 10

We will select the date 1024 for the foundation, as best suiting the conditions
of ecclesiastical architecture in Germany in the first quarter of the Xlth century,
when churches on a large scale still had wooden roofs for the aisles, whereas shortly
afterwards they were occasionally, in the Rhine lands, covered with cross vaulting.

1 Opera historica Chronica insig nis monasterii Hirsaugiensis.

- Monasticon PalatinumLimburgitm ad Hartam monasterium Ord. S. Benedict i.

3 German fa topo-cfr.rono-stetnmalo-graphica sacra et profana.

4 Antiquitates et Annales Trevirenses.

5 Wilrdtwein, op. cit. Limburgum ad Hartam monasterium Ord. S. Benedict!.

6 Man. Germ, hist. Annales Spirenses. 7 WUrdtwein, op. cit. Limburgum ad Hartam.

* Man. Germ, hist. Deutsche ChronikenLimburgerAymalen.

" Sackur, op. cit. ' Manchot, Klosler Limburg an der Haardt.

Fig. 762. Limburg. Abbey Church. North arm of transept
(1024-1045-1058).

312

LOMBARDIC ARCHITECTURE

Nothing more than ruins of the abbey survives. The church had a cruciform
plan with nave and aisles, the former terminating in a square chancel flanked
by two apses projecting from the transept (Fig. 762). At the west end was
a narthex between two towers, to each of which a round staircase turret was
attached. The nave was separated from the aisles by two rows of columns with
Lombardic cubico-spherical capitals, ending with two massive cruciform piers
which, with the responds of the chancel walls, carried the square central tower.

A restoration of the basilica which has been published l shows an octagonal

cupola rising above the
crossing. It is due to the
author's imagination, for
the nave and aisles, transept,
and even the chancel, had
wooden roofs, as is obvious
to any observer, and there-
fore the lantern tower over
the crossing can only have
had a similar covering. An
octagonal cupola carried
on pendentives would have
been an impossibility with
the construction of the
church such as it was.

It is true, indeed, that
an erroneous interpretation
of a passage in an incom-
plete description of the
church printed in Wiirdt-
wein, 2 has given rise to the
idea that the aisles had
barrel vaults. There is,
however, not the least trace
of an impost course on the
aisle walls, and they are not
provided with the supports
for the transverse arches
which would have crossed
the vault at intervals and
given it stability.

Underneath the chancel was a crypt with cross vaulting (Fig. 763). The whole
building was decorated with blank arcading, lesenas, and arched corbel courses.

The distinctive feature of the church is the arrangement of western towers with
staircase turrets attached ; an idea apparently derived from Saint Riquier, where the
two central towers had similar adjuncts. The transept apses, not set exactly in the
line of prolongation of the aisles, are copied from the normal Lombardo-Norman
basilica. They had already appeared at Bernay. Lombardic influence is suggested
by the decorative use of the arched corbel course. This is the earliest dated instance
of its appearance in Germany.

1 Manchot, op. cit. 2 Op. cit.Limburgum ad Hartam monasterium Ord. S. Benedicli.

Fig- 763. Limburg. Abbey Church. Crypt (1024-1045-1058).

THE LOMBARDO-RHENISH STYLE

313

THE CHURCH OF ST. MARIA IM CAPITOL AT COLOGNE, founded about the
year 700 by Plectrudis, consort of Pippin II (f 714), was rebuilt in the first half of
the Xlth century, and consecrated by Leo IX (1049-1055) in IO49. 12 * 4 The latter
fact is confirmed by the actual presence of the Pope in Cologne at the time, and also
by the existence of a Bull of Leo's in a Register among the archives of the church,
granting indulgences to it.

The rebuilding must have taken place in the period intermediate between the
erection of the abbey
of Limburg (1024)
and that of the cathe-
dral of Speyer (1030),
on account of the form
of the supports and
the vaulting of the
aisles.

As designed in
the Xlth century, the
church consisted of a
basilica with nave
and aisles, a three-
lobed choir surround-
ed by ambulatories
and a western tower
flanked by staircase
turrets. The original
outline is shown by
the uniform character
of the masonry in the
portions of the old
facing to be seen in
the western towers,
nave, transept, and
apses.

The arrangement
of the western towers
must have been just
then the fashion in
Cologne, for we read
in Gelenius 5 that
Archbishop Anno II
St. Martin's.

The nave arcades (Fig. 764) have rectangular piers with engaged columns. The
roof was originally of wood, but it has been replaced by vaulting. The aisles,
strengthened on the outside by buttresses connected at the top by arches, have
unraised cross vaulting. At the ends of the aisles are two polygonal towers,

1 Gelenius, op. cit.

2 Schaefer, Das Alter der Parochie S. Maria im Kapitol ( Annalen lies hist. Vereinsf. d. Niederrhein, 1902).

3 Id., Beitrage zur Kiiliier Topographic tind Kirchengeschichte (Ki'ii. Quartalschr., 1904).
* Board, .y. Maria im Kapilcl zu Koln. * Of. cit.

l''ig. 764. Cologne. St. Maria im Capitol. Nave (1049).

(1056-1075) built two towers at the west end of Great

3*4

LOMBARDIC ARCHITECTURE

originally quadrangular at the base and polygonal in the upper part, flanking the
old nuns' choir, which was rebuilt after the fall of the western tower in 1637, and
partially enclosing the bell-tower in the middle of the west end. This tower,
closely bonded in its lower part into the nave walls, formed the narthex from which
the nuns' choir was entered through two doors in the staircase towers, now blocked up.
We may conjecture that it was raised in 1170, and so would be the bell-tower which,
according to Gelenius, 1 collapsed through decay in 1637.

The three-lobed choir (Fig. 765) has suffered from alteration which has affected
the whole of the sanctuary and the upper part of the transept, to which vaulting has
been added. It had, from the first, an ambulatory forming a continuation of the
aisles, and covered with cross vaulting which is old in the transept but reconstructed
in the sanctuary. Whether it had a crypt from the beginning is not known. The

F'g- 765. Cologne. St. Maria im Capitol. Choir (1049 and Xllth Century).

present one contains, indeed, supports just like those in the church, but it is
impossible to say with certainty whether the cross vaulting is of the Xlth century
or the Xllth.

It has been suggested 2 that all this eastern part of the church was at first
designed with cross vaulting in the ambulatory, half-domes for the apses, barrel
vaulting for the presbytery and transept, and a cupola rising immediately from the
arches of the crossing. The nave would have had a wooden roof. The tampering
to which this part of the church has been subjected does not allow of any certain
decision about such theories. To form one would necessitate an inter-comparison of
all the vaulting in the church, based on tests of the masonry made in the different
parts.

Meanwhile we are unable to explain why the architect vaulted the larger and
more complex part of the church, while giving a wooden roof to the smaller, which

1 Op. cit. * Board, op. cit.

THE LOMBARDO-RHENISH STYLE 1

was the easiest to deal with. All the more as this was not a case of extending the
choir into the transept in order to find room for the stalls of a great crowd of
monks, which might have been a reason for not confining a solid roof to the chancel.

Further, it looks as if the ability displayed by the designer of the rational system
of vaulting in the three-lobcd part was something superior to the ordinary construc-
tive and statical knowledge current among the German builders in the first half of
the Xlth century. Nor need we wonder at the application of simple barrel vaulting
to a choir in the second half of the Xllth century (to which the result in question is
attributed), seeing that in Great St. Martin's (Cologne), consecrated in 1172, it was
used not only for the rectangular bays in a choir of the same plan as that of St.
Maria, but also for the first bay of the nave, as can still be seen. In St. Martin's we
cannot say whether the crossing originally had, as now, a cupola resting on triangular
pendentives, for in
1 373 the tower suffered
from a fire in which
the bells were melted,
and was restored a cen-
tury and a half later. 1
If it had, the system
of barrel vaulting in
St. Maria, flanking a
cupola of exactly the
same form, may have
been suggested by St.
Martin's.

Finally, the later
strengthening of the
wall piers from which
spring the transverse
arches of the present
barrel vaulting, is a
reason for suspecting
that the latter is a
subsequent addition, and that originally the transverse arches merely supported the
wooden roof.

It is not known when the changes took place which gave the church its best archi-
tectural decoration, though some would place them at the end of the Xllth century
or the beginning of the next. We will confine ourselves to noticing that the external
open gallery at the summit of the eastern apse (Fig. 766), with shafts in twos and
fours, presents a close analogy with the one in the three-lobed choir of Great St.
Martin's, belonging to the rebuilding after the memorable fire of 1149, and that the
date of St. Martin's may roughly correspond to that of St. Maria. All the more as
the foliated capitals of Pointed style in the blank arcading on the apse of the latter
exhibit a less advanced art than those in the same position in the Church of the
Apostles at Cologne, erected after 1199.

St. Maria is the earliest instance of the aisles prolonged into the choir and
enclosing it on all sides. The plan was soon after adopted in Saint Remy at
Rheims (1036-1044), and rather later in Winchester Cathedral (1079-1093).

1 Gelenius, op. cit.

Fig. 766. Cologne. St. Maria im Capitol. Choir (1049 and Xllth Century).

316 LOMBARDIC ARCHITECTURE

Dchio and Von Bezold have before now instituted the proper comparisons
between the church and various Roman three-lobed structures of either simple or
colonnaded form. We will only repeat what we mentioned before in our accounts of
the Church of the Nativity at Bethlehem (327) and of the Palatine Chapel at Aachen
(796-804), that the three-lobed choir plan, derived from numerous ancient Roman
examples, is of Latin origin. Before its adoption by Justinian (527-565) for the
Church of the Nativity it had been used in the celebrated basilica of St. Paulinus at
Nola (end of the IVth or beginning of the Vth century), the idea being taken from
the small tri-apsidal basilicas (Illrd century) in the Cemetery of Calixtus, dedicated
respectively to SS. Xystus and Caecilia and to St. Soteris, the illustrious ancestress of
St. Ambrose, or perhaps from the " cella trichora " of St. Symphorosa on the Via
Tiburtina (Illrd century). 1

We will conclude our discussion of St. Maria by remarking that the three
towers of the west front may later have provided a suggestion for Hezilo, bishop of
Hildesheim (1054-1079), who rebuilt Alfred's cathedral (851-874). The most
striking feature of his new church was the great bell-tower of three stories at the
west end, flanked by two staircase turrets, between which on the ground floor was
an apse facing east, opposite to the main apse which faced the west ; the whole
corresponding in width to the nave and aisles. The tower was demolished in the
course of the last century as it threatened to fall. 2 3

THE CATHEDRAL OF SPEYER. The most generally accepted date for its founda-
tion by order of the Emperor Conrad II is 1030. By 1039 the crypt was ready, for
we learn from various sources that in that year the founder was buried in it. The
works went on under Henry III (1039-1056),* and they must have made considerable
progress by 1071, for the church was dedicated to the Virgin in that year by
Gundecar II, bishop of Eichstadt. 5 The notices about its completion differ, some
giving the credit to Henry III or Henry IV (1056-1 io6), 6 7 others to his successor
Henry V (i 106-1 125)."

Damaged more than once by fire, it was largely destroyed by the French in
1689. A drawing in the possession of Herr Schwa rtzenberger, the author of a recent
book on the cathedral, executed before the restoration began in 1756, shows that the
only parts then standing were the choir, eastern transept, the nave as far as the
fifth bay, and the narthex. Another view in the Museum shows that other parts
were missing, viz. the upper portions of the transept with its towers, and the whole
of the cupola over the narthex. The western end of the church must have been in
such a ruined state that most of it had to be taken down to save it from falling :
thus, for instance, only the lowest part of the towers is original, as the facing shows.
From another drawing in the Museum, showing the church as it was at the
beginning of the XVI Ith century, we learn that the cupola of the narthex was
octagonal, and that the towers had three stages, with openings of several lights
as in the eastern transept. The restorations and reconstructions from 1756 to 1858
have brought the cathedral to the state in which we see it to-day. 910

1 Stevenson, op. cit. 2 Man. Germ. hist.Annales Hildesheimenses.

3 Bertram, Zur Kritik der iiltesten Nachrichten tiber den Dombau ;u Hilde'.heim.
* Man. Germ. hist. Wipo, Vita Chuonradi II imperatoris.

5 Man. Germ. hist. Guiidecharus, Liber tontificalis Eichstelensis.

6 Man. Germ. hist. Aniiales Spirenses. 1 Man. Germ. hist. Annalcs Hildesheitnenses.
6 Man. Genii, hist. Gotifredits Vilerbiensis, Pantheon.

9 Zimmern, Der Kaiser-dom zu Steyer. 10 Schwartzenberger, Der Dom zu Speycr.

THE LOMBARDO-RHENISH STYLE

317

Beneath the choir and eastern transept is the imposing crypt. The beautiful
cross vaulting springs from columns bearing cubico-spherical capitals with a half-
round in relief on each face (Fig. 767). Beyond it is the new Imperial sepulchral
crypt, in which may be seen remains of the first Merovingian church, and the bases
of two of the piers in the upper church, showing how the original form has been
altered by the addition on the nave side of a half-pier with engaged column, the base
of which has simple spurs at the angles.

Both arms of the transept have ribbed cross vaulting. At the north and south
ends are arcades containing
chapels, taken out of the
thickness of the walls. Two
apses project from the .eastern
side outside the lines of pro-
longation of the aisles. The
western arch of the crossing
is strengthened by a sub-arch,
involving an addition to the
piers, apparently inserted
when the cupola was built.
Above the crossing is an
octagonal cupola carried, with
its drum, on niches which
form the transition from the
square base to the octagon.

The body of the church
(Fig. 768) is divided into nave
and aisles by arches sup-
ported on piers alternately
larger and smaller. Originally
they were of uniform section

and size : cruciform, with two

engaged shafts. The alter-
nate ones were modified
when it was decided to replace
the wooden roof with vaulting.
The shafts of the untouched
piers have cubico-spherical
capitals : those on the enlarged piers (when original) are Corinthianesque with the
leaves treated in a Byzantine manner, recalling those on the outer faces of the large
windows.

The nave, one bay of which corresponds to two in the aisles, has raised cross
vaulting, carried on wall and transverse arches. The aisles retain their original
unraised cross vaults.

To pass to the exterior (Fig. 769), the apse is embellished with blank arcading.
One of the shafts is carved in relief with two animals, men mounted and on foot,
trees, cauliculi, and intertwined snakes ; the whole very rude both in design and
execution (Fig. 770). The summit is encircled by an open gallery, where the capitals
reveal more advanced skill and a different artistic feeling from those in the blank
arcading below. The bases here, as in all the external open galleries, are spurred.

Fig. 767. Speyer Cathedral. Crypt (Xlth Century).

3 i8

LOMBARDIC ARCHITECTURE

The gallery ex-
tends along the
side walls of the
presbytery. The
rebuilt gable at
the east end con-
tains another
open gallery,
stepped so as to
follow the line of
the roof, and also
continued at the
sides. In these
extensions the
arches are com-
pound and not
simple as in the
apse and its im-
mediate neigh-
bourhood.

The towers
flanking the pres-
bytery have the
characte ri stic
helmed tops (a
German crea-
tion), the gables
being pierced
with triplets
enclosed by a
t re foiled arch,
a feature already
The tops are a later addition, as is shown by the masonry.

Fig. 768. Speyer Cathedral. Nave (Xlth and Xllth Centuries).

introduced at Laach.

The eastern transept has buttresses at the
angles of each arm, and another in the middle of
the walls. The carving round the large windows,
more advanced in the south than the north arm, is
superior to the relief we noticed in the apse. There
is the usual open gallery under the eaves, con-
tinued along the nave and round the west front.
Above it runs a cornice band of foliage and flowers,
in part original.

In the north arm of the transept the western
buttress contains at the top a tv/o-light opening
with a shaft supporting a lion. The capital is
cubico-spherical, and the base is formed by a similar
capital turned upside down. The idea of making
a base out of an inverted capital is of great antiquity,
as is proved by a Phoenician carving in the British

Fig. 770. Speyer Cathedral. Carving from
the exterior of the apse (Xlth Century).

THE LOMBARDO-RHENISH STYLE

3'9

Museum, of which I append
an illustration (Fig. 771).

Above the windows in
the aisles runs a corbel course
with lesenas at intervals ;
while the nave walls, as we
have said, are crowned by
the usual open gallery, with
which even the octagonal
cupola at the east end is
provided.

In the angle between the
north aisle and transept the
chapel of St. Afra was erected
by Henry IV between 1103
and HO6; 1 otherwise it was
constructed about 1097, when
Herimannus, bishop of Augs-
burg (1096-1133), gave the
emperor a relic of St. Afra. 2
In any case it was finished

/**T

. 77 1 ' British Museum.
Phoenician carving.

Fig. 769. Speyer Cathedral. North side (Xlth and Xllth Centuries).

by 1106, when the Emperor was buried in it. Opposite
to it on the south side of the church is the chapel of St.
Emmerammus (1088-1091). At present it is used as a
baptistery (Fig. 772).

Let us now proceed to question the Sphinx of the
Rhine, and compel her to reveal the main facts of her
story in the Xlth and Xllth centuries. The result we
will subject to the tests of historical, architectural, and
artistic criticism.

We have already seen that the crypt was finished in
1039, and that the cathedral was consecrated in 1071.
How much of it was really complete at that moment
we cannot say. But \ve do know that most of it had
been constructed when Henry IV, learning that the
stability of the structure was being endangered by its
proximity to the Rhine, commissioned (1082-1084) Benno
II, bishop of Osnabriick (1068-1090), to take the
necessary measures, which consisted in increasing the
thickness of the walls and piling up a bulwark of large
blocks of stone. 3 The type of church that Benno found

1 Schwartzenberger, op. cit. 2 Zimmern, op. (it

8 Alon. Germ. hist. Norberlus, Vila Bennonii // episcofi Osna-
brugensis.

320

LOMBARDIC ARCHITECTURE

in existence is easy to realize. It was a basilica with a wooden roof over nave and
transept, and vaulted aisles. The design showed an intelligent step forwards in the
development of the Lombardo-Rhenish basilica, for at Limburg (1024-1045-1058)
the aisles were still separated from the nave by cylindrical piers, and had wooden
roofs. In short, it was a structure of the type of the almost contemporary St. Maria
im Capitol at Cologne, though Speyer has the merit of more developed, i.e. cruciform
supports in the nave, instead of the simple ~f form.

It is true that Schwartzenberger, 1 who in his book has invented a chronology of
architecture and art, mostly of a fanciful kind, believes that the nave was vaulted as

well as the aisles, on account
of the form of its compound
supports. All the evidence of
actual buildings is against this.
Thus, for instance, at San
Miniato al Monte (1013), a
half-column rises from the
cruciform piers merely to
carry a transverse arch of
the nave, which, like the
aisles, has a timbered roof.
Again, at Mont Saint Michel
(rebuilt in 1020), Cerisy la
Foret (1030- 1066), Saint
F-tienne (1066-1086) and
Saint Nicholas (1080-1093)
at Caen, the compound piers
give rise to a vaulting shaft,
the original function of which
was to carry, not the vaulting
but the tie beams of the
timbered roof of the nave.
And, without going as far as
Italy or France, the architect
of the cathedral of Mainz, as
rebuilt by Henry IV after
the fire of 1081, carried up
vaulting shafts in the nave to
support a wooden ceiling.

Further, we saw that, in
the first half of the Xlth century, no basilica of large size had a complete system
of cross vaulting, even in the countries which had made most progress in vault
construction. And, from another point of view, it is out of the question that the
builders of Speyer should have begun by providing it with vaulting (a view shared, by
the way, even by Choisy 2 ), and then have taken the fancy to pull this down and
rebuild it either on the old lines or on a more intelligent system. Lastly, we cannot
imagine an architect so incapable as to design a nave of such width with cross
vaulting (which in that age was very heavy), and at the same time prepare to receive
it nothing more than vaulting shafts barely 2 ft. in diameter.

1 Op. cit. 2 Histoire de I 'architecture.

Fig. 772. Speyer Cathedral. Chapel of St. Emmeramrnus
(Baptistery) (Xlth Century).

