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

LGMBARBIC ARCHITECTURE

IT 1 :) OR j| WELOPMENT AND
DERIVATIVES - * By G. T. RIVOIRA

nf ilje

Jack Efyrie

QxP

LOMBARDIC ARCHITECTURE

VOL. I

CATHEDRAL CITIES OF ENGLAND. By GEORGE GILBERT.
Illustrated with Sixty Colour Drawings by W. W. COLLINS, R.I.
In One Volume. Demy 8vo. Price i6s. net.

CATHEDRAL CITIES OF SPAIN. Written and Illustrated

with Sixty Colour Drawings by W. W. COLLINS, R.I. In One
Volume. Demy 8vo Price l6s net.

CATHEDRAL CITIES OF FRANCE. By HERBFRT and
HKSTKR MARSHALL. With Sixty Illustrations in Colour by HERHERT
MARSHALL. In One Volume. Demy 8vo. Price i6s. net.

ART IN GREAT BRITAIN AND IRELAND. By Sir
WALTER ARMSTRONG. With many Illustrations. In One Volume.
Cr. 8vo. Price 6s. net.

ARS UNA: SPECIES MILLE. GENERAL HISTORY
OF ART.

HOW TO BUILD OR BUY A COUNTRY COTTAGE
AND FIT IT UP. By "HOME COUNTIES." With a large number
of Photographs, Plans, and Estimates and Introduction by HENRY
NORMAN, M.P. In One Volume. Cr. 8vo. Price6j.net.

LONDON : WILLIAM IIEINEMANN, 21, BEDFORD STREET, W.C.

SAN SALVATORE, BRESCIA.

[Frontispiece.

ORIGIN
DERIVATIVE

PURE

ME NT AND
t RIVOIRA

TRANSLA,

HEINEMA

LOMBARDIC ARCHITECTURE

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

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

WITH OVER EIGHT HUNDRED ILLUSTRATIONS

VOL. I

LONDON: WILLIAM HEINEMANN

MCMX

9572

PREFACE

IT was in the course of one of my numerous artistic pilgrimages in the
countries north of the Alps that there came to me like a flash the vision of
what I call the real origins of the styles of architecture which flourished
in those lands in the Xlth and Xllth centuries. Far from being intimidated
by the importance or difficulty of the subject, I forthwith determined to
devote my studies and researches to the development and completion of the
idea. I set to work without delay, making it my object to follow the path
of truth, which was my only guide, so far as it was revealed to me by those
same studies and researches. And now at last I find myself in a position
to lay before the world of students the results of a labour which can be
truly described as conscientious.

The book is divided into two parts. The first deals with the origins
of the Lombardic vaulted basilica the main stem from which were derived
the shoots whence sprang the Northern styles above referred to. The
second part is concerned with the origins of the chief derivatives of the
Lombardic basilica in the lands beyond the Alps.

The work is based on investigations which are absolutely original ;
and not less original are the conclusions to which they give rise. Some
of these conclusions, I mean those relating to the origins and modifi-
cations of the Byzantine vaulted basilica, though not immediately con-
nected with my subject, will have the effect of opening up a wider and
more rational field of research for the ecclesiastical architecture of the East.
Other paths, as yet untrodden, are pointed out to students of Western
mediaeval art.

The historical arguments, which form an essential part of my work, are
the result of long and patient study of the original sources. Further, all

viii PREFACE

the existing buildings or monuments described in this book have been, with
rare exceptions, personally investigated on the spot. These buildings and
monuments represent but a very small part of those which I have examined.
The rest have been omitted here, either because they did not appear to me
to throw any fresh light upon the subject, or because they were not directly
connected with it.

TRANSLATOR'S PREFACE

THE Author's Preface and the Introductions to the two parts of this book explain so
clearly its origin, method, and scope, that it is unnecessary for me to say more than
a few words with regard to the translation which is now presented to the English-
speaking public.

In the first place I should like to point out that it is not a mere reproduction of
"Le Origin! dell' Architettura Lombarda " (ist edition, Loescher, Rome, 1901 and
1907, 2 vols. : 2nd edition, Hoepli, Milan, 1908, I vol.). The great bulk of the material
and its treatment of course remain the same ; but the Author has taken advantage of
the translation to revise the whole work, with the result that considerable improve-
ments harve been made in the form of correction, amplification, re-statement ; not to
speak of important additions to the subject matter, among which we may call especial
attention to the account of the so-called Temple of the Clitumnus, and the Excursus
on Hadrian as an architect.

In the next place I think it is due to the Author to explain his relation to this
version. Throughout it has had the advantage of his personal supervision in a very
exceptional manner. Signer Rivoira's knowledge both of the English language and
of English architectural and archaeological terminology is such that he has been able
to exercise a real control over every word that I have written. Constantly when
difficulties have arisen (and they have not been few) as to the rendering of passages
or phrases connected, for instance, with such abstruse and technical matters as vault
and dome construction, it is he who has provided the solution. I think it will be
admitted that it is an inestimable advantage thus to get the Author's own version of
his statements, so that the originality and individuality of his presentation may be
conveyed direct to his readers.

With regard to my own part in the translation, it has been one of my main
objects to preserve this individuality; and I would ask those who may feel aggrieved
at new or unusual forms of description or statement, to remember that a large part of
the value to us of works which come from other countries and other intellectual
atmospheres consists in the freshness and novelty of the presentation. Above all I
would ask them to reflect that, in this case, the form of statement and the terms have
been settled by one who comes from the land and belongs to the race which created
the art of vault construction, and fostered its development from Roman times
onwards.

x TRANSLATOR'S PREFACE

As Italy was the cradle of Western achitecture, it is not unnatural to find that
the Italian architectural vocabulary is in some respects richer than our own. Accord-
ingly we have thought it advisable, in the interests of simplicity and convenience, to
Anglicize a few Italian technical terms, such as the following. " Lesena" has always
been retained in place of " pilaster-strip." " Pulvino," of which various renderings
appear in English writers, naturally becomes " pulvin." The rudimentary pendentive
or squinch which is described in Italian as a " raccordo d'angolo," in the same way
naturally appears as " raccord." Other phrases have been translated in the simplest
and most direct manner, so that, for instance, " crociera di sesto rialzato" appears
(without having recourse to the expedient of rendering one foreign language by
another, and using such a term as " surhaussee ") as " raised," and in the converse case
" unraised," " cross vaulting." We can only ask for a kindly reception and considera-
tion of these and similar innovations.

G. McN. RUSHFORTH.
October, 1909.

The great majority of the illustrations which appear in this book are derived
from photographs taken expressly for the work. Many of those in Volume II are by
Lionel Johnson, Esq. For a few I am indebted to Miss P. Bruce and Miss Bulwer ;
also to Dr. T. Ashby, Prof. Camille Enlart, Dr. Paul Gauckler, Dr. Henry Gee,
Harold Johnson, Esq., and F. Tuckett, Esq. ; to all of whom I offer my sincere
thanks.

CONTENTS OF VOL. I

PAGE

INTRODUCTION ........ .3

CHAPTER I

THE ROMANO-RAVENNATE AND BYZANTINO-RAVENNATE STYLES (FROM
HONORIUS TO THE END OF THE KINGDOM OF LOMBARDY) . . 7

CHAPTER II

THE COMACINE MASTERS . . . . . . . .108

CHAPTER III

THE PRE-LOMBARDIC STYLE (FROM THE REIGN OF AUTHARIS TO THE FALL

OF THE KINGDOM OF LOMBARDY) . . . . . .112

CHAPTER IV

ARCHITECTURE IN ITALY AND DALMATIA IN THE TIME OF CHARLES
THE GREAT ......... 151

CHAPTER V

PRE-LOMBARDIC ARCHITECTURE FROM THE CONQUEST BY CHARLES THE
GREAT DOWN TO THE APPEARANCE OF THE LOMBARDIC STYLE, AND
ITS COMPLETION . 161

PART I

VOL. I

INTRODUCTION

unravelling of the tangled skein which involves the dark age of
Italian architecture, from the second half of the Vlth century to the
second half of the Xlth, that is to say, from the descent of the Lombards
into Italy down to the appearance of the Lombardic or Comacine style, has been
a very slow process. The explanation is to be found in the confusion engendered
by the disputes, and sometimes the mistakes and prejudices of not a few of the most
prominent writers on this subject with regard to the age and style of monuments
of that epoch. Another reason is that such writers either study books too much
and the monuments too little, or else they shut their eyes and accept statements
which have no foundation in fact, or wander out of the way, restlessly seeking in
distant lands for the origins, influences, examples, and craftsmen, which they might
have found without any trouble in Italy.

Hence the history of Lombardic architecture remains, on the whole, to-day
a great collection of problems ; and a solution of them which, if not definitive,
shall be at least fairly approximate, is a matter of neither easy nor rapid
accomplishment.

So much by way of preface. I will now indicate in a few words the conception
which forms the basis of the present work.

Lombardy was the cradle of the style which preceded the Lombardic no less
than of the Lombardic itself. It was the product of the Comacine or Lombard gilds,
and its real beginnings must be referred to the days of the Lombard King Autharis
(583-590) and his immediate successors Theodelinda (590-625) and Agilulf (590-615),
when the School of Ravenna, founded in consequence of the transfer to that city of the
seat of the Western Empire (404), had already entered upon its long course of decay.

It is my privilege to reveal for the first time to the world of students this

15 2

4 LOMBARD1C ARCHITECTURE

unknown but important School. It was from the productions of its craftsmen that
the Lombard gilds derived their inspiration, borrowing, to begin with, merely various
original motives of architectural decoration, and then some organic elements of
construction, both of primary and secondary importance. To these productions we
shall devote a special study.

In the days of the above-named sovereigns, when these gilds (whose members-
the Comacine masters though they have been the subject of much discussion, I
shall deal with but briefly) were called in to erect their buildings, the condition of art
in Lombardy, as indeed in every other part of the peninsula subject to the Lombards,
was deplorable. It is true that, ever since Maximian (286-310) had made Milan his
official residence, the Emperors, who were often kept there by the necessity of defence
against various barbarian enemies, had turned their attention to beautifying the city,
restoring existing buildings and erecting new ones. Hence Ausonius could write : l

Omnia giiae magnis operum velut aemula formis
Excellunt nee iuncta f remit vicinia Romae.

In this way a wide field for the exercise of their ability was offered to the crafts-
men of Italy, but especially to those of Lombardy, who for centuries had found no
lack of employment in their own district, particularly at Milan, which from the time of
Augustus has been an important and wealthy city. In 404, however, Honorius
transferred his court to Ravenna, and this was regarded as the seat of government
and the capital of Italy until Odoacer put an end to the Western Empire (476). On
such an occasion it is reasonable to suppose that not a few, and among them the
best craftsmen of Milan, migrated thither.

From that time onwards the fortunes of Milan, and with them the conditions
of Lombardy steadily declined, until a climax was reached in the devastations of
Attila (452). Restored in the days of the first Gothic kings, so that Milan became,
after Rome, the first city of the West for size, population, employment, and wealth,
its prosperity came to a speedy downfall at the hands of Uraias (539). Revived to
some extent by Narses after he had been made exarch (554), its condition sank to
the lowest depths under Alboin (568-572), Clefi (572-573), and the confederate
Lombard dukes (573-583), whose history, like that of the conquered Italians, is one
unbroken tale of oppression, plunder, and bloodshed. To these calamities was
added the double scourge of pestilence and famine which, about the year 556, smote

Monumenta Germaniae his/erica. Ordo urbium Hobilinin.

INTRODUCTION 5

Lombardy in common with the rest of Italy. Next came the transfer of the
episcopal see of Milan to Genoa in consequence of the flight of Bishop Honoratus
(568-572) in terror at the approach of the hordes of Alboin.

Hence we need not be surprised if the first productions of the Comacine gilds
were only such as the unhappy times allowed of. Still, the numerous buildings
which they were called upon to erect, restore, or decorate, during the Lombard
rule (under which they enjoyed special privileges) gave their members opportunities
of exercising hand and brain, and of raising art to some extent from the slough
of barbarism in which it was plunged. Such progress did they make that, in the
VHIth century, we find that, while in buildings of their o\vn creation hardly any
advance in the principles of construction is to be discovered, progress in architectural
ornament and decorative carving is fairly well marked. We find, too, that Lombard
carvers were in such request for work of this kind, that their presence is apparent
in works executed in that age, not only in many parts of Italy, but also in Dalmatia
and in the countries north of the Alps. Later, in the time of Charles the Great,
and after he had become king of the Lombards (774), the Comacine masters had
the opportunity, besides employing their chisels in a far wider field, of taking part,
together with the craftsmen of Ravenna, in the construction of the most important
buildings erected by that monarch, or produced in imitation of them ; and they
acquired in the course of their execution some degree of experience in the difficult
and, to them, unfamiliar art of vault construction.

Fortified by the valuable knowledge thus acquired, the Lombard gilds proceeded
to take part in the erection of the numerous and occasionally sumptuous buildings
raised by the liberality of two magnificent prelates, Angilbert II (824-860) and
Anspert (869-882', in Milan and the districts subject to their spiritual authority :
and it was then that were laid the first solid foundations of the Lombardic ecclesias-
tical architecture of the future. Indeed, from this time onwards, we sec them eager
in searching among the ancient buildings of Rome and Ravenna for elements which,
when developed, would lend themselves, by means of a rational evolution, and supple-
mented by new ones, to transform Roman architecture into a new style, thought out
by themselves, and destined to serve new needs, as well as to adapt itself to changes
of taste.

To facilitate and hasten an evolution of this kind there contributed mainly : the
fear that the world was coming to an end ; the widespread religious movement which
originated therefrom ; the breath of Liberty which stirred the peoples of Italy ; and,

6 LOMBARDIC ARCHITECTURE

lastly, the improved conditions of the country, resulting from the new life infused
into commerce and industry which had begun to make its presence felt there in the
second half of the Xth century.

In this way, when the first quarter of the Xlth century was reaching its close,
the Lombardic architectural organism was already formed, and in the second half of
that century there appeared the earliest types of the vaulted Lombardic basilica.

CHAPTER I

THE ROMANO-RAVENNATE AND BYZANTINO-RAVENNATE

STYLES

FROM HONORIUS TO THE END OF THE KINGDOM OF LOMBARDY

WHILE not denying it would he folly to do so the share of the East
in the birth of the arts of the West, I do not believe, as many do, that,
from the period when Honorius moved the Imperial residence to
Ravenna (404) down to the fall of the kingdom of Lombardy, Italy,
as often as she wanted to produce something not of mere rude workmanship,
found herself obliged to fall back upon artists from the East, whether painters,
mosaic-workers, goldsmiths, carvers, or architects and builders. My view rather
is that the better works, at least in the case of architecture and sculpture,
the two branches of art which have formed my special study, or those which
in any way influenced the origins of Lombardic architecture and were executed
during those centuries in the exarchate of Ravenna, the kingdom of Lombardy,
and the duchy of Rome, ought to be assigned as follows. In the case of
architecture to Italian craftsmen, mainly, however, those" of Ravenna, with whose
productions we shall accordingly be almost exclusively occupied. In the case of
sculpture restricting ourselves to works of purely Byzantine style executed in the
days of Theodoric (493-526) and of Justinian I (527-565) to Greek artists, after
allowing a very modest share to the chisels of the School of Ravenna. And, thirdly,
in the case of sculpture carried out in a form and style of carving which is merely
Byzantinesque, to Italian artists, and before all, those of Ravenna. This opinion I
shall support by historical considerations, but mainly by the study and comparison of
the monuments themselves.

A century after the death of Diocletian (313), the Western Empire, on the eve
of its disappearance, imparted to its last capital, Ravenna (404-476), a splendour
to which Constantinople was as yet a stranger.

In the days of Honorius (395-423) and Galla Placidia (408-451), contemporary
with whom were the archbishops Peter I (396-425), Exsuperantius (425-432 or 439),
Peter II Chrysologus (433 or 439-449 or 458), and Neon (449 or 458-477),! the best
craftsmen of Milan, whose opportunities for exercising their talents and making
money had, since the transfer of the Imperial residence, become few and far between,
flocked to Ravenna, attracted thither by the numerous works of importance in course
of execution, and by the hope of lucrative rewards. And so, long before the erection
of St. Sophia at Constantinople (532-537), the tomb of Galla Placidia was already
gleaming with the gold of its mosaics. Mosaics and marbles not less splendid

1 The dates of the archbishops of Ravenna are derived from the chronological table published by Giani,
Sltidi storici, \\ A. V 1 1. A Iciinc csserrazioni SH la cronologia di Agiiello Kavennalt.

8 LOMBARDIC ARCHITECTURE

were used to decorate, not to speak of other buildings, the Basilica Petriana at
Classis, the extent and magnificence of which made it one of the wonders of the city
(396-425), the basilica of San Giovanni Evangelista (425), the chapel of San Pier
Crisologo (433 or 439-449 or 458), and the baptistery of Neon (449 or 458-477). And
the buildings which contained these mosaics and marbles, together with other
remarkable works which served for the decoration or finish of the structure, showed,
either in their plan, or in their internal decoration, arrangements or motives of an
original character which constitute a new style, to which I give the name of Romano-
Raven riatel

To this style belong the following buildings still existing at Ravenna in a greater
or less state of preservation : the basilica of San Giovanni Evangelista (425), the
basilica of Sant' Agata (425-432 or 439), the chapel of San Pier Crisologo (433 or
439-449 or 458), the tomb of Galla Placidia (about 440), the basilica of San Francesco
originally San Pietro Maggiore (450), and lastly the baptistery of Neon (449 or 458-
477). The now rebuilt Basilica Ursiana or cathedral (370-384) belonged to the same
category.

THE BASILICA URSIANA was founded and completed by Archbishop Ursus
(370-396) who gave it his own name and dedicated it to the Resurrection. 1 It was

rebuilt in the XVIIIth century. From the
notes and drawings left by Fabri '' and
Buonamici 3 we know that it consisted of a
nave and four aisles, all of very spacious
dimensions, roofed with timber, and divided
by four rows of marble columns taken from
older buildings. The capitals were in some
cases sculptured with representations of the
Fig. i. -Ravenna. Basilica Ursiana. Plan eagle and the ram, and they supported pulvins

of Apse (370-384:) (From Biionaiinci, . ,

" Metro folil ana ,ii A'areuna.") ( pUU'ini ) Or impost blocks, marked With

crosses. The nave terminated towards the east

in an apse, five-sided externally, and semicircular internally. The half-dome was
formed of two super-imposed concentric rows of tapering terra-cotta tubes, fitting
one into the other. Above these was the framework of rafters and boards forming
the roof, covered with sheets of lead placed there, it seems, in the Vllth century, to
take the place of the original roofing which apparently consisted of tiles.

This church presents five ' notable features, the creations of the School of
Ravenna, and not of Constantinople as is the universal but erroneous belief.

(1) It is the oldest instance of a basilica with the apse at the east end. Before
this, apses were placed at the west ; and that not only in the Western world but also
in the Eastern, as is shown by Helena and Constantine's church of the Holy
Sepulchre at Jerusalem (327-335), an,d by the Constantinian basilica at Baalbeck,
erected in the great court of the temple of Jupiter (138-249), the remains of which were
brought to light during the recent excavations, 4 when I had the opportunity of
examining them on the spot.