THE LOMBARDO-RHENISH STYLE 321

Apart from the stone bulwark, Benno's work is apparent in the enormous thick-
ness of the outer walls of the crypt, produced by the new facing. This is characterized
by the different spirit in which the exterior of the sanctuary and transept is treated as
compared with that of the body of the church, which has not been altered. The two parts
arc entirely devoid of that intimate connection which marks a building constructed all
of a piece. Its absence may be noticed in the blank arcading round the apse, where
the inner arches, which ought to be the most elaborate, have plain Lombardic
cubico-spherical capitals (which occur in all the oldest parts of the cathedral), while
the outer ones have Corinthianesque capitals, and one of them is even ornamented
with carving.

It might be suggested that, in the course of the operations carried out by the
bishop of Osnabriick, the cross vaulting in the crypt was reconstructed. It is a fact
that this vaulting is of a distinct character, and superior to the nearly contemporary
work of the kind in St. Maria im Capitol at Cologne, and still more that in St.
Michael's at Hildesheim. But the men who worked at Speyer were recruited from
all parts (" fabros et ccmentarios aliosque opifices regni sui, vcl etiam de aliis regnis
in opere ipso habens"), 1 and the hands of master masons from Italy, especially from
the North, may explain the fine quality of the cross vaulting. However this may be,
the supports are the original ones, resting on bases without angle spurs (a feature
which I have not found in Germany before 1052, in the Minster of SchafThausen),
whereas this detail appears in the parts of the cathedral which have been subjected to
alteration.

In the matter of vaulting, Benno added it, or intended to add it, only in the
transept, as is indicated, I think, by the flat buttresses on the exterior of this part of
the church. What its nature was we cannot say.

Meanwhile, the completion of the great pile progressed somewhat slowly owing to
the negligence and fraud of the builders. The Emperor accordingly decided to send
to Speyer (1097) Otto, afterwards bishop of Bainberg (i 103-1 139), who was to put a
stop to this state of things and assume the supreme direction of the works. He set
things in order, and on this occasion he suggested an alteration in the windows. 23
The results of these modifications are obvious in the transept, e.g. the windows with
cable mouldings and spiral shafts, elaborately decorated with carving, and ill according
with the bareness of the recesssed windows in the apse or the absolutely plain ones in
the nave. The carvings which frame these transept windows display a higher grade
of art than the capitals in the blank arcading round the exterior of the apse, and must
be ascribed to some constructive and decorative enterprise of a later date than Bcnno.
Perhaps it may be connected with the erection of the chapel of St. Afra, that is to say
the period of Bishop Otto's supervision, for the carvings in either case are obviously of
the same date.

The bishop of Bamberg did not confine himself to restoring order among the
workmen and regularity in the works, or to providing more light for the choir : he also
gave a great impetus to the completion of the church, as the following passage from
his Life shows : " Non facile dici potest, quanta conservatio rerum facta sit, et quanta
structurae promotio." 4 The finishing touches were given, according to Godfrey of
Viterbo, in the reign of Henry V.

The notices of Herbordus and Godfrey are the latest transmitted to us by

1 a Man. Germ, hist. Herbordus, Dialogus de vita Oltonis episcopi Babenbergcnsis.

3 Man. Germ hist. Ebo, Vita Oltonis episcopi Babtnbergcnsis.

4 Man. Germ. hist. Herbordus, op. cil.

VOL. II Y

322 LOMBARDIC ARCHITECTURE

mediaeval writers concerning the building of the cathedral, which, as the original
parts show, certainly does not remain in the state in which it was left by the last named
Emperor. We will endeavour to supply this gap, if only in an approximate way, on
the sure basis of monumental evidence.

As we said, the cathedral was originally designed to have the main spaces
covered with wooden roofs. The question arises, when were these replaced by
vaulting? My belief is that the change was made after the fire of 1137, which was
most serious : " Ecclesia quoque Spirensis maior, cum parte non modica ciuitatis,
et oppidum Goslariense, eodem die quo et Moguntia, igne consumptae sunt." l And
it was before 1146 when St. Bernard preached the Crusade in the church. It cannot
have been after the fire of 1159, for the artistic features of the old parts of the
building show none of the characteristics of the third quarter of the Xllth century ;
for instance, spurred column bases with leaves of Pointed character, and capitals of
similar character like those in the three-lobed choir of Great St. Martin at Cologne,
which belongs precisely to that period.

At the time when the works of Benno and Otto were in progress at Speyer
the Lombard gilds were giving the finishing touches to the constructive and
statical revolution which had for its object the perfection of their vaulted basilica
type, by the completion of the church at Rivolta d'Adda and the Ambrosian Basilica
at Milan, achieved, the former under Urban II (1088-1099), the latter about 1098.
Having accomplished this they went on to create and apply an essential feature of
both the Lombardic and the Lombardo-Rhenish church ; I mean the elaborate open
galleries running round the exterior. Next, in the first quarter of the Xllth century,
they had succeeded in covering wide spaces of great extent with the cupola of their
creation, brought to perfection and in an embellished form ; and this we saw in our
account of the baptistery of Galliano. Outside Italy, Hezzel, the reputed architect of
the great church at Cluny (1089-1130), had confined himself to covering the ground
floor of each of the two octagonal towers in the main transept, which was only about
33 ft. wide, with an octagonal vault resting on hood-shaped pendentives. The
Italian examples must have encouraged some able architect, called in to repair the
damage suffered by the cathedral of Speyer in 1137, to suggest some safer form of
roofing, and one more in keeping with the conditions of ecclesiastical architecture at
the time. And it seems that this suggestion was acted upon.

While, however, the' structure of the transept, after Benno's alterations, with
the walls strengthened by buttresses, admitted of the erection of vaulting, that of
the nave was unequal to such a burden. Recourse accordingly was had to the
ingenious expedient which I will now describe. Every alternate pier was strengthened
by the addition of a half vaulting pier, as may be seen in the crypt, in order to
produce the Lombardic alternation of larger and smaller supports, and provide
starting points for the transverse arches and cross vaulting, the latter being of the
raised form in order to reduce the thrust, in the manner adopted as early as the
Xlth century in the aisles of San Babila at Milan, and both nave and aisles at Rivolta
d'Adda and Sant' Ambrogio. The bases of these additions to the piers had rude
claws instead of the leaves, sometimes of Pointed character, or animal heads to
be seen in other German buildings of the second half of the Xllth century. The
engaged shafts in two cases have original Corinthianesque capitals of Byzantine
character, recalling the manner of some of those in the transept windows.

The cross vaulting of the nave has been dated shortly before the end of the

1 Trithemius, op. cit. Chronica insignis monasterii Hirsaugiensis.

THE LOMBARDO-RHENISH STYLE 323

Xllth century. But the form which it took in the naves of the Rhenish cathedrals
at that time was very different from this, in proof of which one has only to study
the nave vaulting at Worms, the old parts of which were constructed by bishop
Conrad II (1171-1192), or that at Mainz (1183-1200) with its moulded ribs
and pointed transverse arches. And the same was the case in other German
cathedrals, e.g. that of Bamberg, which, after its consecration in 1012, was twice
burned with its wooden roof, and rebuilt with vaulting between 1185 and 1237.

From the original smaller piers and the larger ones resulting from this alteration
arches were sprung against the old walls to carry an additional facing of the upper
part of the walls on the inside ; and these walls were also raised in height to enable
them to resist the thrust of the vaulting. In the lunette wall spaces thus produced
small windows were formed to make up for the very moderate amount of light
admitted by the openings below, instead of enlarging these at the expense of the
stability of the nave walls. The walls, thus heightened and increased in thickness,
could now admit of the formation of the external open galleries.

At the same time, after first raising the height of the outer walls, the transept was
covered with ribbed cross vaulting, supposing that the existing ribs are not a later
addition to strengthen the vault. This raising in height was carried all round the top
of the building and the apse ; and hence the most ornamented capitals in all the open
galleries are evidently of the same date. As part of the same operations the eastern
gable, with its stepped gallery, was raised, and the helmed tops of the towers, with
their gables and triplet openings, erected. It was these gables that gave rise to the
helmed roof, so dear to the German builders, who must certainly be credited with its
invention, though, so far as I know, there is no specimen in existence older than the
Xllth century. Some people, indeed, have fancied that an example belonging to the
Anglo-Saxon age exists in the tower of Sompting Church (Xllth cent.), but it is not
of such an early date as that, though undoubtedly an importation from abroad.

The last step was to add the octagonal cupolas over the transept and the
narthex. Some writers 1 think that the former was part of the original design. I am
unable to share this opinion. Given a basilica planned for wooden roofs in nave and
transept, the crossing could only admit of a lantern tower roofed in the same way,
just like the central towers of all the Lombardo-Norman churches which we have
examined.

The ground plan of Speyer Cathedral represents the normal plan of the Lombardo-
Rhenish basilica, with its eastern and western pairs of towers and octagonal cupolas
over the crossings. The conception of a church confined by four towers, two at the
front and two over prolongations of the aisles beyond the crossing, might have been
suggested to the architect by San Lorenzo Maggiore at Milan "edita in turribus."
The suggestion may also have come to him from two sources in the following way :
the eastern towers from the cathedral of Ivrea (973-1001 or 1002) (unless it can be
proved that they were introduced still earlier in Cologne Cathedral), or Sant' Abondio
at Como (1013-1095) ; the western from St. Cyriacus at Gernrode (968).

The cupola over the centre of the narthex or western transept must have been
suggested by Angilbert's church of Saint Riquier (793-798), or by Saint Remy at
Rheims (VHIth and IXth cent), or, again, by Cologne Cathedral, supposing that
Willibert's church is represented in the illumination to which we called attention,
showing two towers (no doubt of wood) which must belong to the transepts. Or,
lastly, the source may have been St. Michael's at Hildesheim.

1 Lubke, op. dt.

Y 2

LOMBARDIC ARCHITECTURE

THE CATHEDRAL OF TRIER. As the old church of Agricius and Nicetius
threatened to fall, it was restored by Archbishop Poppo (1015-1047), who also designed
to lengthen it by a third towards the west, and took personal charge of the works.
Death, however, overtook him when the walls had just risen above the ground. The
new structure was continued by his successor, Eberhard (1047-1066), but how far we
do not know, and finished by Udo (1066-1068-1078). l The excavations made by
Wilmowsky 2 showed that the extension was terminated by a plain west front. From
this Archbishop Bruno (1102-1124) threw out the western choir apse, and dedicated
it in 1 12 1 to the Trinity and St. Nicholas. 3

The part of the cathedral which is of interest for our purpose is precisely this
apse (Fig. 773), where the recent restoration uncovered blank arcading round the

interior, while the outside is deco-
rated with arched corbel courses
broken up into sections by lesenas,
but is not encircled by an open
gallery. Now, given the im-
portance of the primatial church
of Gallia Belgica, and considering
Bruno's acquaintance with Italy,
whither he had gone in 1104 or
1106 to receive the pallium from
Paschal II (1099-1118), we may
safely say that the reason why the
latter feature did not appear in the
apse was that, although by tins
time diffused throughout Italy, the
land of its birth, it had not yet
reached Germany, where, in my
belief, it was seen for the first time
at Speyer, after 1137, and at Bonn
in the days of Pope Innocent II
(1130-1143) and the provost
Gerhard von Are (1126-1169).
We may, therefore, until the con-
trary is proved, relegate to the
domain of fable the theory which
would at all costs assign a Rhenish

Fig. 773. Trier. Western apse of the Cathedral (1121).

origin to this decorative motive, and give it a vogue in Northern Europe before it
found its way south of the Alps.

A noteworthy apse is that of St. Castor at Coblenz in the diocese of Trier, dedi-
cated in 1208 by Archbishop John (1190-1 2 1 2) (Fig. 774).* Round the base is a range
of blank trefoil arcading ; and above this another range of arches springing from
attached shafts, four of which rest on the backs of lions which the original specimens
show to have been of rude design and execution. An open gallery crowns the whole,
formed with isolated shafts interrupted at regular intervals by piers with attached
shafts, an arrangement suggested by the alternation of single and grouped shafts in
the open galleries of the apses at Cologne of about the same date.

1 Mon. Germ, hist. Gesta Treverorum.
3 Mon. Germ. hist. Gesta Treverorum.

2 Der Dom zu Trier.
4 Mon. Germ. hist. Notae dedicationum dioceseos Treverensis.

THE LOMBARDO-RHENISH STYLE

325

I may mention here that the apse of St. Castor, which must have been rebuilt
in the last years of the Xllth century, and the eastern choir apse of Worms
Cathedral, furnish the prototypes in Germany of ranges of blank or open arches
ornamented at the base with animal forms.

THE CATHEDRAL OF MAINZ. The new church of St. Martin, erected by
Archbishop Willigis (975-1011), was burned down on the day fixed for its con-
secration (loop). 123 The reconstruction was taken in hand by his successors

Erkenbald (1011-1021)
and Aribo (1021-1031),
while Bardo (103 i-
1051) added the
panelled ceiling (" a
tecto aedificare coepit,
sicque domum Dei
laquearibus, pavimento,
et parte fencstrarum
parietibus dealbatis,
dedicationis consecra-
tion! praeparavit "), and
the church was conse-
crated in 1 036 or 1O37. 45
That it was not com-
plete at that date is
shown by the fact that
the high altar was not
dedicated till 1049.'

After another fire
in 1 08 1, the Emperor
Henry IV (1056-1106)
began the rebuilding,
but did not live to see
its completion. 78 We
are not told how the
work went on after his
death, but it must have
been brought to an end
by Archbishop Adel-
bert I (1111-1137), for
he is stated to have

erected a magnificent wooden roof which was burned in 1137: "ncc mora, civitas
una cum principal! templo quod ipse magnifico tecto munierat, igne cremata est."
Damaged again by fire in the disturbances of 1 160, when Bishop Arnold (i 1 53-1 160)

1 Mon. Ger,,,. hist.Annales Wirziburgcnses. 2 Man. Germ. hist.-Annales Hildesheimenses.

" Man. Germ. hist. Lambertus, Annales.

4 Man. Germ, hist. Vulculdo, Vita Bardonis archiep. Mogunlini.

6 Man. Germ. hist. Marianas Scotttts, Chronicon.

6 Man. Germ, hist. Adam, Gesla Hammaburgensis Ecclesiae pon/ijic urn.

7 Man. Germ, hist. Vita Heinrici IV imperatoris. * A/on. Germ. hist.Annales Oltenburani.
Man. Germ. hist.Annales Palidenses.

Fig. 774. Coblenz. Church of St. Castor. Apse (1208).

326

LOMBARDIC ARCHITECTURE

met a violent death, 1 it was restored in the second episcopate of Conrad (1161-1165,
1183-1200), who saw his cathedral once more burned in iigo. 2 To him are
attributed, with good ground, the nave vaulting, now partly reconstructed, and the
rebuilding of the eastern choir. We also know that he began a new work, thought
to be the western choir, the completion of which he was prevented from seeing by
death. The final touches were given by Sigfried III (1230-1249), who consecrated
the cathedral in I239 34 (Fig. 775).

The oldest portions of the building are the lower parts of the restored eastern
towers. They have been assigned 5 to Willigis, or at latest to Bardo. The simple
and rude external decoration might point to the age of the former, but the fine
internal construction would lead one to ascribe it to the time of the latter.

F'g- 77S- Mainz Cathedral (Xlth, Xllth, and XHIth Centuries).

Next comes the body of the church, where the nave (Fig. 776), with the
exception of the vaulting, still represents the rebuilding of Henry IV. The aisle
walls were cut through in the XII Ith and XlVth centuries for openings to chapels,
and there is nothing left here save two original supports with cubico-spherical capitals
in the north aisle.

The width of the nave is about 50 ft. between the piers, which are of quad-
rangular shape and uniform size, and provided alternately with one and two half-
columns. The half-columns towards the aisles correspond to similar members on the
outer walls, so that we may reasonably infer that the aisles were vaulted from the
beginnning (Fig. 777). The half-columns of the alternate piers on the nave side

2 J a ffe, Bibliotheca rerun Germanicartini Monumenta Moguntina.
8 Joannis, Res Mogimtiacae. 4 Schneider, Der Dom zu Mainz.

5 Schneider, of. cit.

THE LOMBARDO-RHENISH STYLE

327

form awkward imposts for the cross vaulting, which has moulded ribs and slightly
pointed arches, and belongs to the second half of the Xllth century, or more
precisely to the second episcopate of Conrad who, as we saw, restored the church.
Though these half-columns do not form a structural part of the piers, and seem to be
a later insertion, nevertheless the piers are all of a piece, for the material used is the
same, and so is the construction. We may add that the expedient was adopted
for economy in stone, a fact familiar to every one acquainted with the practical side
of building. Above the impost cornice of each nave pier rises a broad pilaster
supporting the arcade
which carries the
clerestory. The idea
was originally sug-
gested by Saint Phili-
bert at Tou rnus
(1008-1019).

It has been
thought that the half-
columns carried up
from the alternate
piers show that the
nave was originally
designed for vaulting,
and that afterwards,
when this was found
to be unsuitable, it
was replaced by the
present vaulting;
though we know for an
undoubted fact that
Henry IV's church
had wooden roofs
over the main spaces. 1
The half-columns
were really carried
up, either to support
transverse arches,
supposing that their
present cubical capi-
tals are original ; or
else, as I believe, to

Fig. 776. Mainz Cathedral. Nave and eastern choir (Xlth and Xllth Centuries).

sustain the principal beams of the roof, in which case the half-columns were reduced
in height when the cross vaulting was constructed in the Xllth century.

I have already pointed out in connection with Speyer what a mistake it is to
suppose that a particular form of compound pier proves the existence of a system of
vaulting. The nave at Mainz was so little adapted for receiving vaulting that it had
to wait for the advent of a system of cross vaulting, the thrust of which was lightened
so as to make up for the defective and unsuitable organic structure of the building,
that is to say the system of raised cross vaulting here employed.

1 Man. Germ. liist.Annales Palidensts.

3 28

LOMBARDIC ARCHITECTURE

F 'S- 777- Mainz Cathedral. South aisle (Xlth and Xllth Centuries).

ordinary continuous unraised cross vaulting

The cathedral of
Mainz teaches a fact of
primary importance for the
history of German archi-
tecture, viz. that an edifice
commissioned by Henry
IV, the reputed Imperial
patron of the Rhenish
revival, was designed with
a wooden ceiling for the
nave. This is fatal to the
theory that, in his reign,
at Speyer and Mainz
vaulting was erected over
the main spans, only to be
taken down and rebuilt at
the end of the Xllth
century.

An interesting feature
is the Chapel of St. Code-
hard, to the north of the
cathedral, built by Adcl-
bert I. Its altar was con-
secrated in 1137 or 1138
by Burchard II (1120-
1149), bishop of Worms. 1
It is derived from San
Flaviano at Montefiascone
(1032), and, like it, is sur-
rounded by an aisle with

of this cathedral we will call attention to a

a gallery over it. Both parts have
(Fig. 778).

Before concluding our account
carving (Fig. 779), of the time of
Conrad's second tenure of the see,
from the eastern choir which he re-
built and caused to be decorated
with paintings, as a specimen of the
state of carving in Germany in the
second half of the Xllth century.

THE AUBEY CHURCH OF
LAACH was founded in 1093 by the
Count Palatine Henry II, with the
co-operation of his wife Adelaide.
After his death the works were at
first neglected by his step-son
Sigfried, but in 1112 he started them again with a generous contribution. On his
death in the next year, the Countess Hedwig devoted herself to the progress of

1 Schneider, op. cit.

Fig. 779. Mainz Cathedral. Carving in the eastern
choir (Xllth Century).

THE LOMBARDO-RHENISH STYLE

329

the building, which was finished in

the time of Abbot Fulbert (1152-1178),

and consecrated in 1156 by Illinus,

archbishop of Trier (1152-1 169), under

the invocation of the Trinity, the

Virgin, and St*Nicholas. 1J

The plan is that of a basilica with

nave and aisles, and a transept and

choir apse at either end. The pres-

bytery is flanked by two square towers,

and the arms of the western transept

arc terminated by two round ones,

Over the principal or eastern crossing

rises an octagonal cupola, while the

western one supports a large square

tower (Fig. 780).

The nave is separated from the

aisles by compound piers of uniform

size (Fig. 781). The capitals on the

shafts are either of cubico-spherical

form, or Corinthianesque carved with

birds, foliage, flowering plants, mon-

sters, billets, studs, interlacing, and

other ornaments. All is in moderate

relief, and of indifferent design and execution ; but we must remember that the stone

used is not
very suitable
for carving.

The crypt
has a nave
and aisles
separated by
supports with
spurred bases
and capitals,
either cubico-
spherical, or
Corinthian-
esque with
foliage, inter-
lacing, roses,
&c. Others
have already
remarke d 6
that the
cross vaulting

Fig. 778. Mainz Cathedral. Gallery of Chapel of
St. Godehard (Xllth Century).

Fig. 780. Laach. Abbey Church (Xlth and Xllth Centuries.
in the presbytery, main transept, and crypt, is more advanced than that in the nave.

1 Browerus, Antiquitatts et annales Trevirenses.
Trithemius, op. cit.Chronica imigitis monasterii Hirsaugiensis.

, . .

Kniel, Die benediktiner Abtei Maria-Laach.
5 Dehio and Von Bezold, op. (if.

.
4 Bucelinus, Gcrmania sacra et pro,

330

LOMBARDIC ARCHITECTURE

Perhaps the church was at first designed without a crypt, and, like San Michele
Maggiore at Pavia for instance, with barrel vaulting in the presbytery and eastern
transept. Later, the changes were made which gave the church its present
appearance.

The narthex or western transept is in two stories, and has an apse. The
atrium in front of it is a subsequent addition. With the exception of the semi-
domes of the two
apses, unraised cross-
vaulting is used in
every part of the
building. The ex-
terior of the eastern
apse and transept is
treated with blank
arcading. Through-
out, except on the
eastern apse, arched
corbel courses occur,
of larger or smaller
dimensions, and
either continuous or
broken up by lesenas.
It is generally
thought that when
the works were re-
sumed in 1 1 1 2, it was
intended to vault the
whole church, includ-
ing the nave. The
point seems to me
beyond the range of
doubt. What is de-
cisive is the fact
that the architect of
the nave at Laach,
which is only about
25 ft. wide between
the bases of the
piers, carried up

Fig. 781. Laach. Abbey Church (Xlth and Xllth Centuries).

pilasters about 4 ft. across to support the roof. Whereas the authors of the naves of
Speyer and Mainz, which are about 51 ft and 50 ft. wide respectively, contented
themselves with shafts of under 2 ft. in diameter for the same purpose ; and this
makes clear the different conception which they had before them in the erection of
their naves ; for it was their intention to cover the main spaces with wooden ceilings,
while the architect of Laach was all the time thinking of cross vaulting.

Nevertheless the church was not the product of a progressive study tending
towards the evolution of a new architectural organism, but the result of an imitation.
The absence from the eastern apse of the arcading which forms the principal
decorative feature of the rest of the edifice shows (as has been apparent to other

THE LOMBARDO-RHENISH STYLE 331

observers besides myself) that this is the earliest part, the remainder being assignable
to the resumption of the work in 1112. Now, by that year, there had come into
existence another Benedictine church, with nave and aisles covered with unraised
cross vaulting springing, as at Laach, from compound piers of uniform size : I mean
the church of Vezelay (1096-1104) as erected by Abbot Artald. Any one can sec
the striking resemblance which exists between the first experiment of the School of
Cluny in the way of constructing cross vaulting over a nave of large size, and the
church of Laach. In spite of its German ground plan, the decoration of the latter is
Lombardic ; and in the disposition of the nave and aisles with their vaulting, it is a
copy of the church at Vezelay.

It was just because it was the result of an imitation that Laach constituted an
isolated example in the German lands at the beginning of the Xllth century. A
long time will have to elapse before the German architects design a great church
planned from the outset, like that of Laach, for a complete system of cross vaulting.

But apart from its being an importation and an isolated instance, the fact
remains that at the time of its conception with its unraised cross vaulting, the churches
of Rivolta d'Adda and Sant' Ambrogio at Milan were already in existence with their
raised cross vaulting, partially ribbed as well ; while San Michele Maggiore at
Pavia was about to be begun, and mark the completion of the normal Lombardic
basilica.

We should observe at Laach the absence of an open gallery round the apse of
1093, which is only embellished with ordinary blank arcading. We have already
seen, while discussing the cathedral of Trier, that such galleries had not made their
appearance in Germany in 1121.

THE MINSTER AT BONN. The ancient collegiate church of SS. Cassius and
Florcntius was rebuilt by the provost Gerhard von Are (1126-1 169), J no doubt after
its property had been confirmed to it by Pope Innocent 1 1 (i 130-1 143). 2 It is generally
believed that nothing is left of his reconstruction except the apse and its two bell-
towers. The transept and the rest of the church are ascribed, by some to a rebuild-
ing after the injuries suffered by the church during the war between Philip of Suabia
(1198-1208) and Otto IV of Brunswick (1198-1212); by others to Gerhard himself
who, in the course of his long tenure of the provostship, was able to give his works the
stamp of the different styles which the church exhibits. One need, however, only glance
at the cross vaulting of the nave at Mainz, which we know was constructed in the
last years of the Xllth century, in order to feel sure that the similar vaulting of the
nave at Bonn, the supports of which were evidently planned to carry it from the
beginning, is not earlier than the XII Ith century.

The Minster is a basilica with nave and aisles, of Latin cross plan, with a choir
apse at either end. In front of the eastern apse is a presbytery of three bays.
Though there has been a good deal of alteration and reconstruction, there undoubtedly
survive of Gerhard's work, in a modified condition, the eastern apse with the towers
beside it, the presbytery, and the crypt beneath it. The apse is covered with a semi-
dome. In the bay immediately in front of it, flanked by the towers, the vaulting has
been reconstructed. The other two bays have been given a new dress in the Pointed
style, but the skeleton is the original one, as the exterior shows.

1 Maacken, Geschichte der Pfarreien des dekanates Bonn.

2 Neues Archiv der Gesellschaft fur iiltcre deutschc Ceschichtskunde, Vol. XIII Perlbach, Aus einem
verlorcnen Codex Iraditionum der Banner Miinsterkirche St. Cassias und Florentius.

332

LOMBARDIC ARCHITECTURE

The crypt (Fig. 782) under this part of the church extends to the towers, and
thus has a cruciform plan. It is divided into aisles by piers and columns, some of which
have cubico-spherical capitals with indentations at the angles, or are hung with plain
festoons and slightly undulated below, suggesting the scalloped capital which is very
rare in Germany.

I believe this crypt to be, on the whole, the work of Gerhard, but it has been
altered in the part under the sanctuary, the floor of which was raised when the bodies
of SS. Cassius and Florentius were exhumed, and the translation of other relics

recorded under the year
1 1 66 took place. 1

The apse (Fig. 783) is
decorated with arcading,
and at the top is an open
gallery with single shafts,
between every two of which
are coupled shafts. The
gable above is not original.
The towers, the tops of
which have been altered,
are covered with blank
arcading, arched corbel
courses, and lesenas. Here
there occur capitals with
crocket leaves, the earliest
which I have met with in
Germany.

An examination of the
side walls of the presby-
tery shows at once that
they were altered when the
transept and present nave
were built. Their structure
is on the whole original.

The element in the
Minster at Bonn which has
most interest for us is the
open gallery round the apse,
the dated prototype (to-
gether with those in the
cathedral at Speyer) of this feature for Germany, and marking the first step in the pro-
gressive history of the motive in the diocese of Cologne. The second is to be found in
Great St. Martin's in the same city, believed to have been founded in the VHIth century,
destroyed by the Saxons and the Normans in turn, restored by Archbishop Bruno
(953-965), and rebuilt by Archbishop Warinus (976-g84). 23 Anno II (1056-1075)
added two towers at the west end. 4 Destroyed by fire in IO49, 5 tne church was
rebuilt and consecrated in 1172 by Archbishop Philip von Heinsberg (1167-1191), and

1 Neues Atchro, loc. cit. - Man. Genii, hist, Calalogi archiefiscoporttm Coloniensium.

3 Man. Germ. hist. Chronicon Sancti Martini Coloniensis.

* Gelenius, op. cit. 5 Klinckenberg, Kiiln und seine Kirchen.

Fig. 782. Bonn. Crypt of Minster (Xllth Century).

THE LOMBARDO-RHENISH STYLE

333

finished by Abbot Simon between 1206 and 121 1. 1 The marked difference of
style observable in the interior of the three-lobed choir and the first barrel vaulted
bay of the nave, in which all the arches are round and the original capitals have
occasionally a suggestion of the Pointed style, as compared with the western part
of the nave (Fig. 784) belonging to the Transitional style, with capitals of well-
developed XHIth century type, provides an excellent reason for believing that
the body of the church was remodelled by Abbot Simon after the fire of 1185.*

The church of

the time of Philip von t ^ '

Heinsberg must be
supposed to have had
barrel vaulting only
in the first bay of the
nave, where the two
vault-shafts which
carried the transverse
arch may still be seen.
Its object was to re-
sist on this side the
thrust of the cupola
over the choir. The
other bays had a
wooden roof. Indeed,
the piers in this part
of the nave are not
designed for vaulting ;
so that when Abbot
Simon wanted to con-
struct its cross vaults
he was obliged to re-
sort to the expedient
of corbelling out high
up the shafts on which
they were carried.

I may notice in
passing that the octa-
gonal font (Fig. 785)
at the west end of
the north aisle, orna-
mented with scroll work, roses, and lions' heads, and traditionally regarded as a gift
from Pope Leo III (795-816), is a work which I should describe as coming from an
Italian hand, but not earlier than the XHIth century.

The same stage in the history of the external gallery is illustrated by the
three-lobed choir of the Church of the Apostles at Cologne (Fig. 786), rebuilt by
Archbishop Heribert (999-1021), and finished by his successor Piligrimus (1021-1036),
but reconstructed afresh by Adolphus I (1193-1205) after the fire of noxp. 3 Here,
just as in Great St. Martin's, the open gallery round the apse has a series of two piers,
each with two engaged shafts, alternating with a pier with four engaged shafts.

1 Bock, Rheinlaitds Baudenkmale. * Man. Germ. his/. Annales Floreffienscs. 3 Gelcnius, op. cit.

Fig. 783. Bonn. Eastern Choir of the Minster (Xllth Century).

334

LOMBARDIC ARCHITECTURE

Another instance at Cologne is the apse of St. Gereon, where there is said to
have been a church since the IVth century. 1 The present structure is to be ascribed
(i) to Archbishop Anno II (1056-1075), who threw out from the old round church a
long choir flanked by towers, with a crypt below consecrated in 1068, the church
being dedicated in the following year ; (2) to a remodelling of this choir and its
crypt in the last part of the Xllth century, completed in 1191, as we know that in
1190 the relics of the martyrs were deposited in the new crypt under the altar of
St. Gereon, which altar was consecrated in 1191 by Bertoldus, bishop of Metz

Fig. 784. Cologne. Nave of Great St. Martin's (Xllth and Xlllth Centuries).

(1180-1212). (3) Finally came the construction of the decagon forming the body of
the church, which was finished in 1227.- A *

Of Anno's work there remains intact the western part of the crypt (Fig. 787),
where the unraised continuous cross vaulting is carried on short columns with
unspurred bases and cubico-spherical capitals. The eastern portion of the crypt, the
" nova cripta " of the Annals of St. Gereon, forms part of the operations which were

1 Gelenius, op, cit. - Man. Germ. hist. Annales Sancti Gereonis Colonicnsis.

3 Moil. Germ. hist. Notae Sancti Gereonis Coloniensis,

4 Man. Germ. hist. Vita Annonis 77 archiep. Coloniensis, auctore monacho Sigbergensi.

THE LOMBARDO-RHENISH STYLE

335

brought to an end by the dedications of 1 190 and 1 191. It has raised cross vaulting
with visible arches, and bases with angle spurs of cither simple or elaborate character.
To the same date belongs the apse encircled by its open gallery (Fig. 788). The
apse is clumsily connected with the lower part of Anno's towers ; and though here
we find three light supports alternating with one heavy one, the gallery is closely
related to those of St. Martin's and the Apostles.

Fig. 786. Cologne. Church of the Apostles (Xlllth Century).

It ought to be mentioned that the arrangement of light and heavy supports in
the apse galleries at Cologne is derived from the arcaded galleries interrupted
by piers on the exterior of the Duomo at Modena (1099-1106) and the apse of
San Pietro in Ciel d'Oro at Pavia (1132), where, however, the galleries are not
continuous.

THE CATHEDRAL OF WORMS was in existence as early as the reign of
Dagobert, and was struck by lightning and burned in 872. The subsequent mis-

336

LOMBARD1C ARCHITECTURE

Fig. 785. Cologne. Font in Great St. Martin's (about
Xlllth Century).

solid foundation?, and beneath it a crypt
buried.

In 1033 Bishop Azzecho (1025-1044)
adjunct to the church. Bishop
Eppo expended large sums in
completing the unfinished works,
and at last the consecration took
place (l 1 10) in the presence of the
Emperor Henry V. Trithemius
gives the date as 1 1 18.

Bishop Conrad II (1171-
1192) repaired the building which
showed signs of falling, and a
fresh consecration took place in
the presence of the Emperor
Frederick I in 1 18 1. In the XVth
century the north-west tower fell
and was rebuilt, and in 1689 the
cathedral was burned by the
French. 1 2 3 4

Such, in brief, is the story of
the vicissitudes through which

1 Schannat, Historia episcopatus IVorm-
atiensis.

2 MS. Chronicle of the monastery of
Kirschgarten in the Archives of the City of
Worms.

3 Trithemius, op. fit. Chronicainsignis
monasterii Hirsau^iensis.

4 Man. Germ. hist. Vita Burchardi
episcopi.

fortunes of the city retarded and im-
peded the efforts of various bishops to
rebuild the church, until Burchard I
(1000-1025) demolished what had been
erected by his predecessors as being
unsuitable and mean, and began the
construction of a new cathedral in 1008.
The unfinished structure, which was
nearly ready for its roof ("iam pene ad
culmen eductum stabat "), was dedi-
cated in the presence of the Emperor
Henry II in 1016.

The western part of Burchard's
church fell in 1018, a fact not
difficult to understand when we re-
member that it had been built so
hastily that the bishop's biographer
thought it rose as if by magic. By
1 020 this portion was rebuilt on
was formed, in which the founder was

erected the chapel of St. Maurice as an

Fig. 787. Cologne. Church of St. Gereon. Western part of
the Crypt (Xlth Century).

THE LOMBARDO-RHEN1SH STYI.K

337

the cathedral has passed. Some J have thought that our church of SS. Peter and Paul
is still, on the whole, the one begun by Burchard I (the western towers being his),
continued through the Xlth century, and consecrated in mo. Others are of opinion
that the body of the church, with the exception of the nave vaulting, which they think
was subsequently reconstructed, belongs to Burchard ; the eastern choir apse, with
the transept and cupola, to Eppo ; and the western choir to a rebuilding in the early
years of the Xlllth
century.

Each of these
views when analyzed
falls to pieces. The
nearest to the truth
are Dehio and Von
Bezold, 2 who argue
that Conrad 1 1 rebuilt
the church with the
exception of two
towers, and that the
western choir was
rebuilt on Burchard's
foundations in the
XI I Ith century. My
belief is that, with
the exception of the
western choir, which
belongs to the early
XII Ith century, the
previous one being
apsidal, the cross
vaulting of the nave,
different in type and
execution from that
in the presbytery,
aisles, and transept,
and perhaps the in-
terior of the eastern
apse, the church as we
now see it is due to
Conrad II (1171-
1192). His work must

Fig. 788. Cologne. St. Gereon. East end (Xlth and Xllth Centuries).

have consisted, not in mere restoration of a ruinous building (" basilicam principem
ruinas hinc inde minitantem maximis sumptibus in priorem statum reduxit" 3 ), but in
its reconstruction.

An experienced eye will readily see the constructive and decorative unity of
the church, always excepting the western portion, which, apart from the towers, one
of which has been rebuilt, is later than the works of Conrad. Thus, arched corbel
courses (each arch being recessed), some continuous, some broken by embellished
lesenas or buttresses, occur on every part of the exterior. Everywhere the openings

1 Liibke, op. cit. Of. fit. 3 Schannat, op. fit.

VOL. II

338

LOMBARDIC ARCHITECTURE

are in groups, and the stringcourses are generally of saw-tooth design. Everything
points to its being a structure of one date, including the Lombardic cupola over the
crossing with its different masonry.

The eastern cylindrical towers cannot be assigned to Burchard's period, at any
rate so far as regards the stages of the old part (with their arched corbel courses
broken by lesenas, and saw-tooth mouldings) which rise above the lofty basements of
the towers. The early Xlth century is not the period which suits the recessed arches

of the corbel courses,
separated by lesenas
which are no longer
plain but moulded,
and in some cases
supported by human
heads, occasionally of
fantastic character.
In fact, one might
search in vain for such
arched courses with
figure corbels on any
Western building of
known date belonging
to the early years of
the Xlth century.

Further, these
towers are clearly con-
temporary with the
choir which they
flank, as is shown not
only by the masonry,
consisting at the base
of roughly dressed
stones, with higher up
a facing of carefully
dressed blocks of
varying dimensions,
but also by its deco-
rative treatment, and,
thirdly, by the carv-
ings of living crea-
tures, seeing that, on

more than one ground, we may regard as contemporary the carved heads on the
towers, the figures in the windows, those which project below the open gallery round
the apse, and the carving on one of its shafts of a man with an animal gnawing at
his head.

The polygonal exterior of the choir may be regarded as a casing over of the old
semicircular apse inside belonging to the church of the Xlth century. Now, for the
reasons given in our discussion of Speyer and Trier, the gallery round the choir, and
also that round the cupola over the crossing, cannot be earlier than the Xllth century.
Moreover, the Lombardic portal on the north side, with its foliage occasionally

Fig. 789. Worms Cathedral. Nave looking east (Xllth Century).

THE LOMBARDO-RHENISH STYLE

339

suggesting the Pointed style, and its figures sculpture, indicate a date not earlier than
that century. The same may be said of the canopy which formerly surmounted it ;
for Lombardic portals or porches with a canopy, or niche, or arcade above them are
not earlier than the erection of the cathedral of Modcna (1099-1106). In any case
the Lombardic portal did not appear in its developed form and elaborate dress before
the Xllth century, as we made clear in our account of Saint Guilhem du Desert.

To turn now to the interior (Fig. 789), the Lombardic arrangement of piers
alternating with piers having engaged shafts suggests a date which is not that of
Burchard I. It is inconceivable that such an important step towards the completion
of the Lombardo-Rhenish basilica should have been taken at Worms between
1008 and 1016, and that afterwards, in 1030, a retrograde one should have been made
at Speyer when Conrad II built St. Mary's to serve as the Imperial tomb house,
and another again at Mainz in 1081 when Henry IV's cathedral was erected.

But there are other reasons for excluding the age of Burchard. There are the
spurred bases of the half-columns, whereas
these adjuncts do not appear in St. Michael's
Hildesheim, which was built in his lifetime,
or in the later St. Maria im Capitol at
Cologne, or in Speyer Cathedral. There are
the ram's heads at the angles of some of the
bases in the transept, indicating a stage of art
which cannot be paralleled in any Western
monument of the early Xlth century. There
are the vaulting shafts, some 4 ft. in breadth,
arranged to receive the cross vaulting of the
nave ; whereas we know that in the naves of
German cathedrals of the Xlth century only
single shafts were carried up to support the
wooden roofs. Speyer and Mainz are ex-
amples. Then there are the well-known and
original carvings on the wall piers in the
presbytery, one of which, signed by Otto,
represents a woman holding the Devil by his
horns while an angel pierces him with a
lance, and another has projecting heads and scroll work (Fig. 790). No one could
possibly ascribe such work to German artists of the early Xlth century, seeing
that at the end of the Xllth they had only attained to the sort of carving found
in the eastern choir of Mainz. And, lastly, there is the cupola over the crossing,
carried on hood-shaped pendentives, first used for such a purpose north of the Alps
in Hugo's church at Cluny (1089-1130) and Saint Etienne at Nevers (1097).

Fig. 790. Worms. Carving in the Cathedral
(Xllth Century)-

The slow evolution of the Lombardo-Rhenish basilica as compared with the
Lombardic was not an isolated and merely local phenomenon, but was common to
all the contemporary German Schools, and to those of the neighbouring German
parts of Switzerland. We may set out the evidence in the form of a review of some
of the best known and dated churches of the period in these countries.

Z 2

340

LOMBARDIC ARCHITECTURE

THE ABBEY CHURCH OF HERSFELD was rebuilt after the fire of 1038 by
Poppo von Stable, and the crypt was dedicated in IO4O. 1 It was not finished till
1 I44- 2 The plan was that of a Latin cross, the nave and aisles being separated
by columns with cubico-spherical capitals, and Attic, unspurred bases. It had
wooden roofs.