(2) It is the first example of an apse, curvilinear internally, and polygonal

1 Monumenta Germaniae Historica Agnelhts, Liber ponlificalis Ecclesiae Ravenna/is.

2 Le sagre mtmorit di Ravenna antica.
' Metropolitana di Ravenna.

4 Puchstein, Schultz, Krencker, Enter jahresberiflit tiber die ausgrabungen in Baalbek. Jahrbuch des
kaiserlich dents f hen archaologischen Instituts. 1891.

THE SCHOOL OF RAVENNA

externally (Fig. i). The credit of this discovery must be given to the builders of
Ravenna, and not to the Byzantines. Indeed, it was only about the middle of the

Fig. 2. Salonica. St. Sophia. Apses (about 495).

Vth century that the Easterns began to substitute the Ravennate plan for the old

form of apse in the Roman civil basilica. As a matter of fact, the apses of the

basilicas at Jerusalem and Baalbeck just referred to, are curvilinear on both faces.

The same is true

of the apse of the

chapel at Nurueh,

also in Syria, which

is ascribed to the

IVth century. 1 At

Constantinople,

the first example of

a Ravennate apse

which can be dated

with certainty, is

afforded by the

church of St. John

Baptist, erected by

B. Porte mcygi
c Porte \Mki\on

Ravenna Basi | ica Ursiana. Section (370-384). (from Bueiiamifi,
" Metropolitana di A'aveiiaa. ")

Studius 2 in 463. " Metropolit

At Salonica, St.

Sophia provides the earliest instance (about 495) (Fig. 2). The older churches of

1 Butler, Publication of an Archaeological Expedition to Syria in 1899-1900. Architecture of Northern
Central Syria and the Djebel Hauran.

3 Corpus scriptorum historiae byzantinae Cedrcnus, Compendium historiarum a mundo condito usque ad
Isaacium Comnentim imperatoreni.

LOMBARDIC ARCHITECTURE

St. George (Vth century), Eski-Djuma, and St. Demetrius (first half of the Vth
century), have apses which are purely curvilinear. In Syria, two very early dated
specimens are presented by St. George at Ezra (515-516) and the cathedral of Bosra
(511-512).

(3) The oldest example of capitals surmounted by the tall pulvins, consisting of
inverted truncated pyramids, which are a characteristic feature alike of the

Ravennate and of the Byzantine
style (Fig. 3).

As a matter of fact, a contem-
porary instance is presented by
the arcaded apse of the Basilica
Severiana, later San Giorgio Mag-
giore, at Naples (Fig. 4), built by
the bishop Severus, 1 who, accord-
ing to Gams, 2 filled the see be-
tween 367 and (about) 387. In
this apse the pulvins are orna-
mented with the cross-monogram,
a form which some would tell us
does not appear after the Vllth
century, though a glance at Plates
LI V and LVII in Mabil!on, 3 or at
a few pages in Kraus,* is enough
to convince us of the contrary.
As Grisar observes, 5 the use of
the monogram lasts all through
the Middle Ages. Thus it appears,
for instance, on the border of the
arches framing the apses of San
Clemente and Santa Francesca
Romana at Rome, the mosaics of
which belong, respectively, to the
first and second halves of the
Xllth century.

Still, taking into considera-
tion the extensive use of pulvins
at Ravenna, and in view of the
fact that the Basilica Ursiana was
consecrated by 384, while the
date at which Severus began his
church cannot be precisely fixed,
it is more natural to refer their
. origin to the School of Ravenna

rather than to that of Campania.
From the point of view of construction, the initial form of the pyramidal pulvin

1 Man. Germ. Hist. Cesta efisfoponim Neapolitanorum. 2 Series episcoponim Ecclesiae Catholicae.

" De re diploniatica. * Real encyklofddie der christlichen al/erthiimer.

5 Niimio Bullettino di Archeologia Crist iana, 1 895 Una 'scuila classica di tnarmorarii medioevali. II tenipio
di Clitunno e la chiesa spoletitia di San Salvatore.

6 Fabri, op. cit.

Fig. 4. Naples.

San Giorgio Maggiore.
(367 about 387).

The old Apse

THE SCHOOL OF RAVENNA

1 1

Fig. 5. Denderah. Temple of Hathor. Portico (1st Century).

is to be found in the cube-shaped blocks, sometimes of considerable height, used by
the old Egyptian builders after the creation, some time under the XVIIIth dynasty,
of the open flower capital with the object of making the weight of the architrave
rest on the centre of the
capital and the column be-
neath it (Fig. 5). The
Etruscans, too, sometimes
employed such cubes rest-
ing on capitals, according to
representations in sculpture
(Fig. 6). The pulvin may,
perhaps, also be connected
with the broken architraves
from which the Romans
sometimes sprang their
arches, the pulvin being the
reduction of these to their
simplest form.

In Giuliano da San-
gallo's sketch-book in the
Vatican Library there is a drawing of arcades belonging to the theatre of Balbus at
Rome (erected and dedicated in B.C. 13), showing exactly the same feature. And in
the mausoleum known as Santa Costanza, outside the walls of Rome, erected for
the princesses of the family of Constantine the Great some time within the decade
326-335, or, to be more precise, between 326 and 329, 1 the arches which carry the
cupola spring from the architraves surmounting the twelve pairs of granite columns
radiating from the centre. These architraves, like the pulvins, serve the purpose of
providing the springers of the arches with a base corresponding to the wall which

they carry, while allowing the support beneath to be
much slighter without impairing the stability of the
structure (Fig. 7).

From this mausoleum was derived the ancient
baptistery, now church of Santa Maria Maggiore,
near Nocera dei Pagani (Figs. 8, 9). Its date, how-
ever, is not that of the age of Constantine, as is
shown by the construction which, though rude, is
more advanced than that of the Roman edifice.
Thus, the spacious dome, constructed of courses
which are horizontal in its lower part and radiating
above, is given a hyperbolic curve so as to diminish
the thrust ; while the circular aisle has a ramping
barrel vault, which therefore presses outwards to-
wards the base of the outer wall ; and it is crossed at
intervals by transverse arches springing from massive
vault piers, thus enabling the wall to be reduced to a moderate thickness. We must
not, however, put the date too late, especially when we consider the extensive use
made in it of Roman columns and capitals. The probability is that it was erected in
the second half of the IVth century, or at latest in the early years of the Vth, and

1 De Rossi, Musaici crisliani delle chiese di Kama anteriori al secolo XV.

Fig. 6 Vollerra. Museo Guarnacci.
Etruscan Sarcophagus.

LOMBARDIC ARCHITECTURE

in origin,

certainty well before 'th- baptistery of Soter, otherwise San Giovanni in Fonte (Vth
century), at Naples, which shows a distinct advance beyond it in scientific construction.
From Ravenna and Naples the pulvin spread over Italy and beyond. At Rome
typical specimens have survived in Santo Stefano on the Via Latina, built, as the
inscription tells us, by Demetria in the time of Pope Leo I (440-461). They may
also be seen in the round church of Santo Stefano on the Celian (Fig. 10), which
protably, was a building' of classical times (65-68) intended for civic

purposes, recon-
structed on the same
plan and with the
same object in the
days of Valens (364-
378) and Gratian
(367-383). damaged
by fire in 410, and
finally restored and
altered by Pope
Simplicius (468-
483), who dedicated
it to Christian wor-
ship. 1 2 To this trans-
formation belongthe
pulvins of the outer
colonnade. At
Perugia they were
used in the round
church of Sant' An-
gelo, which should
be dated about the
middle of the Vlth
century, before the
Lombard invasion
(568), for some of the
capitals recall others
belonging to the
reign of Theodoric
(493-526) (Fig. ii).

In the East pulvins were not introduced before the Vth century ; and if we are
referred to buildings in the Byzantine Empire of earlier dates than the basilicas of Ursus
and Severus, such dates are erroneous. Thus, for instance, the Cistern-basilica at Con-
stantinople ascribed to Constantine 3 though, as Van Millingen observes, 4 what steps
that emperor took to bring water to his new capital (328), is a matter of pure conjecture
is really the work of Justinian I (527-565), who, if he did not build it, restored it. 56
In places it shows the use of pulvins alone, instead of capitals surmounted by pulvins

1 Lanciani, The ruins and excavations of Ancient Koine. 2 Grisar, Storia di Roma e dci Papi ncl itiedio evo.
3 Du Cange, Historia Byzantina. Constantinoplis Christiana. * Byzantine Constantinople.

5 The dates of the Byzantine Emperors are taken from the chronological table in Van Millingen's Kyiantine
Constantinople.

6 Gyllius, De topographia Constant inopoleos et de illiits antiquitatibus. De Bibliotheca Basilica, ct de regia
pot-licit, et de Basilica Cisterna.

Fig. 7. Rome. Santa Costanza (IVth Century).

THE SCHOOL OF RAVENNA

in order to fit the columns
taken from older build-
ings. Its vaulting shows
a close affinity in con-
struction with that of the
other cistern at Constan-
tinople known as " Binbir-
direk" "of the 1001
columns," assigned by
Forchheimer and Strzy-
gowski 1 to the year 528.
The same may be said of
the cistern known as that
of Arcadius, also at Con-
stantinople, where the
capitals betray their
affinity with those of the
cistern - basilica just re-
ferred to, and the stilted groined
In the Eastern Empire

Fig._8. Nocera dci Pagani. Santa Maria Maggiorc (IVth or Vth Century).

vaults indicate an age not earlier than that of Justinian.

no pulvins were to be seen before the Vth cen-
tury, even in mosaics
or sculptured representa-
tions. A proof of this is
the grand mosaic in the
dome of the round church
of St. George at Salonica,
in which the numerous
ecclesiastical buildings
supported by columns
adorned with various
kinds of capitals, do not
exhibit a single pulvin
(Fig. 12). To this day 2
some think that in this
church the external flying
buttresses corresponding
to the sanctuary arch, are
original, thus perpetuating
the error into which
Texier and Pullan fell; 8
whereas they are a later

addition,
however,

This device,
was already

ig- 9. Nocera dei Pagani. Santa Maria Maggiere {IVth or-Vlh Century).

1 Die bywntinischen Wasser-
behalter ran Konstantinopel.

2 Journal of the Koyal Institute
of British Architects, 1907 Gour-
lay, Salonika : the ancient Thessa-
lonica.

5 L 'architecture hyzantine.

LOMBARDIC ARCHITECTURE

Fig. ID. Rome. Santo Stefano al Celio (IVth and Vth Centuries).

Fig. 12. Salonica. St. George. Mosaic (Vth Century).

THE SCHOOL OF RAVENNA

known to the Romans, for we still possess l the plan of a building in which the
frontal arch of the apse and the apse itself are strengthened externally by pilasters
kept up on the outside by buttresses pierced with relieving arches which carry the
thrust (Fig. 13).

The date of this round church of St. George must be fixed after the second quarter
of the IVth century, in view, among other things, of the mosaics in the heads
of some of the semicircular recesses, which show an obvious derivation from the
compartments of mosaic decoration on the annular barrel vault of the mausoleum
of Santa Costanza at Rome.
But it must be earlier than the
middle of the Vth century, on
account of its apse which pre-
sents, just like that of the basilica
of Eski-Djuma at Salonica, a
high plinth on the outside, from
which rise buttresses, instead of
an elegant arcade with marble
shafts, like the apse of St. Deme-
trius in the same city which
must have been erected about
the middle of the Vth century.
We may therefore place it in
the closing years of the IVth or
the early ones of the Vth cen-
tury, after the appearance of the
arcaded choir in San Sebastiano
outside the walls of Rome, the
dated prototype of this arrange-
ment going back to the days of
Pope Damasus (366-384)* and
in the contemporary choir of San
Giorgio Maggiore at Naples (367
and about 387). A precisely
similar design may be observed
in the mosaics of St. George at
Salonica. The representation
of St. Porphyrius. who must be

the Porphyrius of the time of Arcadius (395-408), among the figures of saints
in the mosaic of the dome, would lead us to fix the date of the church prefer-
ably in the first years of the Vth century, and before the construction of St.
Demetrius. Our statement about the comparatively late introduction of pulvins
in the East is also supported by the evidence of the base of the obelisk of
Thothmes III, set up by Theodosius the Great (378-395) in the Hippodrome of
Constantine at Constantinople, where the coarse bas-reliefs show the Imperial
tribune surmounted by an arch which springs from the capitals svithout the
interposition of pulvins (Fig. 14}.

The first appearance of pulvins in the Byzantine world occurred, apparently.

Fig. ii. Perugia. Sant' Angelo (Vlth Century).

1 Vatican Library. Cod. Lat. 3439.
8 Duchesne, Le liber ponlificalis.

LOMBARDIC ARCHITECTURE

.*/<

o
o

o
O

in the old basilica of Eski-Djuma at
*, Salonica, a city of primary import-
ance in the Empire, which must
have been erected about the close
of the first quarter of the Vth cen-
tury, for the following reasons.

First and foremost, the capitals
of the colonnades at the lower end
of the nave (Fig. 15) present a
Byzantine version of the Roman
Composite capital, to which no
certainly dated parallel can be
found before the Vth century. This
version, with its acanthus foliage,
where on the surface of each
, acanthus leaf there appears the
outline of a smaller leaf traced by
drill holes, is earlier than the
more advanced and more purely
Byzantine Composite capitals, with
foliage of the acanthus spinosus
packed into shells and surmounted
by birds, which are to be seen in
I the apse of the basilica of St.

Fig. 13. Rome. Plan of an Ancient Building. Demetrius in the same city.

Next, the two tiers of colonnades consist simply of columns, whereas those in
St. Demetrius
(Fig. 1 6) are
composed of
columns and
piers, an
arrangement
which marks
a construc-
tive advance
beyond that of
Eski-Djuma.

Then, in
St. Demetrius
there may be
seen, in addi-
tion to those
which we have
described, By-
zantine Com-
posite capitals
(Fig. 17) with
single acan-
thus leaves Fig. 14. Constantinople. Base of the Obelisk of Theodosius (378 395

THE SCHOOL OF RAVENNA 17

minutely and sharply serrated, each separate, or with the points meeting arch-wise.
This free movement of the leaves originated later than the time of Antoninus
Pius (138-161) in Syria, 1 where some of the earliest examples may be noticed
in the great square court of the temple of Jupiter at Baalbeck, which court
was erected in the reigns of
Septimius Severus and Caracalla
(193-217) In Italy, its begin-
nings may be seen in a support
of a table in the house of
Cornelius Rufus at Pompeii.
These Composite capitals, taking
into account the novelty of their
design and the fineness of the
execution, are another proof that
St. Demetrius belongs to a later
date than Eski-Djuma.

Lastly, the apse of Eski-
Djuma is still devoid of orna-
ment on the outside, whereas
that of St. Demetrius is decorated
with arcades.

Having settled the question
of the priority of Eski-Djuma
over St. Demetrius, let us see
whether we can approximately
fix the dates.

St. Demetrius presents three
new types of Byzantine capitals :
the cubical or melon-shaped ; the
Composite, with leaves blown by
the wind in two opposite direc-
tions (Fig. 18); and, lastly, the
bird and basket Composite, in
which birds take the place of
volutes (Fig. 19).

Now if the first two of these
types be compared with the
cubical funnel-shaped capitals
(Fig. 20), and the Composite
ones showing the leaves blown
by the wind from right to left,
in St. Sophia at Salonica, com-
pleted in 495, the date given by

Fig. 15. Salonica. Eski-Djuma (Vth Century).

its mosaic inscription, it will be found that the latter, particularly the Composite ones,
which, to my mind, are the most beautiful specimens of this type of the Vth and
VI th centuries which the East can show, reveal an art in a more advanced stage than
that of the capitals of St. Demetrius.

1 Rivoira, Delia scoltura ornanuiitale dai tempi di Roma imferiale al Mille, in the Niiova Antologia,
1904. No. 790.

VOL. I c

LOMBARDIC ARCHITECTURE

Accordingly, the foundation of St. Demetrius, for the exact year of which
no certain evidence exists, with the exception of a fragmentary inscription on marble
relating to a donation of Justinian II 's (685-695), 1 must be somewhat earlier than
that of the before-named St. Sophia, and goes back to about the middle of the
Vth century. The church of Eski-Djuma must therefore have been built before the

Fig. 16. Salonica. St. Demetrius (Vth Century).

middle of the Vth century, and after the IVth century ; in other words, about the end
of the first quarter of the Vth century.

(4) The archetype of the domical vault entirely constructed of tapering tubes
(Fig. 21) inserted one inside the other. This tubular concentric system, which was
also employed in the apse of the basilica of Sant' Agata (425-432 or 439) at

1 Papageorgiu, Un tdit de FEmferenr Justinien II en faveur de la basiliqite tie Saint-Dtmitritis H
Salonique.

THE SCHOOL OF RAVENNA

Fig. 17. Salonica. St. Demetrius.
Capital (Vth Century).

Fig. 18. Salonica. St. Demetrius. Capital
(Vth Century).

Fig. 21. Ravenna. Basilica Ursiana.
Tube from the Apse (370-384).

Fig. 19 Salonica. St. Demetrius. Capital
(Vth Century).

Fig. 20. Salonica. St. Sophia. Capital
(about 495).

C 2

LOMBARDIC ARCHITECTURE

Ravenna, quickly spread in Italy, and was made use of, for instance, by Pope
Hilarius (461-468) when he added three chapels to the baptistery of St. John
Lateran at Rome, rebuilt by Xystus III (432-440); and also in the (presumable)
basilica of Fausta at Milan, where in front of the apse is a transept crowned by a
cupola, regarded as a work of the end of the Vth or beginning of the Vlth

century. 1

Its origin, the
credit of which must
be given to the School
of Ravenna, is to be
sought in the Roman
device of relieving
the weight of a dome
by means of con-
centric rows of am-
phorae, in conjunc-
tion with the other
expedient of hollow
tubes, sometimes em-
ployed by the builders
of Numidia in their
vaulting. I may re-
mark that, before the
Romans and the
Numidians, the Cam-
panians had made use
in their vaulting of
amphorae and terra-
cotta tubes: the
Thermae Stabianae
at Pompeii are there
to prove it. Of the
former method of re-
lieving a dome, a
very early example is
furnished by a poly-
gonal hall standing
near the circular

sepulchral edifice known as the " Tor de' Schiavi," in the Villa of the Gordians on
the Via Praenestina near Rome; a villa which, as a whole, is ascribed to the Illrd
century. 2

Another instance of somewhat later date is afforded by the mausoleum of
St. Helena on the Via Casilina, also in the neighbourhood of Rome (IVth century)
(Fig. 22). The use of amphorae in ordinary vaulting is as old as the time of
Caligula (37-41), in whose palace on the Palatine they are employed to diminish the
weight on the haunches.

As for the hollow terra-cotta tubes inserted one inside the other, the Baths

1 Landriani, La basilica ambrosiana. I resti del/a basilica di Fausta.

2 Nibby, Analisi slorico-topograjico-anliquaria del/a car/a de 1 dintorni di Roma.

Fig. 22. Rome. Mausoleum of St. Helena (IVth Century).

THE SCHOOL OF RAVENNA

of Tabarka (not earlier than about the middle of the Illrd century), demolished
in 1900, but described to me by Gauckler, the late Director of Antiquities in Tunis,
exhibited concrete vaulting which partly rested
on centering of tubes of this description.

(5) The oldest example of a spherical vault
in masonry, with a wooden roof above it. This
treatment is not met with in any dated Eastern
structure of the kind earlier than, or even
contemporary with, the Basilica Ursiana at
Ravenna.

THE BASILICA OF SAN GIOVANNI EVAN-

GELISTA was founded by Galla Placidia 1 in

425.2 In the XVIIIth century, as the building

had sunk, the columns were raised and the

arches of the nave rebuilt, the walls above

them being also raised. In spite of the altera-
tions which it has undergone the church pre-
serves its original form.

The nave (Fig. 23), which looks east, ends

in an apse, curvilinear internally and polygonal

externally. The aisles, on the other hand,

terminate in two rectangular sacristies. This

arrangement, a very early instance of which is

afforded by the church of Musmieh in Syria

(Fig. 24), fitted up, according to De Vogue, 3

before the
IVth century,
in a Roman
Praetorium of

the time of the Emperors Marcus Aurelius (161-180)
and Lucius Verus (161-169), was no new thing in
Italy. Previous examples are: (i) the apse of the
basilica-like structure built of square blocks of
peperino with bonding courses of bricks in the
ruins of the villa known as the " Sette Bassi " on
the Via Latina near Rome, which, as the brick-
stamps show, belong to the period between the years
100 and 155 ; 4 (2) the basilica of the Xenodochium
erected by Pammachius at Porto near Rome, about
398 5 (Fig. 25); (3) the large basilica of Santa
Sinforosa on the edge of the Via Tiburtina, con-
sidered to be not later than about the Vth century ;

1 Man. Germ. Hist. Agnellus, Liber pontijicalis.
1 Fabri, op. cit.

3 Syrie centrals. Architecture civile <t religieusc dtt f au
VII' siicle.
1 Papers of the British School at Rome, Vol. IV. Ashby, Classical topography of the Roman Campagna.

5 Bullettino <t archeologia cristiana, 1865. Lanciani, / monumenti cristiani di I'orlo.

6 Gil studi in Italia, Anno I. Stevenson, La basilica di S. Sinforosa e di suoi settefigli al nono miglio della
Via Tiburtina.

Fig. 23. Ravenna. San Giovanni Evar.gelista

(4*5)-

Fig. 24. Plan of Church at Musmieh
(Ilnd and Illrd Centuries).

LOMBARDIC ARCHITECTURE

Fig. 25. Porto. Xenodochium of
Pammachitis. Plan of Apse in
the Basilica (about 398).

(4) San Salvatore at Spoleto, not later than the IVth

century; and (5) the basilica of St. Paulinus at Nola

(391-402).

This arrangement
originated in Rome. The
basilica of Domitian's
Imperial Palace, begun
by Vespasian after the
year 69, has its apse con-
fined between two ser-
vice rooms (Fig. 26).
Syria took it from Rome,
and applied it not only in

the Praetorium of Musmieh, now destroyed, but also in

the Tychaion at is-Sanamen (IQ2). 1

The walls of San Giovanni at Ravenna, constructed

exclusively of courses of bricks separated by beds of

mortar of irregular thickness, are decorated externally on the sides of the nave with

blank arcades resting on a plinth. The same feature occurs in the basilica of Sant'

Agata erected by Gemellus, administrator of the Church of Ravenna in Sicily under

Fig. 26. Palace of Domitian.
Plan of Basilica (about 69-85).

Fig. 27. Ravenna. Sant' Agata (425-432 or 439).

the archbishop Exsuperantius (425-432 or 439) 2 (Fig. 27). Within every arch a large
round-topped window opened, the head of which, like the arch itself, is framed by a

1 Revue Archiologiqut, 1906 Butler, The Tychaion at is-Sanatnen and the plan of early churches in Syria.

2 Mon. Germ. hist. Agnellus, Liber pontificalis.

THE SCHOOL OF RAVENNA

Fig. 28 Rome. Sketch of an ancient Bas-reliei.

ring of bricks laid lengthwise. The arched corbel course above is an addition of
the XVIIIth century.

This motive of arcading, a favourite one with the School of Ravenna, had already
been applied to other ecclesiastical buildings for example, the church of San Protaso

at Como (391-420),* now turned into a
ww factory, and was borrowed from the
Romans, who employed it from the age
of Augustus onwards, the arcades being
sometimes completely blank, at others
pierced with openings. The octagonal
vestibule of the Piazza d'Oro in Hadrian's
villa at Tivoli (125-135) is decorated with
blank arcades. In Sangallo's sketch-book
in the Vatican Library may be seen a group
of round buildings at Baiae, to be ascribed
to the time of the latter Emperor, one of
the most flourishing periods of that
famous bathing place, decorated internally
with high and narrow blank arcades.

Another drawing shows a portion of the " Crypta Balbi " presenting a range of
blank arcades in its upper story. In the Vatican volume previously referred to, 2 there
are drawings of fragments of reliefs of the classical period representing a triumphal
procession and a sacrifice, with an architectural background displaying, in addition
to series of isolated or continuous pediments, arcades, apparently blank, springing
from pilasters or columns (Figs. 28 and 29). There have recently been discovered
at Terni in front of the Porta Spoletina the basements of three Roman tombs, 3
assigned by some to the family of the
Taciti, 4 the elevation of which has been
preserved to us by a pen-and-ink sketch in
the Uffizi at Florence. One of these tombs
was encircled by a blank arcade. In
Schedel's panorama of Rome (1493), repro-
duced by De Rossi, 5 similar arcades may be
seen on the exterior of two circular buildings
standing to the right of the Flavian Amphi-
theatre. Others are to be observed on three
structures of the same kind adjacent to
the Baths of Diocletian in the Mantuan
bird's-eye-view plan of Rome, also repro-
duced by De Rossi ; and lastly, on two
exactly similar buildings situated to the
right of the aforesaid baths, represented in

a panorama of Rome which I noticed in a picture in the Stadel Institute at Frankfort.
In the ruins of the villa which goes by the name of " Centroni " (Illrd century)

1 Kivista archeohgica delta Provincia di Como, fasc. 25 Chiesa di San Protaso net sokborghi di Como.

2 Vatican Library, Cod. Lat. 3439.

' Bollettino d'Arte, Anno II Sordini, Dei sepolcri dei Taciti in Terni.

4 Alti delta K. Accademia dei Lincei, Anno CCCIV. Notizie degli scavi di antichM.l&na, Terni,
Scoperte nel suburbia.

5 Piante icnografiche e prospettiche di Kama anleriori al secolo Xl'/.

Kig. 29. Rome.

Sketch of an ancient Bas-
relief.

2 4

LOMBARDIC ARCHITECTURE

(Fig. 30), and of another near " Fontana Piscaro " (to be ascribed to the I Ind century),
both on the Via Latina near Rome, the outer fronts can still be seen, faced with blank

arcades springing from
pilasters with engaged
half-columns. In the case
of " Centroni " the arcades
are pierced by circular
openings intended to light
the cryptoporticus within.
The East, too, can
show early examples of
arcades designed to break
the monotony of flat,
uninteresting, cold walls
of buildings. They do
not, however, go back to
such remote times as
some writers suppose ;
Dieulafoy among them. 1
As, for instance, the
palace of Firuz-Abad in

Fig. 30. Rome. Ruins of the Villa called " Centroni " (Illrd Century). Persia, which is not older

than the end of the

Sassanid epoch (226-651), as we shall see in due course. The earliest dated
specimen of this form of architectural decoration in Persia is presented by the still

Fig. 31. Ctesiphon. Palace of Chosroes I (531-579). (From Dieulafoy,
" Uart antique dans la Perse."')

existing fagade of the palace erected by Chosroes I (531-579) at his capital,
Ctesiphon (Fig. 31).

1 Uart antique dans la J'ene.

THE SCHOOL OF RAVENNA

It is appropriate to mention, at this point, a fact which has escaped the observation
of students of the origins of ancient styles of architecture. It would seem from a
statement of Faustus of Byzantium (Bk. V, ch. iv), who lived in the IVth century (in
the Vatican MS. 9545 is a translation of his work), that if the Sassanids left great
monuments, these monuments were erected with the aid of builders from the Romano-
Byzantine Empire. As a matter of fact, this passage, which has been kindly
translated for me by the distinguished Armenian scholar, Mr. Conybeare, relates that
Urhnayr, King of the Albanians, with his army, before entering on a battle together
with the Persians against the Armenians who were allied with the Greeks, carefully
exhorted his own men to spare the lives of the Greek prisoners so that they might be
available for making bricks and mortar, and could be employed as carvers and
masons for the construction of cities, palaces,
and other requirements. From this statement
we may reasonably suspect that, like the
Albanians, the Persians also made use of
Roman builders ; a fact which would support
the theory of direct Greek and Latin influence
on Persian art. The fact of such influence is
confirmed by a passage of Theophylactus, in
whose time (638) it was believed that Justinian
had provided Chosroes I not only with Greek
marbles, but also with the architects and
builders for the palace at Ctesiphon : " They
say that the Emperor Justinian sent to
Chosroes Greek marble, and skilled architects
and master masons who built him a palace in
the Roman style not far from Ctesiphon." 1

Another good instance of wall arcading is
the church of St. John Baptist founded by
Studius at Constantinople (463), where the
three blank arcades of the apse as recon-
structed by Salzenberg 2 are apparently the
result of the restoration which the building

underwent at the time of Constantine Palaeologus Porphyrogenitus, brother
Andronicus II Palaeologus (1282-1 328).*

Externally, the walls of the chapels which form the termination of the aisles of
San Giovanni Evangelista are strengthened at the angles by two buttresses, and are
ornamented at the top by bands enclosing a cornice of bricks set saw-tooth wise.
This form of ornament was not borrowed, as some think, by Ravenna from the
Byzantines, for it was only in the Vth century that the latter began to decorate the
topmost cornices of their ecclesiastical buildings with it; so that, while at Salonica
the church of St. George (first years of the Vth century) is merely finished off at the
top by plain stringcourses, in the basilicas of Eski-Djuma and St. Demetrius (Vth
century) the use of the saw-tooth may be seen (Fig. 32). It was derived, as a matter
of fact, from the Romans, who had used it since the times of Maxentius (306-312),
Valens (364-378), and Gratian (367-383), as is proved by the Heroon of Romulus,

1 Corpus script, hist, byz. Theophylactus Simocatta, Hisloriae.

- Alt-christliche baudenkmale von Konstantinopel vom V bis XII jahrhundert.

3 Du Cange, Hist. Byz.Constantinopolis Christiana.

Fig. 32. Salonica. St. Demetrius.
(Vth Century).

Apse

LOMBARDIC ARCHITECTURE

son of Maxentius, erected in 309 (Fig. 33), and by the drum of the church of Santo
Stefano al Celio at Rome, which clearly belongs to the days of Valens and Gratian.

The apse (Fig. 34), which has been tampered with, was of two stages, the lower
decagonal, the upper heptagonal. The latter is decorated with an arcade of seven
arches resting on pairs of shafts joined back to back by two lateral projections in
which were undoubtedly fixed the transennae which closed the openings (Fig. 35).

The square cam-
panile, the upper part of
which has been altered, is
of a later date than the
church itself. Evidence
of this is to be found in
the fact that it was formed
at the expense of the last
bay of the south aisle, and
in the use of fragmentary
materials in its construc-
tion which has nothing
in common with the uni-
formity of that of the
basilica. If one may judge
from the sculptured
foliage of a pulvin in one
of its windows, the work-
manship of which betrays
a hand near to the Xlth
century, this campanile
should be ascribed to the
Xth century.

San Giovanni Evan-
gelista is distinguished by
two notable character-
istics.

The first is the apse
enriched by arcading sup-
ported by columns, a
decorative feature which
speedily made its way
through the East, where
the oldest instance that
I can cite is the basilica
of St. Demetrius at

Fig. 33. Rome. Heroon of Romulus (309).

Salonica (of about the middle of the Vth century), in which the apse presents
a semicircle of five arches, originally closed by transennae, supported by
columns.

The second is the visible framing line flush with the arches, each of which
is enclosed by a ring of bricks laid lengthwise and fitting exactly. As is obvious,
this is not a question of the rings of brickwork which the Romans some-
times used for constructive reasons, in order to relieve the arch from the weight

THE SCHOOL OF RAVENNA

of the super-incumbent wall, added, as in the present
case, to the desirability of avoiding the defective
junction between bricks set lengthwise and others
radiating vertically. Rather we have here a new decora-
tive motive.

This feature was appropriated by the Lombard gilds,
who embellished it by the use of polychrome materials
in the way which may be seen, for instance, in the
basilica and baptistery at Agliate (824-860), and in the
parish church of San Leo (879-882). The Greeks also
made large use of it in their ecclesiastical buildings of
the Xlth century, with the addition sometimes of one or
two rows of saw-tooth. We may refer in this connection
to the old baptistery, now the church of the Holy
Apostles (to be ascribed to the first years of the Xlth
century), and the churches of Kapni-
karaea (to be ascribed to the Xlth
century, and not to the time of
the Empress Eudoxia [421-460]), St.
Nicodemus, built by Lycos who died
in 1044, and St. Theodore (1049) at

35- Ravenna. San
Giovanni Evangelista.
Column from external
arcade of Apse (425).

Fig. 34. Ravenna. San Giovanni Evangelista. Apse (425).

Athens (Fig. 36). Other instances are the churches of the Virgin (1028) and of the
Apostles (to be ascribed to the Xlth century) at Salonica.

LOMBARDIC ARCHITECTURE

Fig. 36. Athens. St. Theodore (1049).

THE MAUSOLEUM OF GALLA PLACIDIA was erected by order of Galla Placidia
about the year 440, and was dedicated to Saints Nazarius and Celsus. 1 Its plan (Fig.
37) is that of a cross with arms of unequal length, the so-called Latin cross as
opposed to that with equal arms known as the Greek cross : plans which, in either
case, had their origin in Roman tombs, a fact of which anyone may convince himself
by a glance at the sketches of Bramantino, reproduced by Mongeri,' 2 of Montano, 3 of
Serlio, 4 and those in the Vatican volume already referred to. 5

Over the centre rises a square tower closed above by a conical dome resting on
spherical pendentives, each formed by a spherical segment merging in the cupola and

developed from a triangular rib projecting from the
re-entrant angle of the walls (Fig. 38)

The cupola is constructed of bricks (Fig. 39), and
its extrados is covered with amphorae (Fig. 40) set in a
bed of mortar, on which the tiles rest. The arms of the
cross are covered internally by barrel vaults. Externally
the brick walls are decorated with blank arches (Fig. 41).
In the tomb of Galla Placidia I should like to call
attention to the ground plan and to the pendentives of
the cupola.

So far as I am aware, there is no record of churches
or tombs older than this mausoleum, having the form
of a Latin cross with rectangular extended arms, and
not mere apses opposite to one another and starting
directly from the central space. For it seems that the
cruciform Constantinian church of the Apostles at Constantinople was equilateral.

Fig. 37- Ravenna. Plan of
Mausoleum of Galla Placidia
(about 440).

1 Tarlazzi, Memorie sacre di Ravenna.

2 Le ravine di Roma al principle del secolo XVI. 3 Li cinque libri di architettura.

4 De le antiqidta, 5 Vatican Library, Cod. Lat. 3439.

6 Bull, di arch, cristiana, 1866. F. Lanciani, Scoperte negli edijici cristiani di Ravenna.

THE SCHOOL OF RAVENNA

And the very ancient church of SS. Peter and Paul, now Sant' Abondio, at
Como, of the Latin basilica type, 1 was only erected in the middle of the Vth century
(Fig. 42). The basilica, too, of the Holy Cross, built by Galla Placidia 2 in the
shape of a Latin cross, and connected, by means of the portico in front of it, with
the mausoleum of the Empress, was not erected till towards the year 449- 3

Fig. 38. Ravenna. Mausoleum of Galla Placidia (about 440).

Secondly, the pendentives give rise to some important considerations.

The spherical pendentive, of which those of the tomb of Galla Placidia are a
complete type in brickwork, was a fairly ancient invention in Italy, where it begins to
show itself from the 1st century onwards.

At a far earlier period the Etruscan builders had been content to set the circular

1 A'ifista arch, della Provincia di Coma, fasc. 30 Barelli, Basilica di Sanf Abondio net sobborghi di Como.
- Man. Germ. hist. Agnelltis, Liber pontificalis. * Kabri, op. ctt.

LOMBARDIC ARCHITECTURE

base of a cupola, built of horizontal layers of stone projecting one beyond the other,
upon a structure of square plan by the aid of graduated pendentives, as may be seen

in the tomb known as that of the
"Diavolino" from Vetulonia,
now set up in the Archaeological
Museum at Florence, and con-
sidered to be of the Vllth cen-
tury B.C. It may be also noticed
in another remarkable tomb at
Vetulonia, which goes by the
name of " La Pietrera." Tombs
of the Vetulonian type were
also constructed at a late period
in Egypt, the Crimea, &C. 1

Whether the Roman builders
had developed the graduated
pendentive of the Etruscans,
who had been their teachers in
the matter of architecture, into
the triangular spherical one
long before the 1st century, as
seems natural and logical, it is
impossible to say, for the evid-
ence is still wanting.

It is well known that there
are two varieties of the spherical pendentive : in the one, the pendentive and the
cupola belong to different spherical planes and curves ; in the other, the dome is

continuous with the
pendentives.

According to
the evidence of ex-
isting remains, the
first kind had its ori-
gin in some such
structure as the cen-
tral chamber of the
inner west front of
the " Domus Augus-
tana"on the Palatine,
rebuilt by Domitian
about the year 85. 2
We find that in this
square room (Figs. 43
and 44), each side
measuring about 23

feet, the dome, much of which has now fallen in, was
sustained by the aid of triangular spherical pendentives

39. Ravenna. Mausoleum of Galla Placidia. Construction
of the Cupola (about 440).

Fig. 42. Como. Plan of the ancient
Basilica of SS. Peter and Paul, now
Sant" Abondio (Vth Century). (From
Boito, " ArchiUttitra del media evo in
I/a/ta.' )

Fig. 40. Ravenna. Mausoleum of

Galla Placidia. Amphorae from

the Cupola (about 440).

1 Atti del Congresso internazionale di scienze storiche, Koma, 1903 Pinza, Le origini lii alciini tipi delta
architettiira sepolcrale Tirrena nella eta del ferro. 2 Lanciani, The Kuias and Excavations of Aiident Kume.

THE SCHOOL OF RAVENNA 31

formed of lumps of tufa set in irregular courses on a framework of boards and
earth, and backed by concrete, details which have not been noticed hitherto. In
the two lateral rooms (each side about 30 ft.) which flank the square central
chamber, the square of the ground plan, converted into an octagon by means of
four semicircular niches at the angles, passes into the circle of the dome by an
irregular transition of the solid mass which forms the construction, composed, as

Fig. 41. Ravenna. Mausoleum of Galla Placiclia (about 440).

in the former case, of a layer of lumps of tufa set in mortar, forming a sort of case
on which the concrete of the vault was poured (Figs. 45, 46).

Next to the pendentives of the " Domus Augustana" come the similar ones in
the upper story of a tomb of square plan on the Via Nomentana near Rome, not far
from the " Casale dei Pazzi " (Fig. 47). The ornamentation and facing of this tomb
(Fig. 48) suggest a date contemporary with that of the tomb of Annia Regilla in the
" Valle Caffarella," and with other sepulchral buildings near the basilica of Santo
Stefano on the Via Latina near Rome, erected in the time of the Antonines
(138-192).

LOMBARDIC ARCHITECTURE

Typical is the external facing of these sepulchral buildings, which was in use at
Rome only during the Ilnd century. It does not appear on any existing building of

Fig. 43. Rome. Domus Augustana. Plan of
Central Chamber (about 85).

Fig. 45. Rome. Domus Augustana. Plan of Lateral
Chamber (about 85).