THE MINSTER OF SCHAFFHAUSEN was founded by Count Eberhard and his
wife Ida. The choir must have been ready by 1052, for in that year Pope Leo IX

consecrated the

1HT JBH^HHHH'|^I|^BHE "^IHk ^H H^B principal altar. The

church was dedi-
cated by Rumoldus,

^^^^ bishop of Con-

stanz, in 1064. It
was erected from
the plans, and with
the advice of the
priest Liutbald :
" prefigura t ion e
atque adiutorio
cuiusdam Liut-
baldi." 3 *

It is of cruciform
plan, with nave and
aisles separated by
columns, and a
rectangular chancel.
The aisles are pro-
longed beyond the
transept in the
Lombardo- Norman
fas h i o n. The
columns have cu-
bico-spherical capi-
tals and spurred
bases. With the

exception of the semi-domes of the two small apses in the transept, the whole of
the church is roofed with timber.

This appears to be the earliest church of Germany or German Switzerland with
spurred bases for its supports. In those countries the spurs had from the beginning
the form of rudimentary claws, or of leaves. It was only in the Xllth century, after
the Lombardic School had produced forms so advanced as to be almost anticipations
of those of the Pointed style, e.g. in San Michele Maggiore at Pavia, that the German
chisels began to elaborate them with heads, paws of animals, leaves, sometimes
curling over as on the original bases in the transept of the abbey church of
Schwarzach (Xllth or Xlllth century), 5 and other ornamental forms.

Fig- 79 1 ' Cologne. St. Ursula. Nave looking west (Xlth Century).

1 Man. Germ. hist. Lamkrtus, Annales.

3 Man. Germ. hist. Annales Scafhusenses.

6 Sonder-abdruck aus dem Freiburger Diocesan- Archiv, Sauer, Die Abteikirche in Schwarzach.

* Liibke, op. cit.

4 Man. Germ. hist. Notae S. Salvatoris Scafhuscnsis.

THE LOMBARDO-RHENISH STYLE 341

If we were to accept the dates fixed by some writers for St. Ursula at Cologne
(Fig. 791), according to whom the oldest parts of the church go back to the rebuilding
by Archbishop Heribert (999-1020) and Richeza, Queen of Poland (f 1063), and its
completion by Anno II (1059-1075) who, we are told, used to spend much time in
prayer before the relics of the Virgins, 1 in that case the half-columns in the aisles
would have been provided with their spurred bases in the early years of the Xlth
century. But the existence in these aisles of cubico-spherical capitals, some orna-
mented with foliage or other forms, and also of spurred bases, whereas in St. Maria
im Capitol and Anno II's crypt at St. Gereon the capitals of this type are left plain,
and the bases are devoid of any ornament at the angles, affords good ground for sus-
pecting that the nave and transept of St. Ursula have absolutely no connection with
Heribert ; and that though Queen Richeza may really have laid the foundations of a
new church, she never saw it rise above the ground. This suspicion is confirmed by
the abundant and varied architectural decoration of the exterior of the nave and
transept, which can still be seen on the north side. It consists of arched corbel
courses, either continuous or broken by lesenas at various intervals, and blank
arcading. Whereas the exterior of St. Maria im Capitol has nothing more than a
range of blank arches in the aisles, and the original external face of the presbytery
in St. Gereon, one of Anno's works, has only two tiers of blank arcading.

Accordingly, I feel that all that is visible of the nave and transept of St. Ursula,
built as they were for wooden roofs though the aisles have unraised cross vaulting,
must be assigned to a date not earlier than the last years of Anno, and when the choir
of St. Gereon was already in existence (1069).

THE CATHEDRAL OF CONSTANZ. The old church having fallen in 1052, Bishop
Rumoldus (1051-1069) undertook its re-erection, according to Kraus 2 in 1054, and
was buried in it. The consecration took place in
1089 under Bishop Gebhard III (1084-1110).*

The original form was that of a basilica with
arcades and a wooden roof. The substantial
columns rest on Attic bases with rude leaf spurs.
They are surmounted by characteristic capitals,
neither cubico-spherical nor scalloped, but
spherico-polygonal, crowned by an octagonal
abacus (Fig. 792).

THE CATHEDRAL OF HILDESHEIM. The
original church, built by Bishop Alfred (851-
874) and consecrated in 872. was burned down
(1046) in the time of Bishop Azelinus (1044-
1054), who in 1047 began the re-erection which
he did not live to see finished. His successor, Fig. ^.-Constanz. Capital in the

Cathedral (Xlth Century).

Hezilo (1054-1079), continued the work, and the

consecration was performed in 1061. Bishop Bertholdus I (1119-1130) finished or

added the apse. 4 5 6 7

Hezilo's basilica, of Latin cross form, was designed, like St. Michael's, for a

1 Gelenius, op. (it.

- 3 Die Kunstdenkmdler des Grosshenogthums Baden Die Kiinsldentmaler des kieises Konstanz.

* Bertram, Gcschichte des Bisthums Hildeshcim.

6 Id., Zur Kritik der iiltestcn Nachrichten iiber den Dombau ztt Hildesheim.

Man. Germ. hist.Annales Hildesheimenses. 7 Man. Germ. hisl.Chronicon ffildesheimense.

342

LOMBARDIC ARCHITECTURE

wooden roof, and had its nave separated from the aisles by columns, between every
two of which came a quadrangular pier. Modern stucco capitals conceal the old ones.
The bases are Attic, and without spurs.

Hezilo erected another cruciform church in honour of St. Maurice, replacing an older
one of Bishop Godchard's (1022-1038), viz. the church of Moritzbergnear Hildesheim.
It had a wooden roof. Here again, the capitals of the columns are concealed by
stucco. The bases rest on plinths rounded off at the angles for convenience. The
water-leaf capitals in the crypt, being of the type in vogue at Hildesheim in Hezilo's

Fig- 793. Hildesheim. Church of St. Godehard (Xllth Century).

days, suggest that the now concealed original capitals in the upper church, as well as
those in the nave of the cathedral, are of the same pattern.

He also built the church of the Holy Cross at Hildesheim. It has a triforium,
and quadrangular piers in the nave. The main portions have wooden ceilings, but
the aisles are covered by rough barrel vaulting. The latter is to be explained by the
existence of the triforium ; otherwise the wooden ceilings would have been extended
to the aisles. Indeed, well on in the Xllth century at Hildesheim, both nave and
aisles in the church of St. Godehard were provided with ceilings (Figs. 793, 794).

St. Godehard's is a church of considerable interest. Bernward I (1130-1153)
laid the first stone in 1133 ; while Adelogus (1171-1190) carried up the towers at the
west end, where he consecrated a chapel of St. Mary Magdalen, built the western
apse, and finally consecrated the church afresh in 1172.

In the nave there is again the arrangement of two columns alternating with a

THE LOMBARDO-RHENISH STYLE

343

quadrangular pier. The angles of the column bases have the upturned claws (also
in Adelogus's work in St. Michael's), characteristic of the Hildcshcim School of the
Xllth century. Earlier it did not make use of this feature, as may be seen from the
old bases in St. Michael's, as well as those in the Cathedral and at Moritzberg. In
one case four animal paws are introduced. The capitals (Fig. 795) are of cubico-
spherical form, richly decorated with foliage, cauliculi, scroll work, flowers, discs, real
or fanciful creatures, and scenes from sacred history. They are characterized by
the same artistic quality as that of the capitals in St. Michael's, executed for Adelogus

Fig. 794. Ilildesheim. Nave of the Church of St. Godehard (Xllth Century).

before 1 186, and suggest that his munificence was not confined to the cases mentioned
by the chroniclers.

THE ABBEY CHURCH OF ST. AURELIUS, HIRSAU, was begun in 830 by Count
Erlefried and his son Notingus, bishop of Vercelli (827-830), and completed in 837.
As it gave signs of weakness, its reconstruction was begun by Count Adelbert and
his wife Wiltrudis in 1059, and in 1071 the dedication took place. 12 It consisted of a
nave and aisles (Fig. 796) separated by short columns (monoliths of local stone), two
western towers with a narthex between them, and aisles prolonged beyond the
transept, as in a Lombardo-Norman basilica. To each of the isolated columns corre-

1 Trithemius, op. cit. Chronica insifilis nwnastcrii Hirsatigiensis.
- Man. Germ. Hist. Historia monasterii Hirsaiigitnsis.

344

LOMBARDIC ARCHITECTURE

spends a wall shaft standing on a continuous plinth. The capitals are of cubico-
spherical form. There is cross vaulting in the narthex, but the nave had a wooden

ceiling, though the
aisles had unraised
cross vaulting.

Before leaving
Hirsau let us cross
the stream which
flows near the
church, and climb
the hill opposite to
get a view of the
tower (Fig. 797)
which is all that is
left of the church of
the great monastery
of St. Peter, begun
by William, abbot
of Hirsau, in 1082,
and finished in 1091
by the unaided
hands, it is said, of

Fig. 795. Hildesheim. 'Church of St. Godehard. Capitals (Xllth Century). tlle mon ks and lay

brothers. 1 2

This tower is regarded, and rightly, as of the Xllth century; 3 that is to say,
after the monks had recovered from the losses inflicted on them by Henry IV for
having taken the side of Gregory VII, and when they were in a condition to finish
the works of St. Peter's, which
cannot have been entirely com-
plete in 1091.

An interesting feature of the
tower is the band of carving
round the base of the third stage.
It contains atlantes supporting
lesenas (Fig. 798), and beside
them animals, a human being,
and the wheel of Fortune. In
spite of their rudeness, these
reliefs betray a more skilful hand
than that which produced the
carving we noticed on the exterior
of the apse at Speyer, and clearly
reveal Lombardic influence. They
form a piece of demonstrative evidence that the well-known reliefs with the Legend of
St. Vincent, and the arcaded altar front with pairs of Apostles in each arch, built into
a wall in Basel Cathedral, are not, as has been thought, coeval with its erection

1 Trithemius, op. cit. Chronica imignis mcnasterii Hirsaugiensis.

2 Man. Germ. hist. Historia monasterii Hirsaugiensis.

3 Liibke, op. cit.

Fig. 796. Hirsau. Abbey Church of St. Aurelius (1059-1071).

THE LOMBARDO RHENISH STYLE 345

(1014-1019) ' * under Bishop
Adalbert II (999-1021), but
must be dated not earlier
than the end of the XI 1th
century. And the date will
be the same, even if they are
regarded as works of Italian
or French origin.

THE CHURCH OF ST.
JAMES AT BAM BERG was
begun under Bishop Her-
mann I (1065-1075), and the
crypt was consecrated in 1072.
The dedication of the church
took place in IIO9. 84

It is of cruciform plan
with a choir apse at either end
(the eastern one having been
rebuilt in the Pointed style)
and wooden ceilings. The
nave (Fig. 799) is separated
from the aisles by columns
with cubico- spherical capi-
tals and Attic unspurred
bases.

The much altered church
of St. Gangolph in the

same town, erected by Bishop Gunther (io57-io65), 56 which now has clumsy

vaulting, also originally had a wooden

ceiling.

THE CHURCH OF ST. JUSTINUS

AT HOCHST ON THE MAIN. As'thc
old church was in a dangerous con-
dition it was demolished and rebuilt in
1090 by Adelmann, abbot of the
monastery of St. Albanus, in the days
of Ruthardus, archbishop of Mainz
(io88-ii09). 7

1 Robert, Gallia Christiana.
8 Gams, op. cit.

3 Man. Germ, hist. Notae Saudi facobi Baben-
bergensis.

4 5 Jaffe, Bibliolheca reruin Gerinanicartun
Monumenta Bambergcnsia.

6 Man. Germ, hist. Adalbert, Vita ct miracula
Henrici II iinperatoris.

7 Joannis, op. cit.

Fig- 797. Hirsau. Tower of the Abbey Church of St. Peterj
(Xllth Century).

Fie 798. Hirsau. Can-ing on tower of the Abbey
Church of St. Peter (Xllth Century).

346

LOMBARDIC ARCHITECTURE

The church is of cruciform plan, with wooden ceilings (Fig. 800). An inter-
esting feature are the Corinthianesque capitals, surmounted by Ravennate pulvins

of inverted truncated
pyramid form. This is
the earliest instance of
the occurrence of such
pulvins in any German
building.

THE CHURCH OF
ALPIRSBACH, erected in
1095, > s of Latin cross
plan, with the nave
and aisles separated by
columns with cubico-
spherical capitals. It
has a wooden roof.

THE ABBEY
CHURCH OF PAULIN-
ZELLE. We learn that
it was founded in the
time of the Emperor
Henry IV by St.
Paulina (f 1 107), who
was buried there before
the altar of the Holy
Cross, 1 and completed
in 1 119. A fire in the
XVIIth century, and
the abandonment of
the church, have reduced
it to the ruined state
which it now presents.
The nave and aisles

Fig. 800. Hochst am Main. Church of St. Justinus (1090).

were roofed with wood. The columns dividing them have cubico-spherical capitals
with pairs of semicircles on the faces, and spurred bases.

THE LlEBFRAUENKlRCHE AT HALBERSTADT, begun by Bishop Arnulfus
(996-1023), and consecrated in 1005, was still in course of construction in 1020.
Bishop Dietmarus (1089) bequeathed all his property for the completion of the
church. Bishop Rudolfus (1136-1149) rebuilt it, and performed the consecration
in 1145.

It is of cruciform plan, with a very short transept. Square piers separate the
nave from the aisles ; and, with the exception of the apse, the whole church was
designed for wooden ceilings. The cross vaulting in the transept and presbytery was
constructed after the damage suffered at the time of the destruction of the town by

1 Trithemius, op. cit. Chronica insignts monasterii Hirsaugiensis.

THE LOMBARDO-RHENISH STYLE 347

Henry the Lion in il/g. 1 Of its four towers, the western pair, embellished with
arched corbel courses, have German helmed tops.

I conclude with the expression of a wish that the new and wide field which I have
opened up in the domain of Monumental Archaeology may find a worker with the
will and the ability to cultivate it so that it may produce more fruit. Such an enter-
prise would be worthy both of the cause of knowledge and of the investigator ;
because the greatest of all the arts Architecture is the one which, by its creations,
preserves in the most tangible form the memory of great nations all through the
ages.

1 Lucanus, Die Liebfrauenkirche sit Halberstadt.

Fig. 799. Bamberg. Church of St. James (Xlth and Xllth Centuries).

INDEXES

INDEX OF PLACES

The numbers in heavy type refer to the illustrathns. .

The first nu'nbtr gen-.rally gives the nust important referenct to a subject.

Aachen

Palatine Chapel (Vlllth and IXth centuries), ii, 269-278,
717, 718, 719, 720, 721 ; i, 151, 152, 153, 154, 156;
ii, 3, 55, 58, 60, 63, 225, 285, 288, 299, 301, 316

Arcading round the dome (X[IIth century), i, 181 ;
ii, 272

Roman capitals, ii, 271

Abingdon

Abbey church of St. Mary (Vllth and Xth centuries), ii,
151-153, 128, 159, 163, 285, 298

Acerenza
Cathedral (Xlth century), ii, 253

Agliate

Baptistery (IXth century), i, 165, 27, 180, 238 ; ii, 37
Church (IXth century), i, 163, 222, 27, 117, 169, 173,

175 ; ", 37, 49

Agrafe
Castle of the Visconti (XlVth century), i, 166

Alpirsbach
Church (Xlth century), ii, 346

Ancona

Cathedral. Lions at the principal entrance (XHIth
century), i, 218

Angers

Cathedral (Xllth century), ii, 124, 513

Abbey church of Notre Dame de la Charite (du Ronceray)

(Xlth century), ii, 62, 427, 78, 79, 91
Saint Martin (Xlth century), ii, 49, 412

Antioch
Church of Constantine (IVth century), ii, 24

Antrim
Round tower (Xth or Xlth century), ii, 261, 705

Aosta

Castle of Bramafam (Xlth century), i, 166

Cathedral (Xlth, XlVth, and XVIIth centuries), i, iSS

267 ; ii, 32

Roman theatre, i, 247, 357
Sam' Orso. Cloister and crypt (Xllth century), i, 189

Aquileia

Baptistery (IVth century), i, 88

Cathedral (Xlth and XlVth centuries), i, 88, 172, 237,

189, 266, 307
Pluteus(XIth century), i, 214, 307

Ardmore
Round tower (Xllth century), ii, 261

Aries

Saint Trophime. Tower (Xllth century), ii, 36
Chapel of Saint Trophime (near Aries) (Xllh century),
"i, 34

Arliano (near Lucca}

Parish church of San Martino (Vlllth century), i, 116-
121, 159, 161, 162, 113, 123, 137, 156; ii, 299

Armagh
Great church and bell tower, ii, 259

Arshin (Syria)
Apse of church (Vlth century), i, 37

Assist

Amphitheatre, ii, 286

Santa Maria degli Angeli. Phiteus (IXth century), i, 174
Santa Maria del Vescovado. Rose window (Xllth cen-
tury), i, 136; ii, 121

Asti

San Giovanni. Pre-Lombardic figure capital in crypt
(Vlllth century), i, 138

Athelney
Abbey Church (IXth century), ii, 150, 158, 159, 177, 187

Athens

Acropolis. Capitals (Xlth century), i, 140, 192, 193
- Base of monument of Agrippa (1st century B.C.), ii,

37,528
Choragic monument of Lysicrates (IVth century B.C.),

ii, 252, 697
Church of

Holy Apostles (Xlth century), i, 27, 187, 262
Kapnikaraea (Xlth century), i, 27, 140
- Capital (Xth century), i, 217, 316
St. Xicodemus (Xlth century), i, 27, 140, 187, 198,

279, 280

St. Theodore (Xlth century), i, 27, 36, 45, 140, 187,
97

352

INDEX

the Virgin Gorgoepekoos (Xlth or Xllth century), i, t Benian (Aleria\

149, 209

Sculptures built into the exterior, i, 149, 206, 207,
150

National Museum. Sculptured slab, i, 149

Auckland

Church of St. Andrew. Gravestone (pre-Conquest), ii,
175

Basilica (Vth century), ii, 129

Bergamo
Santa Maria Maggiore (Xllth century), i, 159, 221, 241

Bernay

Abbey church (Xlth century), ii, 64-70, 429, 430, 433,
Aversa 434, 435. ii. 12. 72, 73. 75. 77, 7, 84, 160, 217, 230,

Cathedral. Choir (Xlth century), i, 222, 327, 328, 214, Sainte Croix, ii, 64
228, 237 ; ii, 37, 109

Avignon

Bethlehem
j Church of the Nativity (IVth and Vlth centuries), ii,

Notre Dame des Doms. Cupola (Xllth century), ii, 37,

109

Baalbeck
Temples (Ilnd and Illrd centuries), i, 8, 17, 48, 70, 71,

72, 73. 49, 74. 86

Christian basilica (IVth and Vlth centuries), i, 8 ; ii, 13

Babiska (Syria)
Eastern church. Gate in cloister (Vth century), i, 37

Bagnacavallo

San Pietro in Silvis (Vlth century), i, 92-94, 138, 56,
176 ; ii, 160, 251, 299

Ciborium (Vlllth century), i, 93, 139

Baiae

Circular thermal structures (Ilnd century), i, 23 ; ii, 41,
403

B amber g

Cathedral (Xllth and Xlllth centuries), ii, 323

St. Gangolph (Xlth century), ii, 345

St. James (Xlth and Xllth centuries), ii, 345, 799

Ban

Cathedral. Window in the apse (Xllth century), ii, 252,

698
San Nicola (Xlth and Xllth centuries), ii, 251, 253

Capitals, 215, 308, 251, 694

20-22, 385, 69, 316

Biella

Baptistery (Xth century), i, 1 78-^0, 245, 246, 247, 196,

240 ; ii, 37
Old church of Santo Stefano. Campanile (Xlth century),

i, 178

Billingham

Church of St. Cuthbert (IXth century), ii, 136
- Tower (Xlth century), ii, 136, 527

Binbir-kilisse
Churches, i, 47

Blandigny

Abbey church (Xth century), ii, 158

Bobbio

Church of the Virgin (Vllth century), ii, 135, 259

Bolam

St. Andrew's Church. Tombstone (pre-Conquest), ii,
149

Bonn

Minster (Xllth and Xlllth centuries), ii, 331-333, 782,
783 ; i, 244 ; ii, 324

Boschtrville

Barnack , Abbey Church (Xlth and Xllth centuries), ii, 95-100,

Church tower (Xlth century), ii, 180, 596, 597, 189, 190 ! 477- 479, 4.8o, 481, 482, 483, 484, 1 8, 93, 94