Hadrian's time (i 17-138), nor on any later than the age of the Antonines. This
kind of facing is composed of very regularly laid rows of red and yellow broken
fragments of flanged or unflanged tiles presenting to the eye only the edge which is

unbroken, or which being broken has been

Fig. 44. Rome. Domus Augustana. Pendentive of
Dome in Central Chamber (about 85).

made smooth. They are thinned with
the hammer in order that the inner
surface may take a larger quantity of
mortar for holding the bricks together ;
and this made it possible to use very
fine joints of mortar in the visible part
of the facing, so that the latter seemed
to be a homogeneous mass of brick.
The use of materials reduced to this
fragmentary condition was suggested
by the wish to utilise pieces rejected
from the brick kilns, and brick refuse ;
and moreover it was more economical
than using new material broken for the
purpose.

This kind of facing recalls the
moulded brickwork used for cornices,
grooved on the inner face to give a
good hold for the mortar and avoid
its appearing on the outside. A very
early instance of such cornices is pro-
vided by the exterior of the original
curtain walls of the Praetorian Camp
at Rome (23 A.D.).

The earliest example of the second

THE SCHOOL OF RAVENNA

Fig. 46. Rome.

Domus Augustana.
(about 85).

Dome of Lateral Chamber

kind of spherical pendentive is afforded by another sepulchral edifice of the Ilnd
century on the Via Nomentana, popularly known as the " Sedia del Diavolo,"
where the cupola of the
upper story was sustained
by pendentives formed, at
the base of the triangle
by plaster, then by courses
of fragments of broken
bricks laid in such a way
as to second the radius
produced by the penden-
tive, and thirdly, by layers
of lumps of tufa backed
by the concrete (Fig. 49).
Next comes an exam-
ple showing an advance,
perhaps because the tran-
sition was easier to effect,
being generated in a curve
of larger radius. It has a
facing entirely of brick
with concrete above, and is to be seen in a polygonal chamber, strengthened by
a central pier in the Middle Ages, which stands near the circular sepulchral edifice
known as " Tor de' Schiavi " in the Villa of the Gordians on the Via Praenestina near
Rome.

An example showing a further advance, and carried out on a much larger scale,
is afforded by one of the great octagonal halls (Fig. 50) on the south side of the Baths

of Caracal la at Rome (212-216), where,
however, the spherical character of the
pendentives is still not very strongly
marked, but only appears about half-
way up, the lower half forming a re-
entrant angle which continues the lines
of the walls on which the pendentives
rest.

Whether the two kinds of spheri-
cal pendentives, one constructed with
courses of brickwork, as in the case of
the two buildings last mentioned, and
the other made of irregular courses of
lumps of tufa backed by concrete, as
in the case of the " Domus Augustana,"
were ever completely developed by the
Romans, it is impossible to say. My
examination, extending over more than
thirty years, of every possible ruin of
the Roman period in Italy, has not

enabled me to clear up the question. Nor is any further light shed by the drawings
of buildings which have disappeared. Thus, for instance, we know nothing about
VOL. I D

Fig. 47. Rome. Tomb on the Via Nomentana.
Pendentive of Dome (Ilnd Century).

LOMBARDIC ARCHITECTURE

Fig. 48. Rome. Tomb on the Via Nomentana
(Ilnd Century).

these forms of support originated
and pendentives forming part of
a single curve, from the ruins of
Gerasa. It shows, however, an
arrangement exactly like that
of the so-called " Double Gate "
beneath the mosque of
"el-Aksa" at Jerusalem, which
in its turn betrays its relationship
with that of the vaulting of the
" Golden Gate " in the same
place : buildings which must be
ascribed to the Vlth century, or,
to be more accurate, to the reign
of Justinian (527-565). The
period suggested by Choisy,
viz. that of the Early Empire,
while Dieulafoy 3 puts it later
than the Illrd century of the
Christian era, is pure guess-work,
and has no support in facts.

And also what a singular

1 Op. cil.

2 L'art de b&tir chez les Byzantins.

3 Of. fit.

the real construction of the cupolas
resting on pendentives in the
sepulchral structures of which
Montano 1 has preserved the forms,
though he added decorative features
on his own account. Still, we may
observe that, so far as concerns
the first kind, from the pendentives
of the octagonal hall of the Baths
of Caracalla to those of perfect form
of the same kind, is not a long step ;
and it is certain that the builders
of Ravenna boldly employed the
perfect spherical pendentive con-
tinuous with the dome, in a great
cupola such as that of the baptistery
of Neon, between the years 449 or
458-477, before the Byzantines
applied the other variety, in which
the dome and the pendentives
belong to different planes and
curves, to the spacious cupola of
St. Sophia at Salonica (about 495).

There has been an idea that
in Asia. Choisy 2 refers to an example, with cupola

Fig. 49. Rome.

Tomb called " Sedia del Diavolo " on the Via
Nomentana (Ilnd Century).

THE SCHOOL OF RAVENNA

phenomenon so important a
discovery would be and Choisy
in the case of Gerasa regards
it as the result of an Asiatic
conception carried out by a
Roman hand making its
appearance in Syria, perfect
and complete in the days of
the Early Empire, while in that
Empire's decline, that is to
say in the reign of Diocletian
(284-305), those countries,
according to De Vogiie, 1 could
barely show the earliest dated
example, in the chapel of Omn-
es-Zeitun (Fig. 51), finished in
282, of an experimental at-
tempt to set a round dome
on a square base ! In that
experiment the builders, instead
of taking the trouble to find
out scientifically the artistic
manner in which to place a
vault upon a polygonal building,
confined themselves to the
device, both unaesthetic and

F 'g- 5 1 - Omn-es-Zeitun. Pendentive in Chapel
(282).

Fi e- S- Rome. Baths of Caracalla. Octagonal Hall.
Pendentives (212-216).

inartistic, of starting bracket-wise from
the square base the lines of a polygon,
which by gradual multiplication became
assimilated to the circle of the dome.

Choisy mentions other instances at
Sardis, Philadelphia, and Magnesia, in Asia
Minor. But here again the dates are not
known ; and the Roman period, to which
the eminent writer thinks that they belong,
is so uncertain that it cannot be brought
forward to any purpose when we are
comparing one building with another.

It would, moreover, be inexplicable
why the Romans, who were in such direct
contact with Asia Minor and Syria, countries
from which they even got architects,
Apollodorus of Damascus for instance,
laboured for centuries in attempts which
aimed at solving a difficult problem, the solu-
tion of which had already been attained in
those countries, and, in the case of Gerasa,
put in practice by a Roman hand.

1 Op. fit.

D 2

LOMBARDIC ARCHITECTURE

Fig. 52. Kalat Sim-aan. Church of St. Simeon Stylites.
Apse (Vth Century).

the baptistery of Neon (449 or 458-477) close

founded in 450. The merit of the

invention is to be ascribed to the

builders of Ravenna, who hit upon

the idea of combining continuous

rows of small arches, forming as it

were a fringe below the eaves

cornice of a wall, with Roman

lesenas. Such rows of small arches

are a Roman invention, and I

discovered the archetype for them

in the nameless Ilnd century tomb,

proved by its polychrome brick

facing to belong to the age of

the Antonines, which exists at

" Acqua Bollicante," on the Via

Praenestina, near Rome (Fig.

53). Rome, too, at one time

contained instances of them in

the opus sectile decorations of the

basilica of Junius Bassus on the

THE CHAPEL OF SAN PIER
CRISOLOGO was erected by Arch-
bishop Peter Chrysologus (433 or
439~449 or 45 8), 1 as is confirmed by
his monogram in mosaic on one of
the arches of the building.

For us the most notable feature
of this oratory is the external decora-
tion of brick arcading springing from
corbels grouped between lesenas or
pilaster strips (a decorative rather than
a constructive adjunct) ; for though the
upper part has been tampered with,
it was originally decorated with a
course of this kind, the lesenas de-
signed to break it being still preserved.

This architectural innovation,
which is earlier than that of scallop
shells separated by shafts, to be seen
in the apse of St. Simeon Stylites at
Kalat Sim-aan in Syria, believed by
De Vogue to have been erected at the
end of the Vth century 2 (Fig. 52), was
employed at Ravenna, almost at the
same time, not only in the chapel
with which we are dealing, but also in
by, and not far off in San Francesco,

1 Tarlazzi, of. cit.

2 Op. dt.

Fig. 53. Rome. Tomb on the Via Praenestina
(Ilnd Century).

THE SCHOOL OF RAVENNA

Esquiline (IVth century) 12 (Fig. 54), and the mausoleum of Santa Costanza
(IVth century). 3

Fig. 54. Rome. Basilica of Junius Bassus. Inlaid decoration (IVth Century).

In the East, the first ancient example with a fixed date of an arched corbel
course is to be found in a gate in the cloister of the eastern church at Babiska (401).*

The oldest specimen of ^^^^^ ^^^^

continuous rows of
arches formed into
scallop shells, not as
yet divided by sup-
porting shafts as in the
above -mentioned
church of St. Simeon
Stylites, is to be found
in the existing apse of
the church of Arshin,
certainly of the Vlth
century. 5

The motive of the
cornice of arcading
springing from corbels,
or from corbels alter-
nating with or in
groups between lesenas,
passed at a later date
to the Comacine gilds,
and by their means
became a strikingly
characteristic feature
in Pre-Lombardic and
Lombardic architecture

alike.

Fig. 56. Ravenna. Baptistery of Neon (449 or 458 477).

THE BAPTISTERY OF NEON OR SAN GIOVANNI IN FONTE was originally a
chamber in the Baths which stood near the Cathedral. Archbishop Neon (449 or
458-477) remodelled it as a baptistery, and added the decorations. 07

1 Ciampini, Vetera moniinenta.

2 Volume of drawings by Giuliano da Sangallo in the Vatican Library.

3 De Rossi, Mosaic! cristiani delle chiese di Roma antcriori al secolo XV.

4 s Butler, Publication of an archaeological expedition to Syria in 1899-1903. Architecture of Northern Syria
and the Djebel Hauran.

* Man. Germ. Hist.Agiiellus, Liber pontificalis. * Fabri, op. fit.

38 LOMBARDIC ARCHITECTURE

In plan it is an octagon with niches opposite to one another recessed in four of
its sides. The interior has two tiers of round arches, one above the other, springing
from columns (Fig. 55). In the upper tier every arch encloses a triplet of arches, the
middle and largest one being pierced by a window. From this upper tier the hemi-

Fig- 55' Ravenna. Baptistery of Neon (449 or 458-477).

spherical dome starts. It is composed of a double spiral of terra-cotta tubes fitting
into one another and embedded in mortar, and rests on triangular spherical pendentives,
which, in a horizontal section, follow the curve of the cupola. Its lightness allowed
the architect to reduce the walls of the drum to the very moderate thickness of 2 ft.
2 ins. On the outside the walls, with courses of brickwork, separated by layers of mortar

THE SCHOOL OF RAVENNA

of varying thickness, present high up on each face four large corbel arches, divided
into pairs by lesenas, and crowned by a saw-tooth cornice (Fig. 56).

The baptistery of Neon claims our attention by three important peculiarities.

The first is the dome, measuring about 37 ft. in diameter at the base. The
ancient world affords no instance of so wide a vault constructed of tapering tubes.
This method, peculiar to Ravenna, had not made its appearance previously, except in
the case of the semi-domes of apses, as is shown by those of the Basilica Ursiana and
of Sant' Agata at Ravenna.

The second feature is that of the dome resting on broad triangular pendcntives of
perfect form, constructed of courses of brickwork.

The third is the device of single arches, each of which serves to relieve a triplet
of arches of unequal height, a motive which was employed in later times in the West
as well as in the East.

THE CHURCH OF SAN FRANCESCO, begun in the year 450 by Archbishop Peter
Chrysologus, and dedicated to St. Peter with the addition of " the Great," was finished
by his successor, Neon. In 1261, it received the
title of San Francesco. 123 A rebuilding, begun
in 1793, spared only a portion of the original
side walls, the crypt, and the bell-tower.

High up on the outside of the original
south wall of the nave runs a large arched
corbel course resting on simple terra-cotta
brackets, and with a lesena marking off and
dividing each pair from the next.

The crypt is a later addition, as the
materials taken from older buildings which
are used in its construction, testify. Its date
must be the same as that of the campanile,
that is to say, the first years of the Xlth
century, as we shall see when we come to Sant'
Apollinare \uovo, a period in which its Ionic
capitals, the pulvins made to fit their places,
the construction of the vaulting, and the wall
piers of the crypt itself, find their proper place.

The campanile (Fig. 57) is not of the
same date as the church. In fact, it was built
at the expense of the south aisle. Its brick-
work is different from that of the original
parts of the church. Shafts which have come
from elsewhere and every kind of pulvin are
used in its windows.

F'g- 57- Ravenna. San Francesco.
# * * Campanile (Xlth Century).

When the impotence of Romulus Augustulus (475-476) and the valour
of Odoacer 476-493) had brought about the extinction of the Western Empire
and the creation of the first Kingdom of Italy, Ravenna, which since

1 Tarlazzi, op. (it.

2 Rubeus, Historiarum ravennatum, e-

3 Fabri, op. fit.

4 o

LOMBARDIC ARCHITECTURE

the death of Valen-
tinian III (455) had
only afforded its
craftsmen a restricted
field of employment,
was enabled to pro-
vide them afresh
with a very wide one
under the influence
of Theodoric (493-
526).

That illustrious
ruler, who laboured
so strenuously to make
his kingdom appear as
a continuation of the
Roman Empire, em-
bellished the capital of
the Kingdom of Italy
with remarkable
buildings to whose
splendour testimony is
borne alike by the
historians and by ex-
isting monuments.
The most famous of
these structures, the

royal palace, the general appearance of which is represented on a mosaic in

Sant' Apollinare Nuovo, a magnificent edifice surrounded by colonnades and adorned

with the most precious marbles and mosaics, has disappeared. The building which

now goes by the

name of the Palace

of Theodoric is a

structure of later

date, belonging, in

all probability, to

the early years of the

VI I Ith century. The

buildings, however,

which survive are

sufficient to give an

idea of the conditions

of architecture at

Ravenna in the days

of '""heodoric.

Fig- S9- Constantinople. So-called Cistern of Arcadius (Vlth Century).

THE BASILICA

OF SANT' APOLLINARE NUOVO was erected by Theodoric about the year 519,
and dedicated to St. Martin. Owing to its gilded ceiling it was known as " in

Fig. 58. Ravenna. Sant' Apollinare Nuovo (about 519).

THE SCHOOL OF RAVENNA

Fig. 60. Ravenna. Sant' Apollinare Nuovo. Ambon
(Vlth Cenlury).

coelo aureo." 123 In the XVIth

century, as it was sunk considerably

below the surrounding level, the

columns of the nave were raised

and the arcades rebuilt.

It consists of a nave and two

aisles, the former terminating in an

apse which is not original. Ex-
ternally the nave is decorated with

blank arcades surmounted by a

double saw-tooth cornice.

The bulky Corinthian capitals

in the nave (Fig. 58), inscribed

with masons' marks in Greek letters,

and carrying pulvins of the Raven-

nate type, are Byzantine in style and

to be ascribed to a Greek hand, an

origin which would be confirmed by

Fabri's 4 statement that Theodoric

brought from the Greek capital the

twenty-four columns intended for

the nave of St. Martin's. They are closely related to those in the Cistern-basilica

and in the so-called Cistern of Arcadius (Fig. 59) at Constantinople.

These capitals were
not the only ones at
Ravenna to be imported
from the East in the
reign of Theodoric.
For they were Greek
carvers who also pro-
duced the Composite
capitals with protuber-
ant foliage of the wild
acanthus, boldly under-
cut so as to produce
strong contrasts of light
and shade, and pitted
with an endless num-
ber of small holes
made by the drill along
the ribs of the leaves,
four of which, bear-
ing the monogram of

1 Man. Germ. Hist. Af-
nellus, Liber pontijicalis.

2 Muratori, Kcnim italica-
nim serif tores. Spiciltgitun
ravennatis historiae.

Fig. 6 1 - Ravenna. Cathedral. Ambon of Archbishop Agnellus * Tarlazzi, op. cit.

(Vlth Century). 4 Op. cit.

4 2

LOMBARDIC ARCHITECTURE

Fig. 62. Salonica. St. Sophia. Ambon (about 495).

Fig. 63. Constantinople. St. Sophia. Screen (532-537).

THE SCHOOL OF RAVENNA

Theodoric, may be noticed in the ancient colonnade in the Piazza Vittorio
Emanuele II.

To a Greek chisel must also be ascribed the old ambon of Sant' Apollinare
Nuovo (Fig. 60), the panels of which do not exhibit the favourite motive of the
carvers of Ravenna for enriching
liturgical furniture of this sort, viz.
squares with figures of saints, ani-
mals, birds, and fishes (Fig. 61), but
only crosses standing on discs, a
design sometimes used at Rome
in mosaics, as may be seen in Santa
Sabina (Vth century), and also
lozenges with flowers at the angles.

The Byzantines preferred to
ornament their ambons with niches
occupied by human figures after the
fashion of the two fragments from

Fig. 64. Rome. Crypt of St. Peter's. Pluteus
(IVth Century).

The typical

Salonica, now in the Imperial

Museum at Constantinople, belonging to the IVth century. Or else they decorated
them with empty niches, lozenges, crosses, and monsters, as may be seen on the
ambon of St. Sophia at Salonica (about 495) (Fig. 62). They rarely made use, under
Ravennate influence, of squares containing fishes, birds, animals, or other repre-
sentations, except in the case of screens such as that still standing in the women's
gallery of St. Sophia at Constantinople (532-537) (Fig. 63). In this screen the
aforesaid representations have been almost erased, probably after the Turkish
conquest.

ornamental treatment of the ambon of Sant' Apollinare

Nuovo is of Byzantine origin, and
the taste and workmanship dis-
played are just like those of the
parapets of the matroneum of St.
Sophia at Constantinople. It shows
a distinct advance beyond those of
the upper gallery of St. Demetrius
at Salonica (Vth century). After-
wards it spread through Italy. In
the time of Pope John II (533-535)
it was used in the low screens of the
choir and presbytery in San Clemente
at Rome, where it takes the form of
framed panels enclosing lozenges,
stars with eight points, discs each of
which contains a cross, and also the
monogram of the Pope surrounded
by a wreath. The stars recall the
Roman motive of a star formed
by two interlacing squares with a
conventional flower in the centre, an example of which exists in a pluteus
(IVth century) in the Vatican "Grotte" (Fig. 64), one of those which once

Fig. 65. Rome. San Clemente. Capital of the time
of Hormisdas (514-523).