Barton on Humber

Church of St. Peter (Xlth century), ii, 181-185, 598,
599, 186

Basel
Cathedral. Reliefs (about Xlllth century), ii, 344

Bayeux

Cathedral (Xlth, Xllth, and Xlllth centuries), ii, 92,

472, 1 66
Library. Tapestry (Xlth century), ii, 166, 572, 573, 438,

136, 178, 589, 198

Beaune
Notre Dame. Cupola (Xllth century), ii, 109

Beauvais

Old Cathedral ("La Basse OZuvre") (Xth century), ii,
48-50, 413, 415, 61

Bee

Abbey church (Xlth century), ii, 85

Chapter House (Xllth century), ii, 99
Bo sham

Church (Xlth century), ii, 178, 589, 590, 591, 187, 190,
'93

Basra (Syria)
Cathedral (Vlth century), i, 10, 84 ; ii, 13, 15

Bracebridge
Church tower (Xllth century), ii, 186, 602

Bradford on Avon

Church of St. Laurence (Xlth century), ii, 173-175,

582, 583, 189
sculpture (after the Xllth century), ii, 250

Branston
Church tower (Xllth century), ii, 186

Breamore

Church of St. Mary (Xth century), ii, 179, 592, 593,
594

INDEX

353

Brescia
Cathedral, the old or " winter" (Vlllth century), i, 136,

"5 1
Rolomla (Xlth or Xllth century), i, 151, 240 ; ii, 37

Crypt of Sun Kilaslrio (Xlth or Xllth century), i, 136,

'5>

SS. Michael and 1'eter (now San Salvalore), i, 136
San Salvatore(VIIIth century), i, 136, 185, 113, 140,182

capitals (Vllh century), i, 136

MHM.-O Civico. Capital from the old Crypt of San
Filastrio (Vlllth century), i, 136, 187, 151, 164

Britford

Church (Xth century), ii, 179

carving (Vlllth Xth centuries), ii, 180, 595

Brixworth

Church (Vllih, Xth, and Xllh centuries), ii, 153, 555,
556, 557, 166, 189

Bury St. Edmunds
Abljcy church (Xlth century), ii, 163

Caen

Saint Ktienne (Xlth century), ii, 86-91, 464, 465, 466,
467, 468, 469, 470, 471 ; i, 237 ; ii, 37, 73, 85, 93,
94. 99, '75, 194, 196, 206, 218, 237, 238, 320

Saint Nicholas (Xllh century), ii, 93-95, 473, 474, 475,
476, 95- 98, 99, 238: 32

Trinite (Xlth and Xllth centuries), ii, 85, 460, 461, 462,
463- 73, 91, 26

Caenvent
Mosaics of the Romano-British period, ii, 145

Cagliari (Sardinia)
House of Tigellius, i, 84

Cairo
Futuh Gate (Xlth century), ii, 18

Cambridge

Tower of St. Benet's Church (Xlth century), ii, 192, 613,
614, 223

Canopus
Temple of Serapis, ii, 102

Canosa

Sepulchral chapel of Bohemond (Xllth century), ii, 252,
695

Canterbury

Cathedral (Vlth or Vllth, Xth, Xlth and Xllth cen-
turies), ii, 127, 195-197, 619, 620, 621, 91, 152,
170, 229, 232, 247

columns from Reculver (Vllth century), n, 151, 553
Abl>ey church of St. Augustine (Xlth century), ii, 195,

142, 185
St. Martin (Vlth century), ii, 126, 515

Font (Xllth century), ii, 127

St. Pancras (Vlth or Vllth century), ii, 128-131, 517,
518, 126

Carlisle
Cathedral (Xlth century), ii, 215, 225

Cashel

The Rock. Chapel of Cormac (Xllth century), ii, 263,
710, 7"
VOL. II

Casseneuii
Palace chapel (Vlllth or IXth century), i, 281

Castor
Roman building (Illrd Vth centuries), i, 165

Cattaro

Duomo. Archivoll of ciborium (IXth century), i, 157,
218

Cefalii

Cathedral (Xllth century), ii, 253, 699, 254

Cerisy la Foret

Abl>ey church (Xlth century), ii, 74-79, 446, 447, 448,
449- 45'. 4S 2 - ". '2, 33. 7, 72, 73, 4- 9, 93. 95.
7, 37, 320

Cerveteri

Tomb of the reliefs (IVlh century B.C.), ii, 199, 200
Carving in an Etruscan tomb, i, 216, 313.

Chaqqa (Syria)
Basilica (Ilnd or Illrd century), ii, 30

Chartres

Cathedral. Western lowers (Xllth century), ii, 37, 124
Saint Martin au Val. Crypt. Capitals (Vllth and Xllh
centuries), ii, 62

Chester

Cathedral (Vllth, Xth, Xlth, and Xllth centuries),
ii, 226, 664, 231

Chester le Street
Cathedral (pre-Conquest), ii, 176, 178

Chichester

Cathedral (Xllth century), ii, 248, 691
- Sculptured panels (Xllth century), ii, 248-250, 692,
693

Chttr
Cathedral. Crypt (Xllth century), i, 181

Cimitile (near Nola)

Basilica of St. Paulinus (I Vth and Vlh centuries), ii, 20,

29, 316
Chapel of the Martyrs. Porch (Vlllth century), i, 221

Civafe
San Pietro al Monte (IXth century), i, 173, 239

Cividale.
Cathedral. Baptistery of Callistus (Vlllth century),

i, mi. 149, 150, 140

Museum. Capitals (VHIih century), i, 138, 188, 189
Santa Maria in Valle (Vlllth century), i, 97-99, 145,

102, 113, 140; ii, 275
San Martino. Altar of Ratchis (Vlllth cenlury), i, 103,

Civitti Castellana

Cathedral portico. Front of sarcophagus (Vlllth century),
i, 144, 201

Classis (near Ravenna}

Basilica Petriana (IVlh and Vth centuries), i, 8.
Sant' Apollinare (Vllh century), i, 85-87, 131, 132, 45.
47, 5 6 , 6 S- 88 - 90. 209

A A

354

INDEX

Campanile (IXth century), i, 52, 53
Ciborium of St. Eleucadius (IXth century), i. 140,

191. '57

- Sarcophagi (Ylth-IXth centuries), i, 97, 100, 148,
101, 107 ; ii, 58

Ckt

Church tower (Xlth century), ii, 186

Clermont

Basilica of St. Antolianus (Vth century), ii, 60

Notre Dame du Tort. Cupola (Xllth century), ii, 109

Church of St. Namatius (Vth century), ii, 59

Clitumnus

Chapel of the, near Spoleto (IVth century), ii, 131, 522,
523

Clonmacnoise

Crosses (Xllth century), ii, 256

Gravestone, ii, 145

Round tower (Xllth century), ii, 261

Cluny

Abbey church (Xlth, Xllth, and Xlllth centuries), ii,
104-112, 488, 489, 490, 37, 38, 120, 121, 303, 322,

339
Saint Pierre le Vieux (Xth century), ii, 104; i, 210;

ii, 27, 160

Museum. Capitals (Xlth and Xllth centuries), ii, 109-

III, 496, 498, 121

Cobknz

St. Castor (Xllth and Xlllth centuries), ii, 324, 325,
774

Colchester

Museum. Sphinx of the Romano- British period, i, 220,
324

Cologne

Cathedral (VHIth, IXth, and Xlllth centuries), ii, 285,

730 ; i, 187 ; ii, 228, 251, 298, 304, 323
Church of the Apostles (Xlllth century), ii, 333, 786, 315,

335

St Gereon (Xlth, Xllth, and Xlllth centuries), ii, 334,

787, 788, 341
St. Maria im Capitol (Xlth and Xllth centuries), ii, 313-

316, 764, 765, 766, 117, 207, 320, 321, 339, 341
Great St. Martin (Xllth and Xlllth centuries), ii, 332,

784, 313. 315- 322, 33S
Font (about Xlllth century), ii, 333, 785
St. Ursula (Xlth century, ii, 341, 791

Como

Sant' Abondio (Xlth century), i, 206, 294, 295, 168,
227, 187, 188, 240; ii, 95, 113, 185, 204, 294, 309,

3 2 3
Basilica of SS. Peter and Paul, now Sant' Abondio (Vth

century), i, 29, 42, 206 ; ii, 68
San Fedele. Apse (Xllth century), i, 241, 352
San Giacomo (Xlth and Xllth centuries), i, 240, 350;

ii, 37
San Protaso (IVth or Vth century), i, 23, 130

Constantinople

Church of

the Apostles (IVlh century), i, 28

St. Irene (VHIth century), ii, 274. 722, 723, 79

St. John Baptist (Studium) (Vth century), i, 9, 25

St. Mary Diaconissa (Vlth or Vllth century), i, 197 ;

ii, 275

St. Mary Pammacaristos (Xlth or Xllth century), i. 155,
214, 197

St. Mary Panachrantos (IXth or Xth, and Xlllth or

XlVth centuries), i, 155, 212, 213, 187, 197
Myrelaion (convent of), (Xth century), i, 155, 197, 249,

364, 365 ; ii, 41, 252
St. Saviour in the Chora (Xlth or Xllth, and Xlllth

or XlVth centuries), i, 197
St. Saviour Pantepoptes (Xlth or Xllth century),

i, 140, 197, 277
St. Saviour Pantocrator (Xllth century), i, 155, 215,

197,217, 314
SS. Sergius and Bacchus (Vlth century), i, 66, 97, 67,

106, 69, 70, 81, 125, 83, 99, 197, 198; ii, 40, 275
St. Sophia (Vlth century), i, 7, 43, 63, 46, 60, 89, 62,

93, 63, 66, 96, 69, 107, 108, 70, 109, 71, 79, 124,

82, 84, 89, 197 ; ii, 4, 78, 275
Cistern of Arcadius (Vlth century), i, 13, 41, 59
- Basilica (Vlth century), i, 12, 41

of Binbir-direk (Vlth century), i, 13, 60, 88 ; ii, 284
Gate, the Golden (Vth century), i, 119, 160

of Rhegium (Vth century), i, 119
Museum, Imperial

Ambon from Salonica, fragments of (IVth century),

i, 43

Archivolt of cibonum, I, 147, 204
Stylophorous sphinxes from Sindjrli, i, 218, 320
Obelisk" of Thothmes III, pedestal of (IVth century),

i, '5, '4
Palace of Constantine Porphyrogenitus (Xth century),

ii, 49, 414

Constanz
Cathedral (Xlth century), ii, 341, 792

Cartridge

Church of St. Andrew (Vllth century), ii, 150, 551, 552

Cordova

Mosque, now the Cathedral (VHIth, IXth, and Xth
centuries), ii, 39, 238

Capitals from the Church of St. Vincent (Vlth
century), ii, 38, 39

Corneto Tarquinia

Etruscan carving, i, 105, 152 ; ii, 32
Museum. Etruscan sarcophagus, i, 216, 312
Museo Eruschi. Etruscan amphora, i, 216, 311
Etruscan tomb, ii, 66, 432, 83, 459

Cortona
Etruscan tomb, i, 119

Ctesiphon (Persia)
Palace of Chosroes I (Vlth century), i, 24, 31, 25, 195

Daphni (near Eleusis)

Convent church (Xlth century), i, 140, 194, 198, 281,
282

Deerhurst
Abbey church of St. Mary (Xth century), ii, I7'-I73,

577- 578, 579, !2i, '3, '79, 192
Chapel of the Trinity (Xlth century), h, 165, 571, 171,

"73

Denderah (Egypt)
Temple of Hathor. Portico (1st century), i, 11,5

Dijon

Abbey church of St. Benigne (Vlth, IXth, Xlth, and
Xllth centuries), ii, 5-12, 367, 368, 369, 370, 371,
372, 373, 374- 26, 27, 29-34, 36, 59, 60, 61, 64, 69,
91, 107, 113, 160, 211, 271, 285

INDEX

355

DtHtltittg

Wooden church (? Vlltli century), ii, 157

St. Mary in the Castle (Xlth century), ii, 187, 603, 130,
'79

Dowth

Tumulus, ii, 264

Driibeck

Church (IXth and Xlth or Xllth centuries), ii, 294,
743, *96

Dublin

Christ Church Cathedral. Crypt (Xllth century),
ii, 265, 266

Dunham, Great
Church (Xlth century), ii, 191, 611, 6l2, 223

Durham

Castle Chapel (Xlth century), ii, 198-200, 622, 623,

624, 625, 185, 215
Cathedral (Xth, Xlth, and Xllth centuries), ii, 161,

235-242,676,677,678, 679, 680, 681, 37, 88, 114,

123, 194, 200, 254, 303

Cathedral Library. Altar from Lanchester, ii, 138
Coffin of St. Cuthbert (Vllth century), ii, 147

Cross of Acca (Vllllh century), ii, 143, 538

- Cross from Gainford (IXth or Xth century), ii, 161,
566

Cross from Jarrow, ii, 139

-- Cross from Rulhwell (cast) (Xllth century), ii, 143,
539

- Grave-crosses from the old cathedral (Xth or Xlth
century), ii, 161, 564, 565

- Hexham, carving from, ii, 138, 142, 535
Hartlepool, gravestone from (pre-Conquest), ii, 145

- Hog-backed stones (pre-Conquest), ii, 149

Disert Aengus

Round tower (Xlth or Xllth century), ii, 261, 706
Earts Barton

Tower of the church (Xlth century), ii, 188, 606, 607
181, 223

Edinburgh
Museum. Altar from Birrens, ii, 138, 529

Ely

Cathedral (Vllth, Xth, Xlth, and Xllth centuries),
ii, 211-214, 645, 646, 647, 206, 223, 246, 247

Ephesus
Church of the Trinity, i, 63

Eseomb

Church of St. John (Xlth century), ii, 175-177,584,585,
586, 177, 178

Essen

Abbey church (IXth century), ii, 300
- Trilateral choir (Xlth century), ii, 300-302, 751, 752
- Crypt (Xth or Xlth and Xlth centuries), ii, 301, 302

Exeter
Cathedral (Xlth and Xllth centuries), ii, 159, 160, 223

Ezra (Syria}
Church f Si. U-.Mge (Vlth century), i, 10, 84 ; ii, 13, 15

Faltri

di Bi.ve, ii, 173
l'"it:i ili (jiovc, ii, 173

Fani Bergamasca
Church of Autharis (Vlth century), i, 112

Fieamf

Ablwy church (Xth, Xlth, and Xlllh ceni
ii, 46, 230

Ferrara
Cathedral. Portal (Xlllh century), ii, 121, 509

Firuz-Abad (Persia}
P.iUce (? Vlth century), i, 24, 193, 194, ,95, 276

Flavi&ny

Ablxjy church (VIHth and Xlth centuries), ii, 41-4-1
405, 406, 34, 39, 51, 57, 60, 64, 288

Florence

Archaeological Museum

Cippus from Settimello (Vlth century B.C.), i, 236, 348
htruscun carving, ii, 32

from Corneto Tarquinia, i, 105 ; ii, no, 495
I. ion from Vulci (Vlth century B c.), i, 220 323
Tomb of the " Uiavolino" (Vlhh century ri.r. i, i, 30
Inghirami Tomb from Vulierra (Ilnd or Illrd century
B.c ), ii, 284

Baptistery (Xlth century), ii, 69, 437

National Museum. Portion of the Kranks casket (VIHth
century), ii, 147, 544

Fulda

\bbey church (VIHth and IXth centuries), ii, 298
Round church of St. Michael (IXth century), ii 281-285
726, 727, 63, 288, 299, 300

Gainford
Church of St. Mary (IXth century), ii, 161

Gallerus
Oratory (IXth or Xth century), ii, 265, 712

Galliano

Baptistery (Xlth century), i, 190-199, 269, 27O, 179, 18?

202 ; ii, 322
Cnurch of S.in Vincenzo (Xlth century), i, 189, 268, 205 ;

Gerasa (Syria)
Ancient building, i, 34, 35

Germigny des Fres
Church (IXth century), ii, 54-59, 421, 422 423

164

Gernrode

Abbey church of St. Cyriacus (Xth and Xllth centuries),
ii, 295-299, 745, 746, 747. 748, 170, 183, 288, 294,
3o, 323

Glendalough

Cathedral (IXth or Xth and Xllth centuries), ii, 217

258, 259
Round tower (Xth or Xlth century), ii, 257, 258, 260,

704, 261

A A 2

356

INDEX

Church of

Our Lady (IXth or Xth and Xllth centuries), ii, 257,

258, 259

Reefeart (IXth or Xth century), ii, 257, 259
the Rock (IXth or Xth century), ii, 257, 259
the two Sinchclls, ii, 259
the Trinity (IXth or Xth and Xllth centuries), ii, 257,

258, 259

Gravestone, ii, 145
St. Kevin's Cell (not before the IXth century), ii, 257,

258, 264
St. Kevin's Oratory, House, or Kitchen (Xth or Xlth

century), ii, 257, 258, 261-263, 707, 265

Cross (Xllth century), ii, 256

Glentworth
Church tower (Xlth century), ii, 182, 186

Gloucester

Cathedral (VHth, IXth, Xlth, and Xllth centuries),
ii, 216-220, 652, 653, 654, 655, 656, 88, 96, 195,

221, 225, 228

Abbot's Lodging (Deanery) (Xllth century), ii, 220

Chapter House (Xllth century), ii, 220
- Slype (Xllth century), ii, 220

Grado

Baptistery (Vlth century), i, 95, 165

Cathedral (Vlth century), i, 94, 46, 88

Santa Maria delle Grazie (Vlth century), i, 95, 142, 88

Greenstead
Church (Xlth century), ii, 163, 568

Grenoble

Chapel (now crypt) of Saint Laurent, ii, 38-41, 399, 400,
52, 58, 59, 62

Halberstadt
Liebfrauenkirche (Xllth century), ii, 346

Hale, Great
Church tower (Xllth century), ii, 187

Hartlepool

Convent of St. Hilda. Gravestones (pre-Conquest),
ii, 145

Hereford
Cathedral (Xllth cenlury), ii, 216

Hersfeld

Abbey church (Xlth and Xllth centuries), ii, 340

Hexhain

St. Andrew (Vllth century), ii, 139-143, 149, 150, 151,
"S3, I/O

Capitals, Lombardic cubical (Xlth or Xllth century),
ii, 142, 536

Carving (Vllth century), ii, 139, 142, 143, 537

Episcopal chair (Vllth century), ii, 143, 540

Fragment of carving with intestinal interlacing,
ii, 148, 545

Fragment of stringcourse (Vllth century), ii, 142

Hog-backed stone (pre-Conquest), ii, 149, 548
St. Mary (Vllth century), ii, 149

Hildesheim

Cathedral (IXth, Xlth and Xllth centuries), ii, 341,
298, 316, 343

Bronze candelabrum (about the Xllth century),
ii. 309, 760

Bronze doors (above the Xllth century), ii, 309

Holy Cross (Xlth century), ii, 342

St. Godehard (Xllth century), ii, 342, 793, 794, 795,

296
St. Michael (Xlth and Xllth centuries), ii, 304-310.

753, 754, 755- 756, 757, 758, 107, 294, 296, 321, 323,

339, 34i, 343-

Htrstat

Abbey church of St. Aurelius (Xlth century), ii, 343,

796
Abbey church of St. Peter (Xlth and Xlth centuries),

"> 344, 797, 798

Hoc/is t am Main
Church of St. Justinus (Xlth century), ii, 345, 800

Ilsenburg
Church (Xlh and Xllh centuries), ii, 294, 742

Ispahan (Persia)
Sassanid capitals, i, 60

is-Sanainen (Syria]
Tychaion (Ilnd century), i, 22

Ivrea

Abbey church of Santo Stefano. Campanile (Xlth

century), i, 185, 256
Cathedral (Xth and Xllth centuries), i, 183-187, 252,

253, 255, 51, 188 ; ii, 30, 32, 251, 285, 286, 300, 323

Jdk
Church (?XIIIth century), i, 160

Jarroiv
Church of St. Paul (Vllth and Xlth centuries), ii, 138,

53, 531, 136, 182, 192
Collection of baluster shafts and carving, ii, 138, 142,

H8, 153

Jerusalem

Church of

the Ascension (IVth and Vllth centuries), ii, 14

the Holy Sepulchre (IVth, Vllth, Xlth, and Xllth
centuries), ii, 12-22. 375, 376, 377, 378, 379, 381,
382, 384 ; i, 8

St. Anne (Xllth century), ii, 18, 380, 122

St. Stephen. Capital (? Vlh century), ii, 20
Gate

the Double (Vlth century), i, 34

the Golden (Vlth century), i, 34 ; ii, 19, 21, 386
Mosque

El-Aksa. Capitals (Vlth century), ii, 19, 383, 21

of Omar (Vllth and Xlth centuries), ii, 14
Tomb of

Absalom (1st century), ii, 23, 389

the Judges (1st century), ii, 23, 388

the Kings (1st century), ii, 23

Jouarre

Saint Paul. Crypt (Vllth, IXlh, and Xlth centuries),
ii, 50-52, 416, 417, 39, 62.