LOMBARDIC ARCHITECTURE

Fig. 66. Ravenna. Sant'
Apollinare Nuovo. Cam-
panile (850-878).

flo w e r,

or, again, with capitals the angles
of which are hollowed out. One
of the pulvins bears a Latin mono-
gram which, after examination
in situ, I interpret loliannes (fig.
67), for, among other reasons, in
such monograms the two most
striking letters (here I and O)
often indicate the whole name.
Built into the spandrels of the
arches of the third tier of
openings counting from the top,
towards the south-east are terra-
cotta bowls (" ciotole ") (fig. 68).
This is the oldest specimen to
my knowledge of window open-

1 Sarti e Settele, De Vaticanis cryptis.
Appendix.

connected the pedestals of the vine-wreathed columns before
the Confession in St. Peter's. 1

These low screens, the work of a Roman or Ravennate
hand, are not to be confused as to either date or authorship
with the transennae of basket-work design, also to be seen
in San Clemente. These latter, together with an inscribed
epistyle, a fragment of architrave carved with vine-
branches, foliage, and bead and reel ornament, and two
small columns once belonging to the altar erected in the
time of Pope Hormisdas (514-523), decorated with twining
ivy and surmounted by basket-shaped capitals (Fig. 65),
and closely related to capitals in St. Demetrius at Salonica,
the Duomo of Parenzo (535-543), and St. Sophia at Con-
stantinople, must be attributed to Greek carvers on account
both of the designs which are characteristic of that school,
and also of the technique peculiar to it in the Vlth century.
To an Eastern chisel must also be assigned a Composite
Byzantine capital in Santa Maria in Cosmedin at Rome,
founded at the beginning of the Vlth century.

But if the craftsmen who carved the capitals and
the ainbon of Sant' Apollinare Nuovo were Greeks, its
masonry and the characteristic decoration of blank
arcading show that its architects and builders belonged to
Ravenna.

The round campanile (Fig. 66) with its wooden roof,
rising in front of the right aisle of the church, has the shafts
(taken from older buildings) of its double and triple open-
ings surmounted by plain pulvins, presenting across between
leaves, or else ornamented with small capitals of the open
lotus

fry- 67. Ravenna. Sant' Apollinare Nuovo. Campanile.
Monogram of Johannes from a rubbing (850-878).

THE SCHOOL OF RAVENNA

ings decorated in this manner. It was inherited by the Lombard gilds, and by the
Greek builders who in the Xlth century employed it in St. Theodore at Athens

(1049).

The tower was not built at the same time as the rest of the church, a fact
revealed not only by the walls, which show courses of brickwork thinner, and layers of
mortar generally thicker, than those of the church itself; but also by the fragmentary
nature of the materials used in it. If this be so, to what period ought we to ascribe
it ? The answer is,
the IXth century,
and, more precisely,
the episcopate of
John, who filled the
see from 850 to
878, and, according
to Gams, 1 was the
tenth of the name,
though if we follow
the chronology of
Giani, 2 he was only
the eighth. The
history of the church
is intimately con-
nected with this
prelate on account
of his having trans-
lated, or, to be more
accurate, made it
appear that he had
translated, the body
of the patron saint,
about the year 856,
from Sant' Apolli-
nare in Classe out-
side Ravenna to the
present church, in
order to put it out
of reach of the raids
of the Saracens, who
penetrated into the
Adriatic in 829, and

had already plundered the church of Classis of its rich ornaments and treasures.
In consequence of this translation, the church began to be designated Sant' Apollinare
Nuovo. 3 So that the monogram which we noticed above should in all probability be
referred to this prelate, and will give the date of the campanile.

The age of this tower being practically certain, we are in a position, by the help

of legitimate inference and mutual comparison, to fix more accurately than has

hitherto been done the dates of the other ancient bell-towers at Ravenna, all of them

later in origin than the churches to which they belong. This statement is based on the

1 Op. cit. " Of. fit. * Fabri, op. at.

Fig. 68. Ravenna. Sant' Apollinare Nuovo. Campanile. Window and
" Ciotole " (850-878).

LOMBARDIC ARCHITECTURE

fact that the masonry of all these churches of the Vth and Vlth centuries (except
San Vitale) is not of the same date as that of the towers belonging to them ; and that
the towers of San Vitale, though contemporary with the church, were erected in order
to provide access to the galleries, and not to hold the bells.

The oldest churches constructed by the builders of Ravenna in the Romano-
Ravennate or Byzantino-Ravennate styles, had no large bell-tower attached to them.
This is proved by the cathedrals of Parenzo (535-543) and Grado (571-586), which at
the time of their original erection were clearly without such adjuncts. It is also
confirmed by San Vitale at Ravenna (526-547), where the towers were built to hold
stairs ; by San Lorenzo at Milan (Vlth century), where the four angle towers were
erected for purposes of communication, and also for constructive reasons ; and by

Santa Maria at Pomposa (Vlth century), where the original

^^^^^^^^fe tower was merely a lighthouse. I note here that the

I ^r ^^k towers belonging to the facades of San Vitale and San

mf ^B Lorenzo Maggiore may have been suggested by the stair-

gl If case towers in the facades of Roman Baths such as those

Mof Agrippa rebuilt by Hadrian (120-124), or those of Nero
remodelled by Alexander Severus (about 228), or lastly
those of Titus (80). In any case there was no occasion for
their builders to trouble themselves to look for ideas in the
East, for instance in St. Sophia at Salonica (about 495),
where the narthex was flanked by a staircase tower on the
north-west, at a later date increased in height.

It is quite true that the liturgical use of bells is of
great antiquity, going back at least to the Vth century ;
but at first they were hung in modest erections of masonry

w MB* I or woodwork, only just rising above the roofs of the

churches and built up from the main walls. The origins

of the great bell-towers of square or circular plan do not

^L ^HH^^K go back to such a remote period as most people fancy.

The Greek churches were without them as late as the
division of the Empire (1204). Moreover, it seems that
the use of bells did not begin among them till after a

number were sent by the Doge of Venice, Orso Partecipazio, to the Emperor
Basil I (867-886) some time between 877 and 88 1 : " Dominus quidem Ursus dux
efflagitante Basilio imperatore eo tempore XII campanas Constantinopolim misit,
quas imperator in ecclesia noviter ab eo constructa posuit, et ex tempore illo Greci
campanas habere ceperunt." * Their use cannot have been widely extended, seeing
that at Constantinople in 1200 St. Sophia itself was without them : "On n'a pas de
cloches a Sainte-Sophie, mais un petit battoir hagiosidere a la main avec lequel on
frappe pour les matines .... c'est d'apres les preceptes de 1'ange qu'ils ont ce
battoir." 2

If, again, Syria provides very early examples of churches with one or two large
square towers, either incorporated with the facade or flanking the aisles, as in the
basilica of Tafkha (IVth and Vth centuries), and the churches of Turmanin
(Vlth century) and Kalb-Lauzeh (Vlth century), in which the narthex is confined

1 Man. Getnt. Hist.JohannisDiaconichronicon Venetum et Gradense.

2 De Khitrowo, Itiniraires russes en Orient. Antoine archev. de Novgorod, Description des Lieux Saints de
Constantinople (1200).

Fig. 69. Rome. Plan of the
Secretarium Senatus (284-305).

THE SCHOOL OF RAVENNA

between the towers as at Sant' Apollinare in Classe (533-549), such towers were
not intended for bells, but for purposes of communication, and contained staircases.
Communication was also the object of the constructions which close the narthexes

Fig. 70. Baalbeck. Ruins of Temples (Ilnd and Illrd Centuries).

of some of the churches at Binbir-kilisse, the dates assigned to which by Strzygowski *
are wrong. 2 3 Possibly even, some of these constructions were not necessarily towers
but simple chambers, as appears to be the case in the " Secretarium Senatus " in the
Curia of Diocletian at Rome 45 (fig. 69).

1 A'leinasien.

2 Ramsay, Sttuiies in the History and Art of the Eastern Provinces of the Roman Empire. Report on
Exploration in Fhrygia and Lycaonia.

3 Revue ArchMogique, 1906, 1907. Bell, Notes on a Journey through Cilicia and Lycaonia.

4 Lanciani, Forma Urbis Romac.

5 Hiilsen, // fora Koniano.

4 8

LOMBARDIC ARCHITECTURE

It was the Latins who introduced lofty bell-towers and bells into Syria. Mariti 1
gives King Godfrey de Bouillon (1099-1100) the credit of the first introduction of the
latter into Jerusalem, and furnishes interesting information about their use in the
Eastern Church and among the Christians of other denominations subject to Moslem
rule.

In those countries the scheme of a church facade with towers is connected with
and is a survival of a pagan idea of which the imposing ruins of Baalbeck offer

remarkable examples. Thus the
Propylaea of the largest temple of
Heliopolis, that of Jupiter, are flanked
by two towers, the remains of which
exist, which originally hardly rose
above the height of the pediment of
the archway of the Propylaea. The
temple of Bacchus (the dedication of
which is revealed by the sculptured
representations on the flight of steps
in the elevated sanctuary, below which is a corridor or crypt with a barrel vault)
has a fagade with two towers taken out of the angles of the building, and
containing stairs which give access to the roof. The chronicler John Malalas 2
states that the great temple of Jupiter was built by Antoninus Pius (138-161).
" He built at Heliopolis a great temple of Jupiter which was a marvel." But the
erection of a group of buildings of this magnitude (fig. 70) must have taken too long
a time for us to be able to accept the statement without some confirmation. On
the other hand, it is far^more probable that the works carried out by him were
confined to the temple properly so called ; and that the great square court, the
hexagonal court with the Propylaea leading to it, and also the temple of Bacchus,
were erected by his successors, especially, to judge from the coins, Septimius Severus
(193-21 1), Caracalla (212-217), underwhom a well-known inscription tells us that the

Fig. 71. Coin of Caracalla (212-227).

Fig. 72. Coin of Philip the Arabian (244-249).

Fig- 73- Coin of Otacilia (244-249).

Propylaea were in course of construction, and Philip the Arabian (244-249), all of
whom gave themselves credit'therefor on the coins of theColonia Heliopolitana. The
works must have been finished under the last emperor, for while a coin of Caracalla
(fig. 71) struck at Abila, mentioned by De Saulcy, 3 bears on the reverse a hexastyle
temple flanked by two battlemented towers (which proves that in his reign some
great temples were flanked with such towers, both for decorative reasons and as
staircases), two coins of Philip (fig. 72) and of his wife Otacilia (fig. 73) show instead
the Propylaea and its towers with a flight of steps leading up to it.

1 Viaggipcr Pisola di Cipro e per la Saria e Palestina. - Corpus script, hist. byz.Chronog>aphia.

3 Nmnismatique de la Terre-Sainte.

THE SCHOOL OF RAVENNA

The history gathered from the coins is confirmed by the story told by the
sculptures to an eye trained in such matters. These sculptures inform us that the
temple ol Bacchus is later in date than that of Jupiter ; and further, that the
great square court of the latter is more recent by some years than the temple to
which it gives access, for we find in it the motive of acanthus leaves with their
tips arching over. The same feature may be noticed on the capitals of what is
believed to be the temple of Venus (Fig. 74) close by, and enables us to fix its
date at the end of the Ilnd or the beginning of the Illrd century, in other words,
in the reign of Septimus Severus. A remarkable example of it is also afforded
by the remains of the famous colonnades of Palmyra (about Illrd century) (Fig. 75).

Fig. 74. Baalbeck. Supposed Temple of Venus (Ilnd or Illrd Century).

which exhibit capitals the exact counterparts of those in the temples of Bacchus
and Venus at Baalbeck. Again, in North Africa, the scheme of the church facade
with staircase towers (e.g. the great basilica of Morsott in Algeria) was suggested
by pagan prototypes. The mosaics, in particular, discovered in the Roman villa
of El- Alia belonging to the Ilnd century, and now transferred to the Bardo Museum
at Tunis, represent villas flanked by towers which contained the state apartments
of the building (Fig. 76).

The use of bell-towers in the facades of Western churches, on the other hand,
began, so far as we can infer from the monuments, if not with the front of the atrium
of the Constantinian church of St. Peter at Rome, provided with two towers by
Popes Stephen II (752-757) and Hadrian I (772-795), then with St. John Lateran,
the northern facade of which was decorated from early times with two bell-turrets
continuing the already existing staircase towers. Or it may have come from
the church of St. Martin at Tours in the reign of Charles the Simple (893-929),

VOL. I E

LOMBARDIC ARCHITECTURE

which, to judge by the representation on a coin of his period, was provided with
a central tower surmounted by a cross, and two towers in the facade.

As bearing on such towers, I mention in passing that the Senate House (Curia)
of Diocletian, belonging to the first years of the IVth century, 1 was designed with
two of these adjuncts for purposes of communication, a fact which anyone can
still verify.

It was quite in the natural order of things that to Italy should be assigned
the task of diffusing as from a centre the conception of fagade bell-towers Italy,
the birthplace of the great campaniles, forming part of the structure of a church
or rising close beside it. In the forefront stand the tower of Santa Maria clella

Fig. 75. Palmyra. Ruins of Colonnades (about Illrd Century).

Cella at Viterbo (IXth century), and the "Torre dei Monaci" of Sant' Ambrogio
at Milan (789-824). If earlier examples are adduced, the dates assigned to them
are arbitrary. Thus, for instance, the great bell-tower which rises beside Santa
Maria Maggiore at Naples, founded by Bishop Pomponius (5 I4-532), 2 with its brick
construction (except in its lowest portion, where fragmentary materials have been
used) and high pyramidal roof also of brick, and its four one-light round-headed
windows, the bell-chamber having four two-light openings, one of which has been
walled up, is certainly not to be assigned to the beginning of the Vlth century
and the agency of that prelate, as some imagine, 3 but was built after the year 1000.
As a matter of fact the material of which the Vth and Vlth century ecclesiastical
buildings of Naples were constructed was not brick, but tufa with brick courses, as

1 Hiilsen, Die ausgrabttugen aitf dem Fonitn Ronianum, 1902-1904.

2 Man. Germ. Hist. Gesta episcopontm Neapolitanorum.

3 Napoli nobilissima, 1893. Croce, Sommario critico delta storia dell 1 arte nel Napolelano.

THE SCHOOL OF RAVENNA

5 1

in the arcaded apses of San Giorgio Maggiore (367 and about 387) and San Giovanni
Maggiore (554-577). In addition to which, the tall pyramidal roofs made of masonry
did not make their appearance, so far as I am aware, before the Xlth century. And
the typical corbel pulvins of the two-light openings (that is, pulvins corbelled out
to correspond in length to the thickness of the wall) did not come into use before the
beginning of the second half of the Xth century, as we shall see when we come to deal
with the Cathedral of Ivrea (973-1001 or 1002). Lastly, the three small capitals, made
expressly for the tower of Santa Maria Maggiore, which carry these corbel pulvins,
are products of the artistic revival of the early Xlth century. One need only
look at the two imitations of the simplest form of Roman Composite capital, with
the characteristic
central leaf, stiff
and terminating in
a sharp point, in
order to convince
oneself of the fact.
We have still,
it is true, to reckon
with De Rossi's
statement that, as
early as the first
half of the Vth
century, ecclesias-
tical basilicas had
bell -towers con-
nected with the
front or in the rear
of the building; but

we are very much afraid that in this case the eminent Roman archaeologist was
wrong. 12 What indeed are we to say of the buildings encircled by a wall with towers,
portrayed on the triumphal arch of Santa Maria Maggiore at Rome, and representing
Jerusalem, in which De Rossi saw a basilica with its circular baptistery, and
high bell-towers behind and at the side ; as well as a second basilica with a
similar tower flanking the facade ? Whereas, in fact, there is nothing more than
a fanciful group of buildings, two out of the three towers referred to belonging
to the encircling wall, where at the most one might identify Constantine's
' Martyrion " with its three doors, and the " Anastasis " beside it. And what
are we to think of the so-called Temple of Jerusalem under the form of a Christian
Church, with a cross on its front between two bell-towers, to be seen on one of
the carved panels of the well-known and not less discussed doors of Santa Sabina
at Rome, a church which De Rossi himself says 3 was begun by Pope Celestine I
(422-432) and finished under Xystus III, while Lanciani 4 gives 425 as the
date of the building, and Grisar 5 refers the doors to the year 535? For what
are the facts? A pair of towers placed behind the left side of a conventional
church, and perhaps possessing some symbolical character, but certainly not

Fig. 76. Tunis. Bardo Museum. Mosaic from a Roman Villa (Ilnd Century).

1 Bull, di arch, crisliana, 1887. Campana con epig rafe dedicatoria del secolo in circa ottavo o none, trovata
presso Canino. - J/usaici Crist iani delle chiese di Kama, anteriori al secolo X V.

1 Bull, di arch, cristiana, 1887, Camfana, &c.

4 The Kuins and Excavations of Ancient Koine. * Analecta Koniana.

E 2

52 LOMBARDIC ARCHITECTURE

bell-towers, seeing that one of them has no opening in its upper part where the
bell-chamber ought to be (Fig. 77).

But to return to the campanile of Sant' Apollinare Nuovo, let us see whether
it will help us to fix with tolerable accuracy the dates of the others belonging
to the ancient churches of Ravenna. We may, however, confine our examination
to those of Sant' Apollinare in Classe, San Giovanni Evangelista, and San
Francesco, and not without good reason. For the towers of the original
Basilica Ursiana, of Sant' Agata, of SS. Giovanni e Paolo (the original structure
of which is ascribed to the Vlth century), 1 of San Giovanni Battista (a building

of the Vth century, first consecrated by Peter
Chrysologus and erected by Baduarius), 2 and of
Santa Maria Maggiore (originally built between
521 and S34), 3 are not only all of later foundation,
not one of them exhibiting in its oldest parts the
masonry of large specially made bricks which is a
feature of Vth and Vlth century buildings at
Ravenna, but moreover, owing to the alterations
which they have undergone, or from their compara-
tively recent date, they would not contribute to the
elucidation of the argument.

Above all, it is easy to see that the round
form of tower at Sant' Apollinare Nuovo must
necessarily have been that of the earliest bell-
towers of Ravenna. The local builders in the Vlth
century had chosen that form for the characteristic
staircase towers of San Vitale, and it was very
natural that their successors should find in them the
suggestion for the bell-towers of other churches.
And this is what they actually did, in spite of the
difficulties in the way of sound and exact con-
struction in the case of a cylindrical tower, not to
speak of those connected with the insertion of the
numerous openings necessary to let out the sound
of the bells, and the fixing of the frames to hold
the bells themselves, and with complete disregard of the very imperfect connection
and harmony subsisting between towers of this shape and the rectilineal forms of
the church faades to which they are attached. But this was the way in which
the campanile of Sant' Apollinare arose, and the later ones were made after its
likeness.

Next in chronological order to the tower of Sant' Apollinare Nuovo comes that of
Sant' Apollinare in Classe. The materials of its construction are, like that of the
last, fragmentary ; and its pulvins, specially made for their places, as well as the use
of double recessed arches in the heads of the three-light openings, make one think at
first sight that they are contemporary. The masonry, however, shows that they do
not belong to the same date ; and the band of bricks in two colours arranged like opus
reticulatnm, which ornaments the lower part of the Campanile of Sant' Apollinare in
Classe, points to a later period, when the art of decorating these towers had been

Fig. 77. Rome. Santa Sabina. Panel
of Door (Vth Century).

Fabri, op. fit.

2 Man. Germ. Hist. Agnellus, Liber fontif,calis.
3 Tarlazzi, op. cit.

THE SCHOOL OF RAVENNA

developed. Its foundation, then, must be referred to a time subsequent to the
expulsion of the Saracens from Bari (871), or, more precisely, to the last years of the
IXth century.

The towers of San Giovanni Evangelista and San Francesco come next in order.
The square tower, which avoided the inconveniences connected with the round form,
and tends to produce a line which, artistically, is considered much purer, though
chosen by the Lombard gilds for their great bell-towers from about the end of the
VHIth century onwards, was evidently only adopted with reluctance by the builders
of Ravenna. Tradition was the great obstacle to change. It was only brought about,
after those gilds had first created the prototype of the Lombardic campanile, in the
majestic bell-tower of San Satiro at Milan (876), in my belief the oldest example of
such a structure with the beginnings of a definite architectural design.

The two towers in question should, then, be ascribed to a date later than the year
876. That of San Giovanni Evangelista was probably begun after the erection of
the campanile of Sant' Apollinare in Classe, because we do not find in it the local
fashions of the round towers of earlier date, and also on account of the presence of
carving which indicates a time near the year 1000. Finally, though not later than
the first years of the Xlth century, seeing that in 1063 a campanile in the most
elaborate Lombardic fashion had already made its appearance at the neighbouring
abbey of Pomposa (Vlth century) and the presumption is that one of a much
simpler and more primitive style would not have been built at Ravenna if only
separated from the latter by a few years interval rose the campanile of San
Francesco which, with the bands which frame it, and the lesenas merging in an
arched corbel course at the top, proclaims the Lombardic style and marks an artistic
progress beyond the tower of San Giovanni Evangelista. It must, then, be some years
later, and consequently will belong to the beginning of the Xlth century.

THE MAUSOLEUM OF THEODORIC was erected by order of the second King of
Italy, in his own lifetime, about the year Sip. 12 It is due, in all probability, to
the architect Aloiosus and the " marmorarius " Daniel, to whom there are such
interesting references in Cassiodorus. 3

It consists of two stories built of squared marble blocks laid without mortar,
those which form the voussoirs of the arches having joggle joints (Fig. 78). The
irregular adjustment of the blocks at the points of greatest pressure was intended to
increase the stability of the walls, and enable them to meet the thrust of the cupola,
it being well known that such irregularities resist any tendency towards dislocation
of the parts, so that the masonry retains its cohesion perfectly against both outward
thrust and vertical pressure.

The building rests on a platform of brick and stone set in mortar and cement
(" pozzolana "). The lower story externally forms a decagon, and in each of its sides
is recessed a niche of rectangular plan, with an arched head, except in the one
which contains the square-headed doorway. Internally it has the shape of a cross
with equal arms, and is lighted by loops.

The upper story, also forming a decagon, must originally have been decorated on
nine of its outer faces by a sort of high canopies ; not by a loggia or portico,
encircling it in the manner imagined by Buonamici. 4 The tenth side is reserved, as

1 .Mon. Germ. Hist. Aaonymi Valesiani pars posterior.

- Muratori, Kemm ital. script. Spicilegium Kavennatis hisloriae.

3 A/on. Germ. Hist. t'ariae. * Of. (it.

LOMBARDIC ARCHITECTURE

in the lower story, for the door. Internally it is of circular shape, with very
small windows, one of which is in the form of a cross.

The building is covered by a cupola consisting of a single piece of Istrian lime-
stone, the circumference of which is provided with twelve handles, intended, without
doubt, to lift by means of ropes and drop into its place this wonderful inverted basin.

Fig. 78. Ravenna. Mausoleum of Theodoric (about 519).

I cannot imagine a more ingenious or more practical method of performing the
operation.

It used to be supposed that upon these projections, on the outer faces of which
are engraved the names of the four Evangelists and of eight of the Apostles,
the corresponding statues were placed. But the saddle-backs of the projections
are not suited for supporting statues, nor is there any trace of holes for the clamps
which would have been necessary to fix them in place : not to speak of the fact that
the inscriptions are obviously later than the time of the founder.

It has also been imagined that upon the summit of the cupola rested the
porphyry sarcophagus of the great Gothic king. The only fact, however, that

THE SCHOOL OF RAVENNA

is known about the coffin is that in the IXth century it was lying at the foot of the
mausoleum. 1

The designer of the tomb of Theodoric must have derived his inspiration from
some Roman sepulchral edifice (we can catch from Ennodius - something of the
atmosphere of Romanism which enveloped Theodoric him-
self) such as the one here illustrated in the ground plan of
its lower story, as preserved by Bramantino 3 (Fig. 79).
Or he may have followed the type of one of which
Sangallo has left a sketch in his volume in the Vatican
Library (Fig. 80). This would explain the remarkable
ability displayed in its construction, and the well-pro-
portioned relation of all its parts, which are so striking
that some have thought that it belonged to the best age
of architecture.

Nevertheless, with the exception of the wonderful
monolith cupola, measuring more than 30 ft. in diameter
and about I ft. 4 in. in thickness, and also of the cornice-
band carved with a characteristic meander, the mausoleum does not present a single
new idea either in construction or decoration.

# * #

Fig " 7g '~

The impulse given to the arts by Theodoric was bound to produce new and
abundant results. In fact, after his death, first of all and mainly during the regency

Fig. 80. Santa Maria Capua Vetere. Roman Mausoleum.

of the able and intelligent Amalasuintha (526-535), and afterwards in the reigns of
Theodahad (534-536) and Vitiges (536-540), we see the erection of the most important

1 Man. Germ. Hist. Agnellus, Liber pontificalis.
^ Man. Germ. Hist. Magni Felicis Ennodi opera. * Monger!, op. cit.

56 LOMBARDIC ARCHITECTURE

ecclesiastical buildings which the Italian peninsula can show in the Vlth century:
San Vitale at Ravenna, San Lorenzo at Milan, Sant' Apollinare at Classis, and the
cathedral of Parenzo.

These buildings belong to two distinct styles, the Romano-Ravennate, with
which we have already made acquaintance, and the Byzantino-Ravennate. It has,
indeed, been the general practice to give Byzantine builders the credit of all these
structures. But, as we shall see presently, the actual buildings are there to prove that
Byzantine craftsmen either took no part in the erection of the churches we have just
mentioned, any more than they did in the case of buildings of less importance, though
always interesting for the history of art, such as San Vittore at Ravenna, the abbey
church of Pomposa, the parish church of Bagnacavallo, and the cathedral and the
church of Santa Maria delle Grazie at Grado, all of them Italian works of the second
half of the Vlth century, or else were employed merely in the capacities of carvers
and mosaic-workers.

THE CHURCH OF SAN VITALE AT RAVENNA. The erection was entrusted to
Julianus Argentarius by Archbishop Ecclesius (522-532) after his return from

Constantinople (525),
^hither he had been
sent (524) byTheodoric,
together with Pope John
I (523-526) and other
bishops 1 in other
words in the year 526.
" Incoatio vero haedifica-
tionis ecclesiae parata
est ab luliano, postquam
reversus est praecl ictus
Ecclesius pontifex cum
lohanne papa Romam
de Constantinopoli. . . ." z
The founder, then, was
not Justinian. If he had
been, the fact would
surely have been men-
tioned by Procopius.

The building was
not finished till after the
surrender of Ravenna
to Belisarius (540), and
it was consecrated by
Archbishop Maximian

Fig. 81. Ravenna. San Vitale (526-547).

(546-556) in 547. This
took place under Im-
perial patronage, for we may be sure that it was the offerings of Justinian and
Theodora which paid for the mosaic decorations of the sanctuary and, probably, for
the construction of the vaulting over the aisle.

1 Muratori, Annali tf Italia.

2 A/on. Germ. Hist. Agiiellus, Liber pontificalis.

THE SCHOOL OF RAVENNA

Its plan is that of a regular octagon. The interior (Fig. 81) is encircled by an aisle
and gallery, interrupted by the presbytery, and supported by eight massive piers,
the intervals between which, with the exception of the one opening into the chancel,
are filled by semicircular exedras with open arcades, after the fashion of the arcaded
apses of the early Christian
period, as in the church of
San Sebastiano outside the
walls of Rome (366-384) and
the Basilica Severiana at Naples
(367 and about 387). Above the
piers rises the central conical
dome carried at the angles by
pendentives shaped as niches
(Fig. 82), which enable the cen-
tral octagon to pass into the
circular drum which forms the
base of the dome itself. The
latter is masked externally by
the walls of the drum \\hich rise
above it.

The aisle and gallery are
covered with cross vaults. In
the latter may be seen the trans-
verse arches connecting the piers
which carry the dome with the
internal buttresses at the angles
of the building. The vaulting
of the aisle below is not of a
piece with the rest of the struc-
ture, but was added to replace
the original wooden ceiling,
probably before the works were

finished, for the vaulting of the ground floor is constructed of the same materials
and in the same manner as the rest of the building.

The deep apse, curvilinear internally and semi-hexagonal externally, is flanked
by two chambers ending in niches, and by two sacristies with rectangular projections.
The presbytery is covered by a raised cross vault (" crociera di sesto rialzato "\
concave-crowned, i.e. ending in a kind of cap (" calotta "). The object of this
arrangement, of which we have met with no instance earlier than San Vitale, was
to relieve the pressure of the vault, and at the same time to strengthen it in its
weakest point, and provide a better surface for the display of the mosaics which form
its decoration.

Opposite to the apse there opened originally an imposing rectangular narthex
with semicircular exedras facing one another at either end, after the Roman fashion.
Only its shell has been preserved. In contact with it were two round towers, one of
which still retains its conical vault, constructed in exactly the same way as the dome
of the mausoleum of Galla Placidia. They contained the spiral stairs which formed
the communication between the vestibule and the gallery. In later times one of
them was built higher in order to turn it into a bell-tower : the other still retains

Fig. 82. Ravenna. San Vitale. Pendentive of the Dome
(526-547).

58 LOMBARDIC ARCHITECTURE

its form though deprived of its stairs. The narthex was approached through an
atrium or cloistered fore-court.

The building is entirely constructed of brick. The walls, which at the base have
a thickness of about 3 ft., are formed of courses of large bricks separated by layers
of mortar of varying thickness, and finished at the top by a saw-tooth cornice. A
similar cornice marks the division between the two stories of the interior. Substantial
buttresses, measuring about 4 ft. x 5 ft., at the outer angles of the walls, strengthen
the internal ones, and receive the last thrust of the transverse arches which help to

Fig. 83. Ravenna. San Yitale (526-547).

keep the piers of the dome in place. These buttresses, between which are lesenas
of about i ft. x 3 ft. running right up to the top of the wall, like those that survive
in the neighbouring church of Santa Croce (about 449), and interrupting the eaves-
cornice, not only increase the stability of the outer wall, but also have a decorative
purpose (Fig. 83). Two bracket-like projections stand out at the angles of the wall
above the apse and below the pediment ; a decorative feature which cannot be
paralleled in any building earlier than San Vitale. It is a characteristic motive of
the Romano-Ravennate and Byzantino-Ravennate styles of the Vlth century.

The dome is constructed of two concentric rows of terra-cotta tubes, fitted one
into the other and embedded in mortar, which extend in a spiral form up to the
crown. Its curve, the presence of the spirals, each coil of which resists the thrust
of those above it, and, lastly, the fact that the method of construction makes it a

THE SCHOOL OF RAVENNA

homogeneous mass, result in the pressure being almost entirely vertical. This

pressure being relieved by the peculiar material employed, the builders have been

able to reduce the sustaining walls

to a very moderate thickness ; and

the walls themselves, as we have

mentioned, are raised higher than the

base of the dome in order to give

them additional weight and provide

more resistance to the thrust of the

dome itself.

The dome is protected nowa-
days by a pyramidal wooden roof,
which is not of any great age. The
original roof was perhaps formed
of a framework of rafters and boards
covered with thick and broad sheets
of lead, like the roof over the vault of
the apse in the old Basilica Ursiana
at Ravenna.

At first the church was lighted
by very large, round-headed, un-
splayed windows. Those which ap-
pear on the eight sides of the drum
are in every case framed in an arch
recessed within an outer one, which

There is also a large window divided into three

Fig. 84. Ravenna. San Vitale. Capital (526-547).

has a decorative purpose.

lights by small piers carrying pulvins of the Ravennate type, the whole en-
closed by a large arch formed below
the pediment of the eastern end of the
church.

In the five free sides (i.e. those which
do not correspond to the apse and the
narthex) there opens a door relieved by a
triangular arch.

In the interior, the columns on the
ground floor stand on stepped bases, and
are surmounted by cubical funnel-shaped
capitals with the four sides slightly con-
vex and elaborately carved, and carrying
pulvins of the Ravennate type (Figs. 84,85).
There are also some Composite capitals,
like those of the gallery. The capitals of
the columns in the latter, on the other
hand, are either Composite, supporting
pulvins (Fig. 86), or else funnel-shaped
capitals, or, lastly, those of the melon form
(Fig. 87).

The cubical capitals of San Vitale were the first of their kind to be seen in Italy,
and are the work of Greek chisels, as is shown by the Greek letters forming masons'

Fig. 85. Ravenna.

Capital (526-547).

LOMBARDIC ARCHITECTURE

Fig. 86. Ravenna. San Vitale. Capital (526-547).

marks which appear on some of the pulvins and also on some of the columns of the

ground floor.

The Byzantine cubical capital, of quadrangular funnel shape with convex sides

and the angles rounded off, sometimes
lobed like a melon, was modelled by the
Byzantines partly on the Roman funnel-
shaped cubical capital like those of the
Porta Nigra at Trier (believed to belong
to the second half of the IVth century,
or, more precisely, to the reign of Valen-
tinian I [364-373], under whom the city
attained a new splendour), 1 and partly
on the Ravennate pulvin ; certainly not
on a capital of the well-known Sassanid
type from Ispahan (for which see Dieula-
foy), 2 as some think, for the form in
question has much closer affinity with
the Gallo-Roman and Ravennate types
referred to than with those of Persia. Its
simplest form is to be seen in the cistern
of Binbir-direk " of the 1,001 columns "
at Constantinople (Fig. 88).

The principal forms derived from it
are: (i) The simple funnel-shaped type,
the earliest specimens of which are to
be found in St. Sophia at Salonica
(about 495), San Vitale at Ravenna
(526-547), and SS. Sergius and Bacchus
at Constantinople, founded after Just-
inian had assumed the Imperial diadem, 3
as is made clear by the inscription, con-
sequently not before 527, that is to say
at least a year later than the beginning
of the works at San Vitale in Ravenna.
(2) The funnel shape with volutes at
the upper angles a reminiscence of
the Ionic capital. The earliest examples
are found in St. Sophia at Constantinople
(Fig. 89). (3) The melon shape, of
which the prototypes are provided by
St. Demetrius at Salonica (Vth century),
San Vitale at Ravenna, and SS. Sergius
and Bacchus at Constantinople (Fig. 90).
The Composite capitals in San

Fig. 87. -Ravenna. San Vitale. Capital (526-547). Vitale, though they betray the Byzantine

manner of the age of Justinian, are not
inscribed with any Greek carvers' marks, and may be ascribed to Ravennate sculptors

1 Browerus, Antiquitates et Annales Trevirenses. 2 Op. fit.

3 The Bitilder, January 6, 1906. Henderson, SS. Sergius and Bacchus, Constantinople.

THE SCHOOL OF RAVENNA

imitating Eastern fashions. For when Italians, brought up in the local schools,
wanted to produce capitals of the Byzantine kind, they were not servile imitators,
but impressed a peculiar character on their work, together with a certain tendency
towards the classical type. We shall not find it difficult to accept this statement
when we examine the Corinthian capitals in the church of the Spirito Santo at
Ravenna (Fig. 91), believed to be of the age of Theodoric i 1 others of the same
sort in San Martino ai Monti at Rome, founded by Pope Symmachus (498-514) and
completed by his successor Hormisdas (514-523) ; 23 and, lastly, those of the gallery

Fig. 88. Constantinople. Cistern of liinbir-direk (Vlth Century).

at the far end of San Lorenzo outside the walls of Rome, one of the results of the
works of restoration carried out there by Pelagius II (579-590).*

It is the firmly rooted belief of most writers that San Vitale in its entirety, or
nearly so, was the work of Byzantine builders. The plan of the church being quite
new to Italy would support this idea, for a new style of architecture cannot spring by
magic out of nothing ; and, on the other hand, it is well known that the vaulted
basilica of the central type took shape in the Greek Empire, and became the typical
Eastern church, as being best suited to the Eastern character.

Nevertheless, in my opinion the course of things was somewhat different. The
Byzantine vaulted basilica, as it appeared in the time of Justinian I (527-565), was the
result of a gradual but tolerably rapid evolution. Choisy 5 locates its birthplace in the
western part of Asia Minor, in Ionia. My belief, on the other hand, is that it
originated in Macedonia, with the aid of some influence from Ravenna, and, more

1 Tarlazzi, of. fit. * Duchesne, Le liber pontificalis.

3 Mazzanti, La scoltura ornamental* romana nei bassi tempi.

4 Bull, di arch, cristiana, 1864. De Rossi, Le due basiliche di San Lorenzo nelf agro Verano.

5 L'art de b&tir chcz les Byzantins.

LOMBARDIC ARCHITECTURE

precisely, at Salonica, which, even after the foundation of Constantinople (328),
continued to be the real capital of Greece, Macedonia, and Illyria. Ionia can claim
little more than the honour of having produced the two famous architects of
St. Sophia at Constantinople as rebuilt by Justinian I in 532 and consecrated in S37, 12
that is to say, of the highest expression of the Byzantine style.

The first link in the chain which connects the Roman basilica system with the
Byzantine is to be found in the church of Eski-Djuma at Salonica (Vth century).

Fig. 89. Constantinople. St. Sophia (532-537).

Here are used Ionic capitals with pulvins, that is to say a pulvin of the Ravennate
type supported at the angles by volutes intended to conceal the abruptness of the
transition from the square of the pulvin to the round. This new Byzantine kind of
capital, magnificent specimens of which are to be seen in the galleries of St. Sophia at
Salonica (about 495) (Fig. 92), and in SS. Sergius and Bacchus (about 527) and
St. Sophia at Constantinople (532-537) (Fig. 93), made its first appearance in this
basilica of Eski-Djuma, and became a prominent feature of buildings of known date
in other provinces of the Byzantine Empire, though only about the time of Justinian.

1 Corjnis script, hist. byz. loaunis Zonarac epitoutae historianun.

2 Alon. Germ. Hist.- Chronica miiiora. Vol. II. Marcellini coniitis chronicon.

THE SCHOOL OF RAVENNA

The second link is provided by the basilica of St. Demetrius at Salonica (Vth
century), where quadrangular piers break the ranges of columns in the two stories. Here,
too, three new types of capital are used for the first time : the cubical Byzantine melon
form ; the Byzantine Corinthian, with leaves blown by the wind in opposite directions ;
and the Byzantine bird and basket
capital, derived from the Byzantine
Composite with birds taking the
place of volutes, which in its turn
was derived from the Roman Com-
posite capital with birds and animals
supporting the abacus.

The third link in the chain is
the most important of all. Its age
is no matter of hypothesis, like the
churches of Koja Kalessi in Isauria
and of the Trinity at Ephesus, them-
selves important monuments of the
transition from the Roman to the
Byzantine basilica, but has a date
which may be regarded as certain
on the strength of an inscription
alluding to the decoration of the
church. This link is to be found
in the basilica of St. Sophia at
Salonica (about 495), which, though
reduced by a recent fire to the
miserable condition that it presents
to-day, retains enough to make it
a monument of the very first rank,
in which, as ChoHy says, 1 " we find
the typical structure summing up
a whole system of methods, of
which St. Sophia at Constantinople
(S3 2 -S37) offers the grandest and
most perfect expression."

The whole building, with the
exception of the gallery, which has
a wooden ceiling, is covered with r

tig. 93. Constantinople. SS. Sergius and Bacchus.

barrel, domical, and unraised vaults. Capital (about 527).

The square space in the centre

develops into an unraised, spherical dome, provided in its lowest part with a railed

gangway, and resting on spherical pendentives. This is the earliest certainly-dated

example in the Byzantine world of a dome of such size, supported on pendentives

of this kind, constructed with courses of brickwork, belonging to different planes and

curves.