Jumieges

Abbey church (Xlth century), ii, 80-84, 454, 455, 45^,

457, 38,68,88, 109, 113, 167, 308
Saint Pierre (Xth century), ii, 47, 410, 411, 34, 46, 64
Saint Valentin (Xllth century), ii, 96, 478

Kalat Sim-aan (Syria)

Church of St. Simeon Stylites. Apse (Vth century),
i, 36, 52

INDP:X

357

Kalb-Lauieh (Syria)
Church (Vlth century), i, 46, 86, 125 ; ii, 58

Kanawat (Syria)
Church (IVth century), ii, 30

Katura (Syria)
Tomb of Aemilius Reginus (Ilnd century), ii, 24

Kelts

Oratory of St. Columba ("St. Columb's House") (Xlh
century), ii, 263, 708, 265

Killaloe

Cathedral (Xllth century), ii, 263

Church of St. Klannan (Xlth century), ii, 257, 263,
709. 265

Kilimudiiagh
Church and tower, ii, 258

Kilmachedar
Church (Xlth or Xllth century), ii, 265, 713

Kirkdale

Church of St. George (Xlth century), ii, 165, 189
Gravestone (Vlllth or IXth century), ii, 165

Knoiuth
Tumulus, ii, 264

Koja Kalessi (hauria)
Church, i, 63

Kusr en Nueijis (Syria)
Tomb (Ilnd century), i, 66, 98 ; ii, 24, 390

Laach

Abbey church (Xlth and Xlllh centuries), ii, 328-331,
780, 781, 303, 318

La Charite
Abbey church (Xllh and Xlltli centuries), ii, 119, 109

Longford

Church tower (Xlth century), ii, 192, 615, 6l6, 190
Crucifix (Xlth century), ii, 193, 617

Lastingham
Church (Vllth and Xlth centuries), ii, 133, 199, 238

Lattakia (Syria)
Four-faced arch (Illrd century), i, 193, 275

Le Mans

St. Pierre de la Couture (Xth and Xlth centuries), ii, 61,
32, 33. 5', 185

Limburg

Abbey church (Xlth century), ii, 311, 762, 763, 302,
3 '3. 32
Lincoln

Cathedral (Xlth, Xllth, and XHIth centuries), ii, 202-
205, 630, 631, 225, 237,

Font, ii, 246, 687

Sculpture on the west front (Xllth century), ii, 245,
686, 248

St. Mary le Wigford. Tower (Xlth century), ii, 185
St. Peter at Gowts. Tower (Xlth century), ii, 185

l.indii/arnt
Cathedral (Vllth and ? Xth centuries), ii, 134, 135, 163,

Gravestone, ii, 145

Priory Church (XIlh-XMth centuries), ii, 135

Littleton

Villa of the Komano-British period, i, 165
London

Abbey church of St. Peter, Westminster (Xlth and Xllth
centuries), ii, 166-168, 188, 195, 198

Chapel of the Pyx (Xlth century), ii, 167, 205, 215

St. Bartholomew, Smithfield (Xllth century), ii, 21*

British Museum
Cross of Cynibalth from Lancaster (pre-Conquest),

ii, 148

Franks casket (VIHth century), ii, 147, 542, 543
Gravestones from Hartlepool (pre-Conquest), ii, 145
Mosaic of the Komano-British period, ii, 145
Phoenician carvings, ii, 43, 407, 318, yyj

Tower of London. Chapel of St. John (Xllh century)
ii, 207-210, 637, 638, 639, 640/641, 642, 643, 215,

Lorsch

Abbey church of St. Nazarius (VIHth and Xllth

centuries), ii, 286, 287, 28
Sepulchral chapel of Louis III the Saxon (IXth century),

ft, 286-288, 732, 733, 292, 299

Lough Derg

Round tower on Station Island (Xlth or Xlllh century),
ii, 261

Lucca
San Frediano (Xlllh century), i, 241, 243

Lugnano in Tererina
Church of the Crocifisso (Xllth century), i, 135, 184

Lyminge

Church of SS. Mary and Ethelburga (Vllth and Xth
centuries), ii, 133, 160, 563, 162, 170, 176, 179

Lyons

St. Martin d'Ainay (Xlllh century), ii, 34, 43

Chapel of Sainte Blandine (Xth century), ii, 43, 408.
34, 54, 61, 62, 64

Madaba (Syria)
Church (Vlth century). Mosaic, ii, 13, 375

Mainz

Cathedral (Xlth, Xllth, and XHIth centuries), ii, 325-
3 2 8, 775. 776, 777. 779, "'J, 33, 3 2 , 3*3, 33. 33 1,
339

Chapel of St. Godehard (Xllth century, ii, 328, 778

Malmesbury

Abbey church (Vllth or VIII, Xth, and Xllth centuries),
ii, 157, 242, 682, 237

sculptures of the porch (Xllth century), ii, 243-245,
683, 684, 685, 246, 247, 248

Malvern
Priory church (Xllh ccnlury), ii, 215, 650, 113, 231

Afawgan in Pyder

Wheel-head cross from Gwinear (Xth century), ii, 148,
54<5

358

INDEX

Mettlach

Abbey church (Xth century), ii, 299, 749, 750, 50, 190,
283, 291, 296, 301

Milan

Castello Sforzesco. Sarcophagus from Lambratc (IVth

century), i, 207, 296
Church of

Sam' Ambrogio (Vlllth, IXth, Xlth, and Xllth

centuries), i, 227-238, 337, 338, 339, 340, 341, 342.

343, 344. 5, '3', 156. 164, 166, 169, 173, 175,

200, 204, 291, 225, 226, 245 ; ii, 37, 109, 113, 118,

121, 240, 253, 322, 331

Ciborium (XHIth century), i, 146

Cupola (Xllth and Xlllth centuries), i, 73, 229

Basilica of Fausta (supposed) (Vth or Vlth cen-
tury), i, 20
San Babila (Xlth century), i, 199-202, 283, 284, 285,

286, 81, 176, 180, 214, 215, 221, 224, 225, 226, 228 ;

". 32, 37, 79, "4, 3 22
San Calimero. Apse (IXth or Xth century), i, 166,

175 ; ", 37
San Celso. Apse (Xth century), i, 117, 166, 175, 202 ;

", 37
Sant' Eustorgio (IXth or Xth and Xlth' centuries),

i, 174-176, 241, 166, 178; ii, 37
San Lorenzo Maggiore (Vlth century), i, 72-74, III,

112, 46, 56, 65 ; ii, 55, 323
Cupola (Xllth century), i, 72, 196

Chapel of Sant' Aquilino. External gallery round

cupola (Xlth century), i, 240, 349 ; ii, 224
San Satiro (IXth century), i, 168, 169, 228, 229, 232,

53, 164, 172 ; ii, 32, 183, 299
San Vincenzo in Prato (IXth and Xlth centuries),

i, 166-168, 224, 226, 117, 173, 175, 176; ii, 37

Capitals (Vlllth century), i, 138, 168, 225

Monastero Maggiore. Tower (IXth century), i, 169,

230, 231

Thermae Ilcrculeae of Maximian (Illrd or IVth century),
i, 201

Modena

Cathedral (Xlth and Xllth centuries), i, 97, 234 240
351, 243, 245 ; ii, 83. 335, 339

- I orta de Principi (Xlth century), i, 221, 326, 234

Capitals of Pointed character (Xlth or Xllth century)
ii, 122, 510, 511, 512

Crypt. Capital (Xlth or Xllth century), ii, no 404
121

- Sculptures on the fa$ade (Xllth century), ii, ui, 497,

Monasterboice

Churchyard. Cross of Muredach (Xllth century) ii 2U
701

Large cross (Xllth century), ii, 256, 702
Round tower (Xth or Xllth century), ii, 261

Monkwearmouth

Church of St. Peter (Vllih and Xlth centuries), ii 135-
138, 524, 525, 526, 144, I4, 153, 162, 177, 275

- Collection of baluster shafts and carving, ii, 148, 153

Alonreale
Cathedral (Xllth century), ii, 253, 254

Montalino (near Stradella)

Church (Xth century), i, 176-178, 243, 244, 224, 333,
228

Morttefiascone

Sant' Andrea (Xlth century), i, 217, 236; ii, 37, 252
295

San Flaviano (Xlth century), i, 210-222, 300, 301 302

303, 34, 35. 306, 309, 310, IJ5, 202, 222, 223, 22 4 ,'

226, 228, 234, 237; ii, 37, 62, 75, 78, 79, 109, 115,
192, 239, 3 u, 328

Montmajour (near Aries')
Chapel of the Holy Cross (Xlth century), ii, 34

Mont Saint Michel

Abbey church (Xlth and Xllth centuries), ii 70-74
438, 439, 44, 44i, 442, 443, 444, 75, 77, 84, 320

Moritzberg (near Hildesheiin)
Church (Xlth century), ii, 342, 343

Morsott (Algeria)
Great basilica, i, 49

Mt^he'ir (Babylonia)
Temple, i, 247

Murano
SS. Maria e Donato (Xllth century), i, 241, 354

Mitsmieh (Syria)
Church (Ilnd and Illrd centuries), i, 21, 24, 22

Nantes
Cathedral (Vlth century), ii, 60

Naples

Cathedral. Baptistery of San Giovanni in Fonte or
Great Font (Vth century), i, 192, 273, 12, 193, 194
195, 196

- Santa Restituta (Constantinian cathedral) (IVth
century), i, 192

- Stephania (Vlth century), i, 192

- Baptistery of Vincentius or Lesser Font (Vlth
century), i, 192

San Giorgio Maggiore (Basilica Severiana). Apse
(IVth century), i, 10, 4, 15, c;i, 57 ; ii, 29

San Giovanni Maggiore (Basilica Vincentiana). Apse
(Vlth century), i, 51 ; ii, 29, 290

Santa Maria Maggiore. Campanile (Xlth century), i, 50

Nevers
Saint' Etienne (Xlth century), ii, 118, 109, 119, 339

Newcastle upon Tyne

Black-Gate Museum. Gravestones from Hartlepool
(pre-Conquest), ii, 145

New Grange
Tumulus, ii, 264

Niuies

Basilica erected by Hadrian in honour of Plotina (122),
ii, IOI

Nocera dei Pagani

Santa Maria Maggiore (ancient baptistery) (IVth or Vth
century), i, ii, 8, 9

Nona

Churches of San Nicolo and Santa Croce (Xlth century),
i, 198

Northampton
Church of St. Peter (Xllth century), ii, 248

IND1 \

359

Norton
Church of St. Mary (Xlth century), ii, 177, 587, 588. iSi

Norwich

Cathedral (Xlth and Xlllh centuries), ii, 220-224, 657
658, 659, 206, 248, 690, 254

Novaru
Baptistery (Xth century), i, 179, 240; ii, 37

Nurtieh (Syria)
Chapel (IVth century), i, 9

Omn-es-Zeittin (Syria)
Chapel (Illrd century), i, 35, 51

Orleans

Crypt of Saint Aignan (IXlh and Xlth centuries), ii, 54,
, 420, 5, 59
Crypt of Saint Avit (IXth century), ii, 59, 425, 54

Orleansville (Algeria)
Basilica of St. Reparatus (IVth and Vth centuries), ii, 152

Otranto
Cathedral. Capitals in the crypt (Xllth century), ii, 251

Ovingham
Tower of the church (Xlth century), ii, 136

Oxford

Castle (Xlth century), ii, 184

Church of Si. George in the Castle (Xlth century),

ii, 184
Cathedral (Vlllth, Xlth, and Xllth centuries), ii, 162,

567, 229-232, 670, 671, 184
St. Michael. Tower (Xlth century), ii, 184

Padua
Santa Sofia (Xllh and Xllth centuries), i, 151, 152

Palermo
Cathedral (Xlllh century), ii, 254, 700

Paleslrina
Roman tomb, i, 81, 126

Palmyra

Colonnades (about Illrd century), i, 49, 75
Tomb, ii, 24, 391

Parenzo

Baptistery (Vlth century), i, 88
Basilica of the IVth century, i, 130

Cathedral (Vlth century), i, 87-90, 133, 134, 135, 136,
44, 46, 56

Paris

Saint Germain dcs Pres( Vlth and Xth or Xlth centuries),

i, 187, 265 ; ii, 59
Abbey church of Saint Denis (Xllth century), ii, 124

Parma

Cathedral (Xlth and Xlllh centuries), i, 196, 241, 353,

243 ; ii, 294, 744
Lions at the principal entrance (XHIth century),

i, 218

Panlinzelle
Abbey church (Xlth and Xllth centuries), ii, 346

99,

Pama

"I.I Cathedral (Sam.. SU-I..M., and Santa Maria del

Populo) (Xllth century), i, 244
Church of

S.inl' Kusebio. Crypt (Xlth or Xllth century), i, 114
1 15, 156

t ..|.naU in ih,. Crypt (Vlth or VHth century),

'. U.S. I57.IS8. 139

San Giovanni in Borap (Xllth century), i, 244
Santa Maria delle Caccic (Vlllih century), i,
. 146- ."3
San Michele Maggiore (Xllth century), i, 244-247

35<5, 177, 196. 228. 234, 345, 346, 241, 245;

, 303, 33. 33'. 34
the Mother of God (Vllth century), i, 137
San I'ietro in Ciel d'Oro(Xlhh century), i, 135, 177,

2 34, 347- 244. 245. 335
San Teodoro (Xllth century), i, 244
Museum. Tomb of Theodota (VMIth century), i, 100,

147, 102, 103
" Tower of Boethius," ii, 224

Peln shim
Monument creeled lo Pompey by Hadrian (130), ii, 101

Pirigueux

Sainl Fronl (Xth, Xlth, and Xlllh ccnturic-s), ii, 61
"3, "4

Pershore
Abbey church (Xlth century), ii, 226

Perugia

Arco di Augusto, ii, 173
Church of

Sanl' Angelo (Vllh century), i, 12, 12; ii, 150, 550

San Prospcro, i, 138

the University. Cilx>rium from ihe Church of San

Prospcro (VHIth cenlury), i, 138
Museum. Tciracotta urn (Und or Illrd century B.C.),

ii, 200

Porta Marzia, ii, 173, 581

Tomb of Ihe Volumnii (Illrd cenlury li.c.), ii, 37
Cinerary urn (Isl cenlury B.C. or Isl century A.D.),

ii, 199, 626

Peterborough

Cathedral (Vllih, Xth, Xllth, and XHIth centuries),
i'. '33. '34. 232-235, 672, 673, 674, 675. 228

- Fragments of tombstones (licfore the Xlllh cenlury),
'i, '48, 547

Piacenza
Calhedral (Xlllh century), i, 241, 243, 207, 636

Pisa
Calhedral (Xllh, Xllth, and Xlllth centuries), i, 242,

243. 355
San Paolo a Ripa d'Arno (Xllth century), i, 243

Poitiers

Baptistery of St. Jean (Vllth, Xllh, and XIIIili
centuries), ii, 52-54, 418, 419, 50, 58,60, 62, 137, 288

- Tombstones, ii, 148, 149

Saint Hilairc (Xlth and Xllth centuries), ii, 114, 501,
109

Pola

Calhedral. Marble tympanum (IXth century), i, 174,240
Santa Maria Formosa (Vllh cenlury), i, 85

3 6

INDEX

Pompeii

Arches, i, 94

House of Cornelius Rufus. Table support, i, 17

Thermae Stabianae, i, 20

Pomposa

Church of Santa Maria (Vlth century), i, 90, 137, 46,
56, 88

Campani'c (Xllh century), i, 53, 91

Narthex (Xlth century), i, 91

Porto (Tiber ino)

Xenodochium of Pammachius. Basilica (IVth century),
i, 21, 25, 95, 125

Pozzuoli
Octagonal Roman building, ii, 270

Praia (near Avellino)

Church of the SS. Annunziata (Vllth century), ii, 289,
290, 735

Pre Saint Didier ( Val d'Aosta)
Campanile of the church (Xlth century), i, 189

Quedlinburg

St. Servatius (Castle Church) (Xth, Xlth, and Xllth
centuries), ii, 290-295, 736, 737, 738, 739, 740, 741,
50, 164, 170, 186, 298, 300

St. Wipertus. Crypt (Xlh century), ii, 288, 734; i, 184;
ii, 170, 291, 292

Ramsey

Abbey church (Xth century), ii, 136, 158, 159, 168, 177,
187

Ravenna

Baptistery of Neon (San Giovanni in Fonte) (Vth

century), i, 37-39, 55, 56, 8, 34, 36, 67, 69, 74;

. 253
Basilica Ursiana (Cathedral) (IVth century), i, 8-21, I,

3, 21, 39, 52, 59, 84, 95 ; ii, 13, 29, 125
Cathedral. Ambon of Archbishop Agnellus (Vlth

century), i, 43, 61
Campanile, i, 52
Chapel of San Pier Crisologo (Vth century), i, 36, 8 ;

. 253
Church of

Sant' Agata (Vth century), i, 22, 27, 8, 39

Campanile, 52

Sant' Andrea (Vlth century), i, 90
Sant' Apollinare Nuovo (Vlth century), i, 40-45, 58
60, 85, 97

Campanile (IXth century), i, 44, 45, 66, 67, 68,
52, 53

Santa Croce (Vth century), i, 29, 58, 91 ; ii, 27, 68,

130
San Francesco (Vth century), i, 39, 8, 36

Campanile (Xlth century), i, 39, 57, 52, 53 ;
ii, 223

Crypt (Xlth century), i, 39

Sarcophagus (IVth century), i, 106, 155
San Giovanni Battista (Vth century), i, 65

Campanile, i, 52

San Giovanni Evangelista (Vth century), i, 21-27, 23
34. 8 . 85, 13?; ii, 58, 62

Campanile (Xth century), i, 52, 53
SS. Giovanni e Paolo. Campanile, i, 52
Santa Maria Maggiore. Campanile, i, 52

Spirito Santo (Vth or Vlth century). Capitals i 61

PL 136
Santo .Stefano in Olivis (Vlth century), i, 85

San Vitale (Vlth century), i, 56-72, 81, 82. 83, 8, 854,
86, 87, 95, 46, 52, 73, 74, 79, 82, 83, 84, 87, SS,
153, 154, 194, 209, 226, 336; ii, 27, 39, 261, 269,
270, 273, 275, 277

San Vittore (Vlth century), i, 90, 56, 91, 92, 117, 176

San Zaccaria. Epitaph (Vlth century), i, 65
Mausoleum of

Galla Placidia (Vth century), i, 28-31, 37, 38, 39, 40,
41, 7, 57, 71, 119, 194 ; ii, 27, 55', 68, 130

Theodoric (Vlth century), i, 53-55, 78, 121
Palace of Theodoric (Vlth century), i, 40

Guard-house (VHIth century), 96, 143, 40, 186,
221, 275

Piazza Vitlorio Emanuele II. Capitals (Vlth century),
'. 43

Reculver
Church (Vllth century), ii, 151

Repton

Church of St. Wistan (Xth or Xlth century), ii, 168-171,
574, 576, 130, 179, 181, 184, 291

Rheims

Saint Remy (Vlllth, IXth, Xlth, and Xllth centuries),
ii, 115-118, 502, 503, 504, 505, 107, 158, 307, 315,
323

Ripon

Cathedral (Vllth and Xllth centuries), ii, 150, 170

Rivolta d'Ad(fa

Santa Maria e San Sigismondo (Xlth century), i, 223-227,
33. 33L 33 2 . 335. 8', 214, 228, 235, 236, 237, 244;
n, 37, u8, 121, 122, 123, 240, 310, 322, 331.

Rochester

Castle (Xlth century), ii, 207

Cathedral (Vllth, Xlth, Xllth, and XHIth centuries),
ii, 132, 210, 644, 196, 246, 248, 249, 250

Rome

Amphitheatre, Flavian (1st century). Composite capitals,

i, 141, 196
Baptistery of St. John Lateran (Vth century), i, 20, 67;

ii, 149

Bronze door (Xllth century), i, 210
Basilica

Aemilia (Ilnd century B.C.), ii, 30

Julia (1st century B.C. and Illrd and IVth centuries
A.D.), i, 74, 113, 114, 228; ii, 22, 30

of Junius Bassus (IVth century), i, 36, 54

of Neptune (117-138), ii, IOI

Nova, or of Constantine (IVth century), i, 66, 102, 74,
78, 122, 80 ; ii, 21, 22, 25, 270

Ulpia (Ilnd century), ii, 152
Bas-reliefs, fragments of, i, 23, 28, 29
Baths

of Agrippa (120-124), i> 4^> 66, IOO, So; ii, 101

Caracalla (Illrd century), i, 33, 50, 34, 82

Diocletian (IVth century), i, 75-80, 116, 117, 118,
119, 121, 123, 87, 201 ; ii, 103

Nero (Illrd century), i, 46, 66, IOI, So

Titus (1st century), i, 46, 125, 167

Halls and rooms in, i, 66, 99, 67, 103; ii, 41, 404,

224

Bridge of Hadrian (Pons Aelius) (136), ii, 101
Cemetery by Sant' Agnese outside the walls (Vllth

century). Buttresses (Vllth century), i, 247, 360
Church of

Sant' Agnese outside the walls (Vllth century), i, 183 ;

ii. 25, 30, 140, 532, 141

Capitals (ist century), i, 142, 197 : (Vth or Vlth
century), 142

INDKX

SS. Apostoli. I'lulei in the portico (VMItti century),

i, 1 37
Santa Cecilia in Traslevcrc. Mos;iics (IXth century),

ii, 58

Apse (IXth century), ii, 140

S:m Cesario nl I'.ihuino (supplied) (Vth century),

ii, 129, 519, 520
Sun Clcmunte (Xllth century), ii, 141, 228

Altar of the time of Hormisdas (Vlth century),
i, 44. 65

Mosaic of apse (Xllth century), i, 10

Plu'ei of choir enclosure (Vlth century), i, 44 ;
ii, 143. 54'

San Cosimato. Cloister. Capitals (VIII century),

i, 142
SS. Cosma e Damiano. Apse (Vlth century), ii, 29,

290
Santa Francesca Komana. Mosaic of apse (Xllth

century), i, 10

San Giorgio in Velabro. Apse (IXth century), ii, 140
San Ciiovanni in Laterano. Ambulatory of the apse

(Vth and IXth centuries), i, 183, 184, 254 ; ii, 289

Bell-towers of northern fa9ade (Xth and Xllth
centuries), i, 49, 209. 299 ; ii, 104

Cloister Well-head (Xth century), ii, 127, 516
San Giovanni a Porta Latina. Well-head in Atrium

(Xth century), ii, 127

SS. Giovanni e Paolo. Round windows (Xllth
century), ii, 141, 228
- Apse (Xllth century), i, 241

Lions at the entrance (Xllth or Xlllth century),
i, 218, 319

San Lorenzo in Agro Verano (Vlth century), i, 183 ;
ii, 30, 141, 534

Capitals (Vlth century), i, 61

Lions at the entrance (Xlllth century), i, 218, 318

- Portico. Capitals (XHIlh century), i, 135; ii, 70
Santa Maria ad Martyres (Pantheon), ii, 27

Santa Maria Antica. Archivolt of ciborium (VHIth

century), i, 147, 205

Santa Maria in Cappella (Xlth century), ii, 183, 601
Santa Maria in Cosmedin (Vlllth and Xllth centuries),

ii, 84, 141

Capitals in crypt (Xlth century), i, 140, 195, 141,
143, 206

Capital (V'llh century), i, 44

Column in the Campanile (Vlllth century), i, 129,

176, 143

Plutei (Vlllth century), i, 128, 172

Santa Maria in Domnica. Capitals (Xllh or Xlltli
century), i, 141

Mosaics (IXth century), ii, 58

Santa Marii Maggiore (I Vth and Vth centuries), i, 51 ;

ii, 29
San Martino ai Monti (Vth and Vlth centuries), i, 61 ;

ii, 141
Santa Petronilla (suburban) (IVth century), i, 129 :

ii, 289
San Pietio in Vaticano (IVth century), i, 44, 220, 325 ;

ii, 21, 69, 79, 453, 125, 127, 128, 130, 170, 275,

280, 307, 759

Atrium, towers of the (Vlllth century), i, 49;

ii. '57

Crypt ("GrotteVaticane"). Pluleus (IVth century),

i, 43, 64

Mausoleum of the Anicii (IVlh century), ii, 211

Mausoleum, Imperial (Sam' Andrea or Sanli
Maria della Febre, Sant' Angelo, and Santa Petro-
nilla) (Vth century), i, 82, 128, 129, 130 ; ii, 22,

IS2. 2 75
Santa Prassede (IXth century), ii, 84, 140

- Chapel of St. Zeno. Capital (Xlth century), i, 206,

293

Mosaics (IXth century), ii. 58

Santa Pudenziana. Campanile (Xllth or Xlllth

century), ii, '83, 6oo
SS. Quattro Coronati (Vllth and Xllth centurio),

, 140, 533; i> l8 3: ii, 3

Latina (Vth century), i, 12;

S.uiia S.ihin.i (Vth century), i, 51, i ;

(Vth century), i', SL 77 ; ii, 309

Plutei (Vlllth and IXth centuries), i, 128, 173.

174- '29
San Scbastiano outside the walls. Apse (IVlh

century), i, 15, 57 ; ii, 29, 107
SS Simplicius, Kaustinus, and Viatrix (IVlh century),

ii, 290
Santa Sinforo.ti (IVth or Vth century), i, 21, 95, 231 ;

ii, 125, 129, 130

Cella trichora (Illrd century), ii, 316
SS. Sisto e Cecilia (Illrd century), ii, 316
Santa Sotere (Illrd century), ii, 316

Santo Stefano Kotondo (IVlh and Vth centuries), i, iz,

10, 26 ; ii, 26, 149, 549
Santo Stefano on the Via I.

11, 26, 129

San Valentino on the Via Flaminia (IVth century),

i, 130

Sant' Urbano alia Caftarella, ii, 224
Circular buildings, ii, 23, 26, 393, 394, 224

- " Tempio di Siepc" (Ilnd century), i, 82 ; ii, 40, 401,
102, 224

- "Tempio di Venere Sallustiana," ii, 41
Circus of Maxentius (IVth century), ii, 284
Column of Trajan (Ilnd century), ii, 309
"Crypta Balbi " (1st century B.C.), i, 23
Curia of Diocletian (IVlh century), i, 50, 120

Forum

of Augustus (1st century B.C.), i, 119 ; ii, 100
Romanum. Capitals (Vlllth or IXlh century), i, 43

German Institute. Lions from Vulci (Vlth century B c. ),
i, 218, 321, 322

" Heroon " of Romulus son of Maxentius (IVth century),

i. 25, 33

House of Nicola Crescenzio (of Cola di Rienzo, of Pilate)
(Xllth century). Consoles (Illrd or IVth century),
i, 121, 164

Mausoleum of

Santa Costanza (IVth century), i, II, 7, 15, 38, 204;

ii, 24, 25. 392. 29, 283
Hadrian (136), ii, IOI
St. Helena (IVth century), i, 20, 22; ii, 22, 387, 24,

Romulus son of Maxentius, on the Via Appia Antica
(IVth century), ii, 283

Museum

Latcran. CapilaU (Vlllth and IXth centuries), i, 143,
198

Christian sarcophagi (IVth and Vth centuries),
ii, 62, 428, 1 10, 288

- Mosaic (Illrd century), i. 105, 153
Vatican. Galleria Lapidaria. Shrine from Todi,

i, 105, 154

Nymphaeum of the Licinian Gardens (" Minerva
Medica") (Illrd century), i, 67, 105, 81 ; ii, 22, 275

Palatine

Basilica of Domitian's Palace (1st century), i, 22, 26
" Domus Augustana" (1st century), i, 30, 43, 44, 45,

46, 33, '94 ; ii. '2
Palace of Caligula (1st century), i, 2O

of Septimius Severus (Illrd century), ii, 103
Pulvinar, Imperial, ii, 103

Stadium (so-called). Ambulatory round exedra (Ilnd

or Illrd century), i, 184

Pantheon (120-124), ii, Io , IOI > 486, 4^7. 2 7
Portico adjoining the Theatre of Marcellus, i, 75, 115
Praetorian Camp (23), i, 32
Saepta Julia (117-138), ii, 101
Secretarium Senatus (Illrd or IVth century), i, 47, 69

Temple of

Ceres and Faustina (Sant' Urbano) (Ilnd century),
ii, 224

Trajan (117-138), ii, IOI

Venus and Rome (121-135), Io - lo2
Theatre of Balbus (1st century B.C.), i, II

362

INDEX

Tomb
of Annia Rcgilla in the Valle Caffarella (Ilnd century),

',31

Eurysaces (1st century H.C.), ii, 222,660
on Via Appia, i, 225, 334 ; ii, 130, 521, 149

Via Labicana, i, 94

Via Nomentana near the " Casile dei Pazzi" (Ilnd

century), i, 31, 47, 48

" Sedia del Diavolo " (Ilnd century), i, 33, 49

Via Praenestina (Ilnd century), 36, 55

Via Salaria, i, 93, 247, 359

or funerary edifices, i, 15, 13, 55, 79, 80, 67, 104, 187,

263, 264, 247, 358 ; ii, 152, 554
Villa

" Centroni " (Illrd century), i, 24, 30

of the Gordians (Via Praenestina). Mausoleum (" Tor

de' Schiavi") (Illrd century), i, 20, 33, 82 ; ii, 41,

224, 283
"SetteBassi" (Via Latina) (Ilnd century), i, 21, 94,

140 141, 175, 242, 248, 361, 362. 363 ; ii, 58, 103
near " Fontana Piscaro " on the Via Latina (Ilnd

century), i, 24
Mattei. Sarcophagus (Illrd or IVth century), ii, 59,

424
Walls (Illrd-XIXth centuries), 1/165

of Aurelian (Illrd century), ii, 69, 436

Ramsey

Abbey church (Xllth century), ii, 193

Crucifix (Xllth century), ii, 193, 618

Rouen

Abbey church of Saint Ouen (Xlth and Xlltb centuries),
ii, 47, 409

Ruthwell
Cross (Xllth century), ii, 143, 539

St. Albans

Cathedral (Xlth century), ii, 200-202, 627, 628, 629, 96,

98, 185, 186. 206, 211, 219, 231
St. Michael's Church (Xth century), ii, 160, 561, 562,

162, 170, 201, 299

Saint Benolt sur Loire
Abbey church (Xlth and Xllth centuries), ii, 33

St. Gall

Abbey church (IXth century), ii, 285, 61, 270, 298, 299

Saint Ghieroux
Church (Xth or Xlth century), ii, 63

Saint Giiilhem du Desert (Gellone)
Abbey church (IXth, Xlth, and Xllth centuries), ii,
35-37, 397, 398, 339

Chapel (IXth century), i, 167 ; ii, 36

St. Macdara's Island
Chapel (Xth century), ii, 265

Saint Remy

Tomb of Gallo-Roman period, ii, 252

Saint Riquier (Centula)

Abbey church (Vlllth century), ii, 60, 426; i, 178;

ii, 36, 50, 63, 107, 158, 307, 312,323
Church of St. Benedict (Vlllth century), ii, 426
Church of St. Mary and the Apostles (Vlllth century),

ii, 60. 426

Saint Savin

Abbey church (Xlth. century), ii, 115

Salonica

Church of

the Apostles (Xlth century), i, 27, 155, 216, 197, 217,

3*5

St. Demetrius (Vth century), i, 16-18, 16, 17, 18, 19,

10, 15, 25, 32, 43, 44, 60, 63, 126, 154 ; ii, 308
St. Elias (Xlth century), i, 155, 156, 197
Eski Djuma (Vth century), i, 15-18, 15, 10, 15, 25,

62, 97, 126, 154

St. George (Vth century), i, 13-15, II, 10, 25, 154
Pluteus (IXth or Xth century), i, 149, 150, 208
St. Panteleemon (Xlth or Xllth century), i, 197, 278,

140
St. Sophia (Vth century), i, 63, 9, 2, 17, 2O, 18, 34,

43, 62, 46, 60, 62, 92, 64, 94, 71, no, 154, 238
the Virgin (Xlth century), i, 249, 366, 27, 197, 217,

317

Sanf Ambrogio ( Valle di Susa)

Campanile of the church (Xth century), i, 186

San Benigno

Abbey church of Fruttuana (Xlth century), i, 185, 257,
258, 259 ; ii, 5, 32

San Giorgio in Valpolicella

Church (Vlllth century), i, 113

Archivolts and capitals of the ciborium (Vlllth
century), i, 138, 190, 139, 146, 202, 203

San Leo

Church (IXth century), i, 171-174, 234, 235, 236, 238.
27, 117, 176; ii, 294, 299

Santa Maria Capua Vetere

Roman Mausoleum, i, 55, 80

Tomb known as " La Conocchia," ii, 252, 696

San Michele della Chiusa
Abbey church (Xllth century), ii, 74, 445

San Miniato al Monte (near Florence)

Church (Xlth century), i, 202-206, 287, 288, 289, 290,
1 8 1, 190; ii, 84, 99, 305, 308, 320

Sarvistan (Persia}
Palace, i, 193, 194, 195

Scartho
Church tower (Xlth century), ii, 186

Scattery Is/and
Round tower (Xlth or Xllth century), ii, 261

Schaffhausen
Cathedral (Xlth century), ii, 340, 321

Schwarzach
Abbey church (Xllth or Xlllth century), ii, 340

Seligenstadt
Abbey church (IXth century), ii, 280, 278

Sens
Cathedral (Xllth century), ii, 37, 124

Siah (Syria)
Temple of Baalsamin (1st century B.C.), ii, 66

INDKX

363

Sikhester

li.isilica

Civil (Romano- British period), ii, 152
Christian (IVtli or Vlh century), ii, 125, 514

Mosaics of the Romano-British period, ii, 66, 431, 145

Skellig Michael

( >rut(iries and monastic cells (from the IXth century
onwards), ii, 264, 265

Sampling

Church. Tower (XMlh century), ii, 189-191,610,182,

323
-7 Carving (aljout the Xlllth century), ii, 190

Southwell

Cathedral (Xllth century), ii 226-229, 665, 666, 667.
668, 669, 237, 246, 248

Spalato

Baptistery (IVth century), i, 120, 163
Palace of Diocletian (IVth century), i, 94, 97, 144, 120,
158, 181

Speyer

Cathedral (Xlth and Xllth centuries), ii, 316-323, 767
768, 769, 770 ; i, 244 ; ii, 294, 303, 313, 327, 330,
332. 338. 339, 344

Chapel of St. Afra (Xllth century), ii, 319, 321

Chapel of St. Emmerammus (Baptistery) (Xlth
century), ii, 319, 772

Spoleto

Church of San Salvatore (del Crocifisso) (I Vlh century),
ii, 27-29, 395, 396; i, 22, 183 ; ii, 30, 125, 131, 132
Cathedral. Doorway (Xllth century), ii, 29

Steinlxuh (near Michelstadt)

Church (IXth century), ii, 278, 724, 725, 63, 228, 288,
294

Stmv

Church of St. Mary (Xllh and Xllth centuries), ii, 163-
165, S69, 57- 178, 179, 181, 190, 193

Stnusburg

Cathedral (Xllth and Xlllth centuries), ii, 310

Sueida (Syria)

Church (Vth century), ii, 13
Tomb of Hamrath (1st century B.c

ii, 24

Susa

San Giusto (Xlth century), i, 186, 260, 261, 209
Santa Maria (Xlth century), i, 209

Town walls. "Opus spicatum " (IV'th-VIlh centuries),
i, 165, 223

Tabarka (Tunis)
Baths (Illrd century), i, 21

Tafkha (Syria)
Church (IVth or Vth century), i, 46 ; ii, 13, 30

Taraston
Chapel of St. Gabriel (Xllth century), ii, 34

Ttrni

Tombs of the Taciti, i, 24

Teu-kesbury

Abbey church (VIMth and Xllth centuries), ii, 224-3J6,
661, 662, 663, 248

Tivoli

Villa of Hadrian (125-135), ii, 100-102, 485; i, 23 77
120, 81, 127, I 5 6, I 9 2. 271, 272, .93, 274; ii, 40,

402, 41, 169,575, 276, 289

Torcello
Santa Fosca. Apse (Xlth or Xllth century), i, 242

Toscantlla

Alil>ey of San Giusto, i, 122

San Pietro (Vlllth, Xlth, and Xlllh centuries)
i, 121-136, 165, 166, 177, 178, 179, 180, 181, 113,
"7, 137, 156, 171, 238; ii, 70, 78, 253, 299

Carving (plutei, &c.) (VIMth and IXth centuries),
i, 126 129, 169, 170, 171, 175

Santa Maria Maggiore (Xlllth century), i, 135

Toulouse
Saint Semin (Xlth century), ii, 118

Tournus

Abbey church of Saint Philibcrt (Xth, Xlth, and Xlllh
centuries), ii, 112-114, 499, 500, 109, 207, 215, 327

Tours

Church of Saint Martin (Vlh, IXth, Xth, Xllh, Xlllh,
and Xlllth centuries), ii, 107, 108, 491, 493; i, 49,
184 ; ii, 59, 253

"Tour Charlemagne" (Xlth, Xlllh, and Xllllh or
XlVth centuries), ii, 108, 492

Church of SS. Peter and Paul (Vlh century), ii, 59

Trail

Calhedral (Xlllth century), i, 159, 22O, 160
Church of Santa Barbara (Xlllh century), i, 198

Trier

Basilica (Illrd or IVth century), ii, 286, 731

Cathedral (IVlh, Vllh, Xllh, and Xlllh centuries),

ii, 267-269, 714 716, 324, 773, 331, 338
Imperial Palace (I\ th century), ii, 268, 715, 276
Porta Nigra (IVth century), ii, 284, 728 ; i, 60

Trieste

Cathedral (Vlth, Xth, and XI Vth centuries), i, 208, 298 ;
ii, 77, 450

Tuam

Churches (Xllth century), ii, 256

Cross of O'Hoisin (Xlllh century), ii, 256, 703

Tunis

Bardo Museum. Mosaic from Roman Villa at El-Alia
(Ilnd century), i, 49, 76

Turin

Palazzo Madama (XVlli century), i, 188
Porta Palalina (1st century H.c. or 1st century A.D.),
ii, 284, 729

Tiirmanin (Syria)
Church (Vlth century), i, 46

Uses
Saint Theodoril. Tower (Xllth century), ii, 36

Vaison

Cathedral (Xth and Xllth centuries), ii, 38
Saint Quenin (Xllth century), ii, 37

364

INDEX

Venice

San Marco (Xlth century), i, 207. 297, 241, 242;

ii, 114
Sant' Eufemia alia Giudecca. Capitals (Xlh century),

Sun Giovanni Decollate. Capitals (Xlth century), ; Churc t h _f ? e . I , Io '/._ Cros ;f and St ' Laurence (Xllth

i, 209

Venosa
Abbey church of the Trinila (Xlth and Xllth centuries

>i, 253
Capital (Xlth century), i, 223, 329

Verona

Volterra
Porta dell' Arco, ii, 173. 580

Waltham Abbey

:h of the Holy Cru
century), ii, 216, 651, 228

Winchester
Cathedral (Vllth, Xth, and Xlth centuries), ii, 205-207,

632, 633, 634, 635, 157, 159, 210, 215, 231, 315

- Font (Xllth century), ii, 246, 688

Wing
Church (Xlth century), ii, 189, 608, 609

Wittering

Church of

San Giovanni in Fonte. Capitals (VHIth century),

i, 136, 186, 140, 182
Santa Maria Matricolare (Old Cathedral). Capitals | Church (Xlth centurvl ii I7Q