The description just given makes it clear that in St. Sophia at Salonica Byzantine
architecture reached its full development. In order, however, to arrive at St. Sophia
of Constantinople, another link must be added to our chain, and we must look for it in

1 L'art de bAtir chtz Us Byzantins.

6 4

LOMBARDIC ARCHITECTURE

a vaulted basilica, with aisles and galleries, and embodying the feature of exedras
with open colonnades introduced in the space below the central dome. Such a con-

Fig. 91. Ravenna. Spiiito Santo. Capital
(493-526).

Fig. 92. Salonica. St. Sophia. Capital
(about 495).

necting link is provided by San Vitale at Ravenna, begun, as we saw, at least a year

before SS. Sergius and Bacchus.

The points which it has in common with St. Sophia at Salonica are the following.

The women's gallery, originally
designed with a wooden floor ;
the apse, semi-hexagonal ex-
ternally, flanked by two sacris-
ties ending in recesses which
project beyond the main wall
(Fig. 94) ; the line of the in-
ternal galleries indicated by a
saw-tooth course on the exterior ;
and, lastly, the arcaded lower
story.

Its author was undoubtedly
Julianus Argentarius, who has
been made to figure in every
capacity a prefect, a treasurer
of the Church of Ravenna, a
wealthy merchant, a banker, a
money-changer, everything, in
short, except his real character,
viz. an architect of the first rank.
The family of the Argentarii is
F 'g- 93. Constantinople. St. Sophia (532-537). mentioned in an inscription (cited

Fig. 94. Salonica. Plan
of St. Sophia (about 495).

THE SCHOOL OF RAVENNA 65

by Fabri l in his account of the church of San Zaccaria) from a tombstone of
the time of the Emperor Tiberius Constantinus, i.e. Tiberius II (578-582), which
records a " Georgius Argentarius filius Petri Argentarii." Peter and George must
have belonged to the same family as Julian the architect. And he was not the first
ecclesiastical architect of Ravenna ; for, in the preceding
century, the church of San Giovanni Battista, consecrated
by Peter Chrysologus, had been built by one Baduarius :
"Consecravitqueetiamecclesiam sancti loanniset Barbatiani
quam Baduarius haedificavit." - Julian's profession of
architect comes out in the following passages of Agnellus :
" Incoatio vero haedificationis ecclesiae parata est ab luliano
ecclesia beati Vitalis martyris a luliano Argentario .con-
structa est Beati martyris Vitalis basilica, mandante
Ecclesio viro beatissimo episcopo, a fundamentis lulianus
Argentarius aedificavit Beati Apolenaris (basilicam) . . .
mandante vero beatissimo Ursicino episcopo, a funda-
mentis lulianus Argentarius aedificavit lussitque et
ammonuit hie sanctus vir, ut ecclesiam beati Apolenaris ab luliano fundata et
consummata fuisset." His recognised ability in this capacity was also recorded in a
metrical couplet in silver mosaic letters in San Vitale, quoted by the same chronicler :

Tradidit lianc primus luliano Ecclesius arcem,
Qui sibi commissum mire perfect f opus.

Others 34 have noticed these points before me, and have regarded Julian as an
architect, or at least as having a knowledge of architecture, but have not adduced the
ample proofs which I have furnished.

As a matter of fact the two churches designed by Julian betray a common author-
ship. The identity is revealed in the new feature
of the characteristic bracket-like projections at the
top of a wall. But, above all, it is shown by the
systems of resistance adopted in order to counter-
balance the principal internal thrusts, represented
in San Vitale by the central dome, and in Sant'
Apollinare in Classe by the great chancel arch.

From the beginning San Vitale was regarded
as a wonderful building : " Nulla in Italia ecclesia
similis est in aedificiis et in mechanicis operibus." 5
Without doubt Cassiodorus 6 had it and San Lorenzo
Maggiore at Milan specially in mind when he
praised the boldness and lightness of the new style
of building: "Quid dicamus columnarum iunceam
proceritatem ? Moles illas sublimissimas fabricarum
quasi quibusdam erectis hastilibus contineri et sub
tanta aequalitate concavis canalibus excavatas, ut
magis ipsas aestimes fuisse transfusas, ceris iudices factum, quod metallis durissimis

Fig. 95. Ravenna. Plan of San Vitale
(526-547).

Of. fit.

Man. Germ. Hist. Agnellus, Liber ponlificalis.

3 Hiibsch, Die altchristlichen A'ircken nacfi den Baudenkmalen und iiltcren Beschreibungen.

4 Cappelletti, Le chiese cT Italia.

5 Moil. Germ. Hist. Agnellus, Liber pontificalis. " Man. Germ. Hist. Variae.

VOL. I F

LOMBARDIC ARCHITECTURE

videas expolitum, marmorum iuncturas venas dicas esse genitales, ubi dum falluntur
oculi, laus probatur crevisse miraculis."

It was the result of suggestions taken from both Pagan and Christian monuments
of Rome and Ravenna, and also from the East ; and it presents characteristics which

confer on it a claim to constitute a new

Off -^ w ^^ ^ -M style, the style which I describe as

f * *^*9 ^-^' i "Byzantino-Ravennate," and are suffi-

( \ cient to prove that not only its architect,

but also its builders were Italians trained
in the School of Ravenna.

Coming to facts, it is, first and fore-
most, a very rare instance of a purely
octagonal church (Fig. 95). In the
case of the best known contemporary
Byzantine churches still surviving, such
as St. Sophia (Fig. 96) and SS. Sergius
and Bacchus at Constantinople (Fig. 97),
the octagon which carries the dome is
combined with an external wall not of
octagonal but of quadrangular form.

In the next place, the essential
origin of its plan did not come from
the great rectangular halls of the Roman
Baths, nor from the Roman vaulted
basilica, as is the case, for instance, with
St. Sophia at Constantinople, the ground plan of which is supposed by some
writers to have been a development from structures such as those shown here in
plan (Figs. 98, 99). Whereas it has, in fact, a marked affinity with the plan of a hall in
the Baths of Agrippa, rebuilt by Hadrian (120-124) (Fig. 100), and with another
in the Baths of Nero, reconstructed by Alexander Severus (about 228) (Fig. 101),

Fig. 96. Constantinople. Plan of St. Sophia
(532-537).

o .

Fig. 97. Constantinople.
Plan of SS. Sergius and
Bacchus (about 527).

Fig. 99. Rome. Bath Room, (from
a sketch by Brainantino in the
Ambrosiana. )

Fig. 98. Kusr en Nueijis. Plan
of a Roman Tomb (Ilnd
Century).

as well as with that of the " Basilica Nova " (Fig. 102), begun by Maxentius (310-312)
and finished by Constantine. 1

On the contrary, San Vitale was modelled on the plan of Christian baptisteries like

1 Lanciani, The Jim'ns and Excavations of Ancient Rome.

THE SCHOOL OF RAVENNA

those of Neon at Ravenna, and of the Lateran at Rome, or else on that of some bath-
room of the type reproduced in Fig. 103 : a polygon with niches recessed in each of

Fig. 101. Rome. Plan
of Hall in Baths of
Nero (al>out 228).

Fig. 100. Rome. Plan
of Hall in Baths of
Agrippa (120-124).

Fig. 1 02. Rome. Plan
of the Basilica Nova
(310-312).

the sides, and apparently decorated with wall-shafts. The drawing by Baldassare
Peruzzi, to which attention has not been previously called, is preserved in the Uffizi

at Florence. Or else it may have been
derived from the plan of some sepulchral
edifice of the form shown in Fig. 104, or,
again, from the Licinian Nymphaeum at Rome,
known as "Minerva Medica " (253-268)' a
suggestion made before me by Isabella 2 -
with the addition, however, of an eight-sided
outer wall to make the construction of an
internal gallery possible. From this decagonal
hall (rearranged it
seems in the IVth
century) the archi-
tect of San Vitale
also borrowed the
form of the narthex,
and the idea of the
buttresses at the
angles of the poly-
gon (Fig. 105).

This last deriv-
ation makes it easy
to explain the
family likeness
which Choisy 3

found between the church of SS. Sergius and
Bacchus at Constantinople and the Licinian
Nymphaeum, for the church is in its turn

Fig. 104 Rome. Plan of
a Tomb, (from Sir Ho,
" De it antiquitA")

Fig. 103. Rome.
Bath Room.

1 Lanciani, The Ruins and Excavations of Ancient Rome.

2 Les Edifices circtilaires et les dimes.

3 L'art de bAtir chez les Bvzantins.

F 2

LOMBARDIC ARCHITECTURE

derived from San Vitale at Ravenna. There can be no doubt that the plans
of the latter were known to the architect of Justinian's building.

Fig. 105. Rome. Licinian Nymphaeum. " Minerva Medica" (253-268).

Fig. 106. Constantinople. SS. Sergius and Bacchus (about 527)-

In the next place, the narthex is of a characteristic Roman type, with a niche at
either end, and has no analogy in any earlier Byzantine church. Then, the arrange-

THE SCHOOL OF RAVENNA

ment of the two tiers of arcades, opening out of the exedras, is clearly derived from
the internal arrangement of the baptistery of Neon, where we see all round, arches
alternately framing the

four niches and decor-
ating the four wall-
spaces, with, above
them, a second tier of
large arches, each en-
closing three smaller
ones, and forming the
base of the dome. The
derivation must be
obvious to any eye
accustomed to compare
ancient buildings with
one another. The first
occasion when exedras
of this kind were in-
troduced in the East
was in SS. Sergius
and Bacchus at Con-
stantinople, perhaps by
Anthemius, who after
gaining practice in this

earliest of Justinian's buildings, was in a better position to undertake the great
task of St. Sophia.

The dome, again, has a conical form ; it is constructed in the Ravennate fashion
with tapering tubes, and its stability partly depends on the walls of the outer drum

being carried up above half the height
of the dome itself. This, however, was
not done in the Roman fashion, because
the superstructure was not filled in to
form one solid mass with the cupola,
but in the Ravennate manner, previously
followed in the baptistery of Neon.
Whereas the great domes of the Byzan-
tine churches which are contemporary,
or nearly so, with San Vitale, derive,
under Roman influence, some of their
stability besides external weighting,
obtained by the raising of the outer

Fig. 107. Constantinople. St. Sophia (532-537).

Fig 108. Constantinople.
Dome of St. Sophia (558-563).

drum above the haunch, and filling
either from external buttresses set

against the drum and a portion of the
cupola itself, as in SS. Sergius and

Bacchus at Constantinople (Fig. 106), or else from buttresses encircling its circum-
ference, as may be seen in St. Sophia (Fig. 107), the dome of which (Fig. 108) was
rebuilt by Isidorus the Younger between 558, the year in which it fell, 1 and 563,

1 Corpus script, hist. by:. Theophanis chronographia.

70 LOMBARDIC ARCHITECTURE

when the church was restored. 1 These domes are, moreover, of a different type,
and are constructed in a different way. Thus, St. Sophia has a flattened dome with
ribs which show on the inner surface, and form continuations of the external support-
ing buttresses. And its material is brick. The dome of SS. Sergius and Bacchus,

Fig. 109. Constantinople. St. Sophia (532-537).

following a Roman suggestion, has its internal spherical surface divided into
compartments which are alternately flat and concave ; and this feature reappears,
though ill-formed, on the present external covering of the dome. Moreover, it is
provided with ornamented ribs (it is not known whether they are in stone or plaster),
which arch into one another at the crown, so as to leave a ring in the centre. Besides,
brick is the material used.

1 Corpus script, hist. byz. Chronicon Paschale.

THE SCHOOL OF RAVENNA

Again, the dome of San Vitale is not carried by triangular spherical pendentives,
as in the case of St. Sophia at Constantinople (Fig. 109), where a recent re-examina-
tion of the building has persuaded me that the cupola of Anthemius was depressed,
and, as in the great semi-domes of the hemicycles, supported by pendentives of the
Romano-Ravennate type, continuous with the dome. The same was the case
with St. Sophia at Salonica, where the external arched buttresses are a later addition
(Fig. no). The dome rests on a portion of drum forming a perfect junction at the
angles where the dome meets the rectilineal faces of the polygon, by means of a
recess or niche taken out of the angle at the point where
the drum having become circular would be in want of some
kind of support. This junction was an entirely new idea,
without any analogy in earlier buildings.

The origin of this pendentive is to be traced to the hemi-
cycles sometimes used by the Romans as supports for domes.
The way in which this was carried out may be seen in the two
lateral rooms of the internal west front of the" Domus Augustana"
on the Palatine (about 85).

Fig. no. Salonica. Si. Sophia (about 425).

Further, in San Vitale the graduated bracket-like projections of the eastern
pediment are a Vlth century decorative motive of the School of Ravenna, which has
absolutely no analogy in any earlier building, either in Italy or outside it. Finally,
its masonry, with the quality of the bricks and the way in which they are laid, as well
as the use of mortar composed of lime, sand, grit, and pounded brick, is the result
of local traditions. The last ingredient is freely used in the vertical walling, but
sparingly in the vaulting, where sometimes it is altogether absent. The same traditions
are also revealed in the construction of the still existing vault of one of the towers of
the narthex, copied, as we saw, from that of the dome of the neighbouring
mausoleum of Galla I'lacidia (about 440).

To conclude, San Vitale, finely thought-out example of the central architectural

7 2 LOMBARDIC ARCHITECTURE

type though it was, but, on account of its concentric form, ill-adapted to the Western
character and the liturgical wants of the Latin Church, passed, like its brother after
the flesh, San Lorenzo at Milan, merely like a brilliant meteor across the sky of
Italian architecture. Nevertheless, it must be regarded as a building which was both
in design and construction the work of craftsmen of the School of Ravenna, though
some of its decoration was due to Greek artists. For to the latter, if we may judge
from the present condition of the church, we must ascribe not only the capitals of the
colonnade on the ground floor and of the presbytery, but also the mosaics of the
sanctuary (probably purely Byzantine productions), as well as the original carving of
the altar and the screens which enclose the chancel.

BASILICA OF SAN LORENZO MAGGIORE AT MILAN. The analogy and family
likeness which the basilica of San Lorenzo at Milan presents to San Vitale at
Ravenna suggests the idea that they are not only contemporary (a view mentioned
before me by De Dartein), 1 but designed by the same architect, who could not have
been a Byzantine, inasmuch as the plan of San Lorenzo has no resemblance to that of
any church erected by Greek builders in the VI th or preceding centuries. The
vicissitudes which Milan went through in the Vlth century justify us in fixing its
erection before the siege by Uraias (538).

The building the beauty of which makes Arnulph exclaim : " O templum cui
nullum in mundo simile !" suffered from fire in lO?!, 3 and this gives reasonable

ground for suspecting that, as in the case
of San Vitale, the lower colonnade and
the dome were originally covered by
wooden ceilings. The damage, however,
was quickly repaired. In 1123 part of
the church collapsed, and the restora-
tion had scarcely begun when the catas-
trophe was consummated by a second
fire in 1124.* In the course of the restor-
ation which followed this new misfortune,
the lofty dome was buttressed by
ramping arches in the way shown in
a curious though not very trustworthy
print in Giulini's book. 5 After a fresh
and complete restoration it suffered from
another disaster in 1573, when a large

Fig. III. Milan. San Lorenzo Maggiore
(Vlth Century).

portion of the vaulting fell in. There

was now no question of a restoration, but

of the rebuilding of the principal part of the basilica, and this was carried out between
1574 and 1591.

The church preserves its original form, and rests upon the old foundations. It is
an octagon, encircled by an aisle with galleries, and supported on four of its sides by
square towers (Fig. ill). From the outside (Fig. 112) the original walls are seen to
be built of brick, and they are strengthened at the angles by substantial buttresses,

1 Etude snr F architecture lombarde.

3 Muratori, Rerum ital. script. Armdphi historia Meliolancnsis.

' 6 Giulini, Memorie spettanti alia storia, al Governs, cd alia dcscrizione della citta, e della cainpagna a
Atilano, ne' secoli bassi.

THE SCHOOL OF RAVENNA

while the intervening spaces are decorated with lesenas and a cornice marking the
floor-line of the internal gallery of the church, just as in San Vitale at Ravenna.

The best preserved of the square towers, the object of which was to increase not
only the stability but also the decorative effect of the building, has its outer angles
strengthened by returned lesenas, while its walls are constructed of regular courses of
bricks separated by layers of mortar of varying thickness. The internal angle at
the point nearest to the dome is strengthened inside by a boldly projecting buttress.
The original dome was not an octagon of bricks, supported at the angles " by a
number of small arches one above the other, each projecting a little further out than
the one below it, in the manner still to be seen in those at Sant' Ambrogio," to

Fig. 112. Milan. San Lorenzo Maggiore (Vlth Century).

quote the account left by Bassi, 1 for the original dome must have had a conical vault
constructed of terra-cotta tubes arranged in a spiral, like that of San Vitale at
Ravenna, and supported at the angles by niches. Compound conical pendentives, in
a perfect form as in Sant' Ambrogio at Milan, did not make their appearance till about
the early years of the Xllth century.

The interior of the tower is lighted by rows of large, round-headed, unsplayed,
windows. The structure of the walls shows that they belong to the same date as
those of San Vitale at Ravenna. They might in fact be described as the work of the
same builders. And the form as well as the distribution of the windows tell us that in
the Vlth century churches had not yet the adjunct of towers embellished by groups of
windows divided by shafts, with their heads sometimes recessed ; nor of towers which
were the expression of some artistic idea. So that the bell-towers attached to Vth and

1 Pareri e dlspcreri in materia di architetlura e prosptttira.

LOMBARDIC ARCHITECTURE

Vlth century churches of the Romano-Ravennateand Byzantino-Ravennate styles are
necessarily of a later date than the erection of the churches to which they belong.

Before finally leaving San Lorenzo, San Vitale, and the
bapistery of Neon, the three buildings in which the School

\ i / of Ravenna reached the zenith of its brilliancy in con-

struction, I should like to say a few words about the direct
descent of that School and the contemporary Byzantine
School from the School of Rome.

In the first three centuries after Christ and for part of
the IVth, the architects of Imperial Rome faced and solved
little by little, wholly or in part, the vastest problems of
construction and equilibrium that the world, so far at least
as we can judge, had as yet attacked and mastered. Rome,
on the eve of yielding up her sceptre to Constantinople,

Fifi;. 113. Rome. Basilica , .. . . ... .. . , .

Julia. Plan of Pier (284-305). emitted a brilliant ray of light in which all her dying

greatness was concentrated.

These solutions were later appropriated, developed, perfected, applied to new
ends, first of all by the builders of Ravenna, next by the Byzantine architects,
lastly by those of the Middle
Ages.

Before now it has been observed
that " every product of Egyptian,
Oriental, and Greek architecture,
appears as child's play by the side
of the fully developed Roman
arch " ; l and it has also been
remarked that the Basilica Nova
begun by Maxentius (310-312) was
the first example of a vaulted
basilica, and that in it " was solved
the problem which had kept the
whole of Western architecture in
anxious suspense." *'

Our statement can be verified
by anyone who cares to do so,
provided always that he has
mastered the science of construc-
tion and equilibrium, and is
acquainted with the great styles of
architecture in vogue amongst other
peoples, before and during those
centuries. To do so, it will be
sufficient for him to examine the
remains of the imposing baths,
villas, palaces, and tombs, which
Rome and its environs still preserve. And we must not forget the basilicas. In the
Basilica Julia in the Forum Romanum, as rebuilt by Diocletian (284-305), the
cruciform piers at the corners of the middle row are provided with angle supports

1 Wickhoff, Roman Art. '-' Riegl, Stilfragen.

Fig 114. Rome. Basilica Julia (284-305).