(VHIth century), i, 136, 140, 182
Santo Stefano. Apse (Xth century), i, 182, 250, 251 ;

. 3. 79

Santa Teuteria (VHIth and Xllth centuries), i, 113

San Zeno Maggiore (Xllth century), ii, 121
Museo Archeologico. Pre-Lombardic cubical figure

capital, i, 138
Columns of ciboriuin from San Giorgio in Valpoli-

cella (VHIth century), i, 138
Forta dei Borsari (Illrd century), ii, 284

Vetulonia
Tomb called " La Pietrera," i, 30

Vezelay

Abbey church of La Madeleine (Xlth and Xllth
centuries), ii, 119-123, 506, 507, 508, 37, 253, 303,
331

Vicenza

SS. Felice e Fortunato (Xth century), i, 180-182, 248,
249, 176, 178, 202, 205, 221 ; ii, 84, 99, 251, 308

Villanova Veronese
San Pietro. Pluteus (VHIth century), i, 143, 199

Viterbo

San Giovanni in Zoccoli (Xlth century), i, 135, 183, 171
Santa Maria della Cella. Campanile (IXth century),

i, 169-171, 233, 50
San Sisto Vecchio (Xllth century), i, 135, 182

Worcester

Cathedral (Vllth, Xth, Xlth, and Xllth centuries),

ii, 214, 648, 649, 246, 689
Slype (Xlth century), ii, 215, 649

Worms

Cathedral (Xllth and XHIth centuries), ii, 335-339,

789, 790, 113, 303, 323, 325
Chapel of St. Maurice (Xlth century), ii, 336

Worth
Church (Xlth century), ii, 187, 604, 605, 179, 182

York

Cathedral (Vlllh, VHIlh, and Xllth centuries), ii, 132,

157, 244, 245, 248
St. Mary Bishophill Junior. Tower (Xlth century),

ii, 136

Ythancaestir (Bradwelt)
St. Peter on the Wall (Vllth century), ii, 133

Zara

Cathedral (XIHth century), i, 159
San Donato (IXth century), i, 152-157, 210, 211
- Museum. Carvings, i, 157
San Grisogono (Xllth century), i, 159, 219
San Lorenzo (Xlth century), i, 198
San Pietro Vecchio, i, 198

GENERAL INDEX

Adam magister, probable architect of Sant' Ambrogio,
Milan, i, 235

Aisles prolonged into choir and transept, ii, 69, 206, 315

Aloiosus, architect of Theodoric, i, 53

Altar frontals, carved, i, 103, 126 ; ii, 344

Amlxins, i, 43

Animals, realistic or imaginary

holding object IxHwecn their paws, i, 218 ; ii, 256

introduced in carving, i, 105, 146, 149, 202, 215-217 ;

ii, 7, 62, 79, 87, 89, 105, 121, 200, 246, 248, 293,

325, 339

pairs of, with single head, ii, 1 10
rampant, on shafts and capitals, i, 225, 236
stylophorous, i, 218-220

Anthemius of Tralles, architect of Justinian, i, 79, 62,
69, 71, 81

Apollodorus of Damascus, architect of Trajan, ii, loo,
102, 103

Apse

Ambulatory, arcaded or not, round apse or exedra,

i, 10, 57, 69, 183-185 ; ii, 29, 107, 285, 289, 290
Arcading round, closed by transennae, i, 26
Blank arcading round, i, 154-156, 25, 190
Buttresses round, i, 15, 175 ; ii, 55

Chapel projecting from (" Lady Chapel"), ii, 9, 10,
196

facing another apse at opposite end of the church,
ii, 151, 285, 298

frontal arch of, enclosed by a triplet of arches, ii, 54, 62

decorated with shafts, ii, 41, 58, 62
niche in centre of, for bishop's seat, ii, 289
orientation of, i, 8 ; ii, 13, 29, 125, 126, 127, 132, 140
polygonal externally, i, 8 ; ii, 13, 189

range of arched niches round, i, 131, 164, 167, 175,
201, 238; ii, 37

three-lol>ed apse or structure, ii, 316
Arcade of nave or choir supported by

alternate columns and piers, i, 16, 180, 203, 204 ;
ii, 84, 251

alternate larger and smaller piers or supports, i, 224 ;
ii, 84

cylindrical piers, ii, 1 1 2, 215
Arcading and arches

blank, external, i, 23-25 ; ii, 91, 94, 100, 102, 107,
170, 174, 188, 198, 221, 224, 264, 330

round apses and chancels, i, 154-156; ii, 94i

97, loo, 174, 189

internal, i, 125, 143 ; ii, 191, 206, 221

Arcading and Arches tontinned

round apse or chancel, ii, 44, 54, 58, 62, 77,

100, 324
- intersecting, ii, 204, 219, 221, 238, 253

corbel or pensile, i, 97, 158

pointed-arched, ii, 228

triangular-headed, ii, 62, 63, 154, 170, 183, 189,
288

triplet, ii, 206
Arch-
animal forms or heads ("protome") on impost,

keystone, spandrels of, ii, 37, 121, 173, 181, 192

dentils round, i, 124

horse-shoe, i, 171 ; ii, 58, 59

pointed, ii, 34, 109, 121, 218, 220, 241

polychrome voussoirs of, i, 164 ; ii, 49, 271

ring of bricks framing, i, 26, 27

screen of columns tilling, ii, 271

transverse, i, 57, 176, 180, 203, 204, 225 ; ii, 97, 99,

112

two or more arches enclosed by a single, i, 39 ; ii, 48,

95. 239. 253

Arched corbel course, continuous or broken by Icscnas or
shafts, horizontal or stepped, i, 36-38, 39, 119, 131,
156, 172, 177 ; ii, 32, 36, 91, 214, 253, 302, 312, 337

Architecture, importance of, ii, 135, 347

Augustine, archbishop of Canterbury, mission of,

ii, 125-129, 133
Axe, use of by Anglo-Saxon carvers, ii, 202, 211, 229,

248

Baluster shafts, ii, 138, 142, 153, 183, 184, 186, 188,
20 1, 202

Baptisteries, i, 37, 66, 88, 95, 101, 165, 178, 190, 192 ;

ii, S 2
Base of column

bulbous, ii, 164, 170, 206, 215, 220, 291, 292

formed by inverted capital, ii, 318

spurred, i, 158, 181, 221 ; ii, 51, too, 216, 225, 272,
306, 310, 321, 322, 340, 341, 343

Basilica

Byzantine vaulted, i, 61-84

Christian, i, 21, 51 ; ii, 13, 20, 26, 30, 69, 107, 125

Lombardic vaulted, i, 161-250; ii, 276, 322

Lombardo-Norman, ii, 64-99, 251, 253

Lombard')- Pointed, and Pointed, ii, 114, 123, 124,

241, 242

Lombardo- Rhenish, ii, 303, 323, 328
Ravennate vaulted, i, 55-84, 153, 154
Roman, i, 21, 22, 66, 74 ; ii, 30, 103, 286

366

GENERAL INDEX

Bee-hive structures in Ireland, ii, 264
Bells, liturgical use of, i, 46
Benedict Biscop, ii, 135, 138, 146
Benedictine Order, ii, 4, 46, 168, 274

Buttress-
Babylonian, i, 247
Byzantine, i, 69, 79, 82, 249

Lombardic and Pre-Lombardic, i, Si, 175, 201, 225,
227 ; ii, 115, 119, 122

Lombardo-Norman, ii, 74, 81, 82
Ravennate, i, 57, 58, 79, 83, 94, 251
Roman, i, 67, 76, 78, 83, 175, 247, 270
ramping or flying, i, 13, 15, 77, 78, 201, 248
Capital
bulbous, ii, 142, 154, 178, 215

Byzantine, i, 16-18, 41, 44, 59-61, 62, 63, 85, 87, 88,
94, 136; ii, 19-21, 251, 284

Composite, i, 141, 205, 206

crocket (" con foglie angolari a punta bifida accar-
tocciata"), ii, 8, 68, 80, 185, 199, 332

Gallo-Roman cubical funnel-shaped, i, 60 ; ii, 284

Lombardic cubico-spherical, i, 207 ; ii, 95, 142, 185,
195, 238, 294, 301, 308, 309

Lombardic, figure, i, 181, 200, 202, 215, 232 ; ii, 51,
62, 79, 105, 120, 294

Lombardo-Apulian and Sicilian, ii, 251
Lombardo-Norman, ii, 75, 79, 95
Merovingian, ii, 51

Neo-classical (Xlth-XIIIth centuries), i, 141, 205;
ii, 70

of Pointed character, ii, 121

Pre-Lombardic cubical, i, 138-140, 169, 171, 204;

", 32, 43, 44, 48, 57, 63, 153, 165, 189, 192, 284
figure, i, 138; ii, 32, 34, 64
Ravennate, i, 136, 168, 182

scalloped (Anglo-Lombardic cubico-spherical), ii, 95,
96, 98, 185, 202, 204, 217, 219, 220, 238, 309, 332

spherico-polygonal, ii, 341

" storied," ii, 228

volute, ii, 185

Carving, details of, left unfinished, i, 203, 204
Cauliculi, i, 107, 146, 157
Chancel (presbytery, sanctuary)

flanked by sacristies, i, 21 ; ii, 29, 125, 129

raised above the floor of the church, i, 85, 129, 173

rectangular, ii, 161, 176

separated from the nave by columns, ii, 128, 129, 151

Chapels radiating round apse or choir, i, 21 1, 222;
ii, 106, 107-109, 195, 217, 220

Charles the Great, i, 113, 151, 273-278, 280
Chequer pattern as decoration, ii, 83

Ciboriums, archivolts and capitals of, i, 129, 138, 140,
146, 147, 148, 157, 171

" Ciotole," i, 44

Circular vaulted buildings with annular aisle, ii, 14, 23-26
283

of more than one story, ii, 26

Coins with architectural or decorative representations
of Caracalla, i, 48

Coins, &c. continued.

Edward the Elder, ii, 159

Philip the Arabian anil Otacilin, i, 48
Roman Republican, ii, 30

of St. Martin of Tours, i, 50; ii, 108

Septivnius Severus, i, 48

Teos, i, 220

Columns, spiral, ii, 131, 170, 290

Comacine Masters and Gilds, i, 37, loo, IOI, 108-111,
123, 128, 145, 146, 157, 159, 166, 171 ; ii, 31, 145,

169, 273, 275

Corbels carved with figures, i, 119 ; ii, 237, 338
Croatia, i, 158-160
Crosses

Cornish, ii, 148

Irish, ii, 255-257

Northumbrian, ii, 143, 148, 161
Crypt, i, 85, 173, 190; ii, 170, 171

aisled, under chancel and transept, i, 185, 203 ; ii, 60,

170, 205, 289, 294

Dome (cupola) and half-dome

Byzantine, polygonal, i, 196-198, 249

in compartments ("a superficie sferica ondulata"),
i, 70, 81 ; ii, 39-41, 102

constructed of amphorae or terracotta tubes, i, 18-21,
39, 69, 84 ; ii, 22

drum of, carried up above the impost line, i, 82

encircled by columns, ii, 251
lantern of, i, 178

Lombardic, i, 191, 196, 198, 199, 247 ; ii, 338
ribbed, i, 70, 81, 82, 83
windows in, i, 82 ; ii, 224, 275
wooden, ii, 14, 15
wooden roof above, i, 59 ; ii, 275
Doorway

lunette above, i, 119, 133; ii, 262
niche above, i, 97
triangular-headed, ii, 192

Dunstan, archbishop of Canterbury, ii, 155, 156, 161,
176

Einhard, Minister of Charles the Great, ii, 276-280

Etruscan art, motives of, borrowed by later Italian
Schools, i, 105-107, 145, 216, 218-220 ; ii, 173

Fonts, ii, 127, 246, 298, 333
Galleries

external arcaded, i, 240-244 ; ii, 97, 224, 253, 322,

3 2 4, 331, 33 2 333
internal (matroneum, triforium), i, 183, 237; ii, 31,

271
Gates of Gallo-Roman cities, importance of, ii, 284

Gilds, Lombard or Comacine, i, 45, 53, 173, 178, 179,
181, 186; ii, 32, 112, 145, 245, 274, 275, 277,299,
33. 322

Gregory VII, Pope (Ilildebrand), ii, 194

Gundulf, bishop of Rochester, ii, 207, 210

Hadrian, Emperor, ii, 100-103

Hadrian I., Pope, i, 128; ii, 275

Herring-bone work (opus spicatum), i, 165, 175, 178;
ii, 77

Hezzel, architect of Cluny, ii, 104, 109

GENERAL INDEX

367

Hog-backed i;i.m^t'>iH--;, ii, 149

llmiun li^urt-, treatment nf, in carving, i, IO2, 144:

ii, 32, 62, 75, 90, 109-111, 120, 146-148, 161, Hi?,

200, 220, 244, 245, 246, 247, 248, 249, 250, 25(1.

3'. 317, 339, .544
Illuminated Manuscripts

Anglo-Saxon, ii, 143, 144-146, 154, 160, 164, 169,
>3

Knglish, ii, 197

German, ii, 285, 288, 295, 309

Irish, ii, 144-146

in the Library of St. Gall, ii, 146, 279, 292, 295
Interlacing, decorative use of, i, 105, 126, 143-150'

", 89, '43. >44-l49
- intestinal, ii, 14^, 165, 255, 256
Isidorus of Miletus, architect of Justinian, i, 79, 62
the Younger, architect, i, 69, 82
Julianus Argentarius, architect of Ravenna, i, 64, 87, 89
Keystone of vaulting, carved, ii, 37
Lady Chapel, ii, 167, 196, 211

Lanfranc of Pavia, archbishop of Canterbury, ii, 34, 84,
85, 86, 90, K)S, 194, 195, 200, 210

Lanfrancus of Modena, architect, i, 97, 221, 240; ii, 83

Lesena (pilaster strip), i, 36 ; ii, 169, 170, 177, 183, 299

Liutprand, Lombard king, i, 112, 117, 123

Long and short work, ii, 136, 162, 176, 271, 272, 275

Marmorarii (marble workers), i, 53, 135, 143; ii, 131

Melioranzi), Umbrian carver, ii, 29

Milan, Imperial residence at, and later vicissitudes, i, 4

under archiepiscopal rule, i, 5, 161, 230

Monogram

the cross, i, 10 ; ii, 131

of founder in Raveniute buildings, i, 41, 43, 44, 87
Moulding (stringcourse, cornice)

billet, ii, 63, 67, 89, 190, 191

cable, i, 177

chequer, ii, 83

double cone, ii, 224

embattled, ii, 89

recurring volute ("corridietro"), i, 107

saw-tooth, i, 25

zigzag (chevron), i, 158 ; ii, 89, 238
Nicholas, " magister," i, 221 ; ii, 121
Oratory, Irish two-storied, ii, 261-264

Orientation of churches, i, 8, 121 ; ii, 13, 125, 126, 140,
211

Pediments, decorative motive of continuous or isolated,
i, 23 ; ii, 54, 62, 288

Pendentive

Campano-Loml>ardic and Lombardic, conical or hood-
shaped ("a scuffia"), or formed as a niche, i, 71,
73, '78, 179, 191-196, 198 ; ii, 37, 82, 109, 339

compound ("a piu riprese"), i, 73, 192, 247
spherical (triangular), i, 29-35, 39, 63, 194, 198
graduated ("raccordo a risega"), i, 30, 35

Persia, Greek and Roman builders employed in, i, 25
Pier-
cylindrical, ii, 112, 205, 207, 215, 217, 227, 238

compound, ii, 206

of irregular outline, ii, 26, 69

Lomliwdic compound, 1, 75, 171, i-\. IM. i<>-,
248 ; ii, 70

T shaped, i, 90, 92, 175, 176
Plan of church

central or concentric, i, 71, 153; ii, 33, 273

Latin cross, i, 28 ; ii, 68, 129, 130, 168

octagonal, i, 66
Porch

Roman and LomUirdic, i, 97, 218 221

churches with more than one, ii, 128, 130, 150
Portal

Franco-Lombardic, ii, 121

Lombardic, i, 217, 236; ii, 37, 91, 204, 252, 294,
338

Projections, graduated bracket-like, at angles of Kaven-
nate buildings, i, 58, 71, 86

Pulvin

Ravennate, i, 10-18; ii, 39, 183, 346

corbel or crutch-shaped ("a stampella "), i, 51, 186 ;
ii, 182, 183, 190, 300

Raccord (see Pendentive), i, 192; ii, 15, 27, 80, 82, 102

Ravenna, importance of, i, 4, 7, 95

Salonica, importance of, i, 15, 62

Sarcophagi, i, 100, 101, 106, 107, 144; ii, 62, 286, 288

Saw-tooth ornament ("denti di sega"), i, 25

" Scholae peregrinorum " at Rome, i, 1 10

Screens of choir or chancel (plutei), parafiets, carved
slabs, etc., i, 43, 44, 101, 102, 126-129, 143, 149, 150,
214

Sexes, division of the,' in churches, i, 90, 117

Sicily and South Italy, characteristics of the churches of,
ii, 251-254

Star ornament, ii, 49, 191
.Suger, abbot of Saint Denis, ii, 124
Theodelinda, Lombard queen, i, 112 ; ii, 290
Theodoric, king of Italy, i, 40, 55

Towers (Campanile), for bells or stairs : form, posi-
tion, etc.

Anglo- Lombardic, ii, 128, 157-160, 172

atrium of church provided with, i, 87 ; ii, 157

central, ii, 27, 60, 83, 158, 159, 177, 187, 312

four : building set between, i, 72, 75, 76, 79 ; ii, 323

Krankish, ii, 60

German, ii, 299, 303, 312

- "helmed," ii, 190, 191, 318, 323
Irish round, ii, 260, 261

Lombardic, i, 169-171, 185-189, 209; ii, 251, 286

Lom'rardo- Roman, ii, 183

porch, ii, 128, 171, 172

Ravennate, i, 39, 44-53, 57, 73 ; ii, 270, 285

western, single, ii, 158, 159, 187, 313, 316

- pair of, ii, 27, 105, 159, 285
Transept

apsidat chapels projecting from, ii, 34, 69, 109, 312
double, ii, 107, 196, 307

gallery, loggia, or portico at ends of, ii, 79, 307
quasi-, ii, 172, 179, 187

368

GENERAL INDEX

Triforium, pseuclo-, or quasi-, ii, 109, 221, 225, 226, 231
Vaulting

annular barrel, springing from central pier, ii, 283, 284

cross, with diagonal ribs, visible or not, i, 215, 223,
226, 237, 248 ; ii, 37, 225, 228, 235, 239-242

raised, concave-crowned ("crociera cli sesto rialzato
a calotta"), i, 57, 199, 224, 226, 248

with visible arches springing from corbels, i, 97
ramping, ii, 113

Venice, artistic revival of the Xlth century at, i, 207

Wall-passage (" galleria di servizio "), i, 126; ii, 69, 78,
109

West front of church

buttresses of, indicating internal divisions, i, 90, 91,
93, n8; ii, 251

decorated with reliefs arranged unsymmetrically, ii, 245
niche or recess in, containing entrance, ii, 225

Wilfrid, bishop of Vork, ii, 135, 139, 146, 149, 150
Wiligelnuis magister, i, 221

William the Englishman, architect of Canterbury Cathe-
dral, ii, 196

William of Sens, architect of Canterbury Cathedral,
ii, 196

William of Volpiano, abbot and architect, ii, 5, 30, 45,
46, 64, 67, 70, 73, 77, 84, 231

Window or opening

cross-shaped (luminous cross), i, 54, 121, 131, 163, 167

Lombardic, i, 131

Lombardo-Apulian, ii, 252

rose, i, 135 ; ii, 34, 121, 253

round, ii, 107, 117, 222-224, 228, 279, 285

splayed on both sides, i, 93, 131 ; ii, 160, 175, 299

triangular-headed, ii, 181, 190
Wooden churches, ii, 133, 134, 135, 163, 176, 259

RICHARD CLAY & SONS, LIMITED, BREAD STREET HILL, E.C., AND BUNGAY, SUFFOLK.

NA Rivoira, Giovanni Teresio

1119 Lombardic architecture

L8R513
v.2

PLEASE DO NOT REMOVE
CARDS OR SLIPS FROM THIS POCKET

UNIVERSITY OF TORONTO LIBRARY

f. :-,-^y - HEBE

// I

KMI^^9III -f
.1'' " " 1