THE SCHOOL OF RAVENNA

for the springing of the groins of the quadripartite vaulting, thus anticipating the
compound Lombardic pier: a point which will have been noticed by others before
me, though they have not taken the trouble to publish it (Figs. 113, 114). Above
all, he must look at the Baths of Diocletian and Maximian. And the study must
be completed by an examination of drawings that have been preserved of these
monuments and of others which have been destroyed. For example, in a sketch
attributed by Hiilsen to Fra Giocondo, in Sangallo's volume in the Vatican Library,
we see a Roman portico adjoining the Theatre of Marcellus, having cruciform piers
(composed of a. pier with two pilasters and two half-columns attached to it)
with elaborate supports
at the angles for the
springing of the groins
of the intersecting vault-
ing. And these piers
alternate with columns

(Fig. US)-

I have specially
selected the Baths of
Diocletian (Fig. 116)
because, to my mind,
they sum up, so to speak,
all the great principles of
construction and statics
attained by Imperial
Rome ; and also because
it was to them that the
builders of succeeding
ages mainly had re-
course. It will be
enough if we give the
plan of the Tepiclarium
and the chambers im-
mediately adjoining,
which form the nucleus
of this stupendous
building (Fig. 117); and
also a section taken at the most important point of this central part, viz. the
great hall with its three rectangular bays, having a rotunda to the west preceded
by a semicircular recess, and its eastern side looking on to the Piscina (Fig. 118).

Many drawings (Fig. 119) of the Baths of Diocletian are in existence, and
various plans, as well as a few reconstructions, cither strictly architectural like
those of Palladio, 1 or made with a purpose partly architectural and partly artistic,
but mainly the latter. 2 No one, however, it appears has yet brought out the real
importance of the influence exercised by this structure on the great architectural
styles of later times both in its ground plan and in the principles of construction
and equilibrium on which it is based. For instance, attention has never been
called to the arrangement by which the circular hall connected with the
Tepidarium is confined between four towers, of which the round ones are simply

1 Le termc di Kama. Paulin, Thermes de DiocUtien.

115. Rome. Sketch of 1'ortico adjoining the Theatre of Marcellus.

7 6

LOMBARDIC ARCHITECTURE

staircases, while the rectangular ones are not only staircases but also act as the
outermost buttresses of the great hall. This circular hall apparently provided the

Fig. 1 16. Rome. Baths of Diocletian (306).

idea for the architect of the octagonal church of San Lorenzo Maggiore at Milan,
"edita in turribus " l like its Roman prototype.

In the same way the marvellous system of equipoise applied to the Tepidarium
has passed unnoticed. Here it is not merely a question of the simple and
intelligent grouping of the surrounding structures with the object of resisting the
thrust of the vaulting, as was carried out in Roman Baths. It is not even a case of such

A grouping assisted by the use of but-

i tresses at the external angles where

the thrust of the vaults is not counter-
balanced by barrel vaults, an expe-
dient a very early instance of which
is afforded by the great Baths in
Hadrian's Villa at Tivoli. On the
contrary, it is a rational system of

h\^^Jt AM % thrusts and counter-thrusts such as

Jr^ ? '"ti i ~ no great building had exhibited up

to that time. For though based on
the fundamental principles of equi-
librium applied in the past by Roman
builders, it still contains elements
hitherto unknown, which impress on
it a character of absolute novelty.
These auxiliary elements combine, in

one direction to secure the stability of the structure, and in another and here comes

out the practical Roman spirit to supply its needs.

1 Muratori, Serum ital. script. Versus de Mediolaito.

Fig. 117. Rome. Baths of Diocletian. Plan ot
Tepidarium and adjacent parts (306).

THE SCHOOL OF RAVENNA

These elements may be seen
applied, with this double purpose,
to the support of the three cross
vaulted bays of the great central
hall. The intersecting vaults
are of rectangular plan, the
proportion of the sides being
about 2 to 3, while the chord of
the elliptical arc of the diagonal
groins of the middle bay
measures about 106 ft. The
vaulting springs from columns
set against the side walls and
in the angles of the hall, a
device which has been described
as Byzantine, but is really Roman
and used as far back as the
1st century in the Baths of
Titus (80). In the Great Baths
of Hadrian's Villa at Tivoli
(125-135) stone corbels were
used instead of shafts. These
corbels are shaped somewhat
like Ravennatc pulvins, and
are stuccoed and painted all
over (Fig. 120).

Given the enormous span
of this vaulting, and considering
the instability of the diagonal
depressed ribs, the master archi-
tect of the building was not con-
tent with using concrete com-
posed of light materials in order
to make the thrust less danger-
ous. He did not confine himself
to springing the powerful, stilted,
diagonal ribs from columns sur-
mounted by an entablature.
But in order the better to guaran-
tee the structure against disin-
tegrating and dislocating move-
ments at the haunches of the
vault, to which it would have
been liable during the settlement
of the solid vaulting, he had
recourse to the following expe-
dients for ensuring its stability,
the western side he set four ramping buttresses, over 14 ft. thick,

(0
each one relieved by an arched passage opening, and with its back forming a flight of

LOMBARDIC ARCHITECTURE

Fig. 119. Rome.

Baths of Diocletian. {From a Drawing by Dosio in the
Uffizi.)

steps which gave access to the roof of the great hall. The weight of the two end
ones was transmitted outwards on to massive piers (Fig. 121), while the two inside

ones had the support
of substantial rect-
angular staircase tow-
ers. Pierced ramping
buttresses of this kind
were employed short-
ly afterwards, quite
likely by the same
architect, in the
Basilica Nova (Fig.
122), where we get a
repetition of the
arrangement of Dio-
cletian's Tepidarium,
with the great hall of
three bays covered by
quadripartite vault-
ing, flanked by six
compartments with

barrel vaults. The greater simplicity of the buttressing in the Basilica is to be
explained by the lessons learnt from the construction of the Tepidarium. It should,
however, be borne in mind that this simplifi-
cation was not for the benefit of the Basilica,
grand as it was, for the earthquake of 1348
brought down the nave and one of the
aisles ; 1 while Diocletian's Tepidarium, with
its more complicated but more stable con-
struction, is still there, practically intact, to
tell us what a great architect its designer
must have been. An account by Boni of
the recent excavations in this Basilica is in
course of preparation.

(2) On the eastern face he set four rect-
angular buttresses, arched like the others,
and enclosing stairs. Externally they are
strengthened by ponderous projecting piers
built up above the level of the buttresses,
and two of these contain spiral service-
stairs.

(3) He closed the two sides of the hall
by chambers with quadripartite vaulting,
again strongly ribbed (Fig. 123), which is
kept in place by a double set of massive
internal supporting piers at the angles, and
by equally solid external buttresses.

The system here described was the

1 Lanciani, The Golden Days of the Renaissance in Rome.

";*"

Fig. 120. Tivoli. Villa of Hadrian.
Great Baths (125-135).

Hall in

THE SCHOOL OF RAVENNA

Fig. 121. Rome. Baths of Diocletian (306).

source of numerous and important suggestions. Thus, the architect of St. Sophia
at Constantinople took from it the idea of the four buttress towers on the north
and south, which by
means of the same
number of arched
buttresses (in San
Vitale at Ravenna
Julianus Argentarius
had already provided
support for the piers
of the dome by con-
necting them through
arches with the
massive buttresses at
the external angles)
receive the thrust of

the great arches transmitted through the piers of the central dome of the most
famous of Justinian's buildings (Fig. 124). For everything leads us to think that
Anthemius, who is described by Procopius 1 as the master builder, must be regarded
not only as the builder, in partnership with Isidorus of Miletus, of Justinian's church,
but also as the originator of the plans for it. In fact, it appears from Procopius
that Isidorus was not the author of the design, but rather the associate of Anthemius,
and an architect capable of carrying out plans already prepared. Of Anthemius
we read in the Silentiary's poem that he was " skilled to draw a circle and set out

a plan." 2 Gyllius 3
had noticed the fact
before me : " Quam-
quam Anthemius,
qui aedem Sophiae
architectatus erat."
So that everything
leads us to believe
that Anthemius
studied on the spot
the great buildings
of Rome in order to
base on them his
plans for St. Sophia ;
and this is all the
more likely because
one of his brothers,
Alexander, followed
the profession of
medicine at Rome. 4

In this way would be explained the family likeness, to which we have already
called attention, between the plan of St. Sophia and the two halls of the Baths of

1 Corpus script, hist. byz. De aedificiis dn. [ttstiniani.

2 Migne, Pair. Gr., Vol. 86. Paulas Silentiarius, Descriptio Sanctae Sophiae.

122 Rome. Basilica Nova (310-312).

Of. fit. De Templo Sophiae.

Corpus siript. hist. bys. Agathiae Scholastid Myrinensis historiae.

LOMBARDIC ARCHITECTURE

Agrippa and Nero, as well as the Basilica Nova of Maxentius and Constantine.
And we must not omit the system of dividing the great windows by isolated uprights
followed in the Baths of Diocletian.

Fig. 123. Rome. Baths of Diocletian. Ribbed Vault (306).

Fig. 124. Constantinople. St. Sophia (532-537). (From a drawing in San Gallons volume
in the Vatican Library, made before the Turkish Conquest.}

The Lombard architectural gilds were also in touch with the principles of
construction and equilibrium as applied in the Tepidarium of the Baths of Diocletian.

THE SCHOOL OF RAVENNA

They borrowed from it, for instance, the arched ramping buttresses carried up above
the roof from the transverse arches of the aisles, as applied in San Babila at Milan
(Xlth century) and the church of Rivolta d'Adda (Xlth century). This expedient
was the origin of the flying buttresses of the Pointed style. Another thing that they
borrowed was the system of groining with diagonal ribs.

Anthemius of Tralles was not alone among Justinian's architects in deriving
suggestions from the great Latin mother-city. As a matter of fact, the designer of
SS. Sergius and Bacchus
at Constantinople (about
527) borrowed so freely
from the Licinian Nym-
phaeum that Choisy 1 re-
marks that the plan of
either building might be
a copy of the other. Nor
did he omit to notice the
radiating ribs forming part
of the concrete mass of the
Nymphaeum ; and they
suggested to him the ribs
which, as it seems, form
the structural skeleton of
the dome in SS. Sergius
and Bacchus (Fig. 125). It
appears that ribs standing
out from the inner surface
of the dome were them-
selves an idea borrowed
from Roman buildings. In
fact, if we may judge from
some of the illustrations
in Montano '-trustworthy
so far as the ground plans
are concerned, though the
elevations are based, partly
on remains then in exist-
ence, and partly on the
imagination it will appear

that the salient lines of the dome carried down on to the columns below were indica-
tions of the radiating ribs which formed the essential structure of the dome itself
(Fig. 126).

And this was not the only source, for he derived from another Roman building,
the Serapeum of Hadrian's Villa at Tivoli (Fig. 127), the idea of a dome, the surface
of which is (as others 34 have also noticed) a rhythmic sequence of flat and concave
sections unsupported by pendentives, simply flush with the course of the drum from
which they start, and not an alternation of segments of circles more or less concave

1 L'art de bAtir chcz les Byzanlins. * Op. cit.

3 Lethaby, Mediaeval Art.

4 The Builder, Jan. 6, 1906. Henderson, SS. Sergius and Bacchus, Constantinople.

VOL. I G

Fig. 125. Constantinople. SS. Sergius and Bacchus (alxjut 527).

LOMBARDIC ARCHITECTURE

to the interior, as has been supposed. 12 Further, he did not overlook the application
to his dome of the device of carrying up the drum above the haunch, so that by this

increase of weight he might
obtain greater resistance to
the thrust of the dome itself.
This expedient had been
employed on a grand scale
by the Roman builders ever
since the time of the Em-
peror Hadrian, and it reached
its climax in San Vitale at
Ravenna. Lastly, he did
not omit to pierce a circle
of openings in his dome,
after the manner of so many
buildings at Rome, for
instance the so-called
"Tempio di Siepe " (117-
138), the mausoleum of
the Gordians known as
"Tor de' Schiavi " (Illrd
century), and the great
circular hall of the Baths
plans were taken from San

Fig. 126. Palestrina. Roman Tomb.
(From Montana, "Li cinque libri di anhitettura")

The final touches of his

of Caracalla (212-216).
Vitale.

When Isidorus the Younger at a later date rebuilt the dome of St. Sophia at
Constantinople, he strengthened it with external buttresses, and provided it with
visible radiating ribs
suggested to him by the
"Mausoleum Augusto-
rum," a work of the
Vth century, in which
Honorius (395-423) was
buried, and to which the
body of Theodosius II
was brought from Con-
stantinople in 451.

This Imperial
Mausoleum consisted of
a pair of rotundas. One
of them, known as Sant'
Andrea or Santa Maria
della Febre, was con-
secrated by Pope Sym-
machus (498-514), and
demolished in 1776 (Fig.
128). The other, known as Santa Petronilla, was dedicated by Stephen II (752-757)

1 Choisy, Vart de b&tir chez les Byzantins.

2 Records of the Past, Dec. 1906. Marquand, The dome of SS. Sergius and Bacchus at Constantinople.

Fig. 127. Ti

( I2 5~'35)-

THE SCHOOL OF RAVENNA

and Paul I (757-767), and destroyed under Paul III 0534-I55O). 1 2 Of the latter
I reproduce an interesting sketch preserved by Giacomo Grimaldi 3 (Fig. 129).

The cupolas of these rotundas
were strengthened externally by
powerful buttresses, and internally by
visible radiating ribs springing from
brackets, as we learn from a sketch
preserved by Cancellieri 4 (Fig. 130).
Such ribs must have been supported
originally by wall-columns in the
manner shown by Fig. 126.

The School of Ravenna continued,
and at the same time improved on the
traditions of scientific construction as
practised by the builders of the Roman
Empire ; and the connecting link is
to be found in the architects and con-
structors who settled at Milan after
Maximian had fixed his official
residence there. The failure of all
preceding writers to recognise the
existence of this School, so totally dis-

Fig. 128. Rome. Imperial Mausoleum near St. Peter's
(Sant 1 Andrea) (Vth Century). (From a painting in
the Vatican Library.}

tinct from the Byzantine, has resulted in the invention by so many of them of an

imaginary Byzantine style in Italy. On the

F v .'. ^ \ . ^ contrary, so far as architecture is concerned,
^^_ /X 'V "i tnat style was, in its earlier or Romano-
Ravennate phase, the creation of Italian
builders, seeing that not one of its distinctive
features had previously made its appearance
in the East ; while in its second stage, it
became the Byzantino-Ravennate style, based
on principles derived from Rome and Ravenna,
together with suggestions drawn from the
School of Salonica.

Another result has been that, in order to
explain the presence in the monuments of
Ravenna of essential elements which are not
to be found in contemporary or earlier build-
ings of the Byzantine style, some of these
writers were compelled (De Dartein, 5 for in-
stance, in the case of the massive buttresses
at the external angles of San Vitale at
Ravenna) to fall back on a supposed influence

Fig. ia9.-Rome. Imperial Mausoleum of the foreign soil in which Byzantine archi-
near St. Peter's (Santa Petronilla) (Vth
Century).

1 De Rossi, Inscriptiones Christianae tirbis Romae.

" Nuovo Bullettino di arch, crist., 1905. Rohault de Fleury, Saint And rJ au Vatican.
3 Rome, Biblioteca Casanatense. Cod. MS. 242l.Catalogtis sacrarum rtliquiarum Vaticanae tasilicae
principis Aposlolorum.

* DC Secretariis novae basilicas Vaticanae. Lib. II. 5 Op. at.

G 2

84 LOMBARDIC ARCHITECTURE

tecture was endeavouring to take root. Or else they found themselves obliged
to assert, for instance, that domes constructed with tapering terra-cotta tubes,
like that of San Vitale, were of Byzantine origin ; whereas, on the contrary, the
first to make use of such terra-cotta tubes in vaulting were the Campanian
builders, and those of the Roman province of Africa, and of Sardinia as proved by
the excavations of 1876 in the so-called house of Tigellius at Cagliari. Moreover,
the oldest example on record of a dome so constructed is that of the apse of the old
Basilica Ursiana at Ravenna (370-384). Again, we know how the domes of Eastern
churches were constructed in the period to which San Vitale belongs : St. Sophia and

Fig. 130. Rome. Imperial Mausoleum near St. Peter's (Sant' Andrea) (Vth Century).

SS. Sergius and Bacchus at Constantinople are there to tell us. Unless, indeed, such
spaces were covered by wooden roofs, like St. George of Ezra (515-516) and the
cathedral of Bosra (511-51 2).

When others who have taken the trouble, as I have done, to make themselves at
home in the science of construction and the builder's craft, shall have studied the
vaulted architecture of Rome and Ravenna with the same devotion that has been
lavished on the contemporary Eastern styles, and with as great or greater thorough-
ness, it will at last be determined whether the East exercised on Italian architecture
the influence ascribed to it by Cordero, 1 Cattaneo, 2 Strzygowski, 3 and so many others ;
or whether, on the contrary, it was Roman principles of construction, a creation of
the Latin mind, that, together with those of Ravenna, were infused into Byzantine
architecture, which is the conclusion to which I have come and, as I believe, have
proved.

1 DelT Italiana architettura durante la dominazionc Loinbarda.
2 I? architettura in Italia dal secolo VI al Mille circa. 3 Orient odcr Rom.

THE SCHOOL OF RAVENNA 85

THE BASILICA OF SANT' APOLLINARE IN CLASSE, erected by Julianus
Argentarius at the order of Archbishop Ursicinus (533-536), was consecrated in 549
by Archbishop Maximianus, the builder of Santo Stefano in Olivis at Ravenna, and
of Santa Maria Formosa at Pola. 1

It consists of a nave and two aisles (Fig. 131), with wooden roofs, separated, like
Sant' Apollinare Nuovo and San Giovanni Evangelista, by twelve marble columns on
either side, the number of the Apostles, surmounted by capitals of Composite character
with protuberant leaves of the acant/ius spinosns, deeply undercut so that the shadows
are strongly accentuated, and treated in a monotonous manner with rows of small
holes made with the drill along the ribs of the leaves. These capitals, which carry

Fig. 131. Classis (near Kavenna). Sant' Apollinare (533-549).

the ordinary Ravennate pulvins marked with crosses on their outer faces, must be
ascribed to Byzantine chisels, not only on account of the design and technique, but
also because they do not all exactly fit their columns ; so that it may be reasonably
inferred that they were not made on the spot.

The nave ends in an apse, semicircular internally and five-sided externally,
flanked by two sacristies which form prolongations of the aisles, and have apses of the
same form as the principal one. The raised chancel with the crypt beneath it, which
some believe to be contemporary with the church, are really works of the Xllth
century, carried out after the relics of St. Apollinarts were removed from beneath the
altar of the Virgin, and deposited in a more conspicuous position in the central part of
the church. 2

Originally, the oldest churches of Ravenna possessed neither crypts nor elevated
presbyteries. As for those of considerable elevation, like that of Sant' Apollinare

1 Moil. Germ. Hist. Agnellus, Liber ponlificalis.

3 Fabri, op. fit.

LOMBARDIC ARCHITECTURE

which is raised as much as eleven steps above the ground floor, I have never yet
found a trace of such in any church of certain date before the second half of the Xth
century. Presbyteries of moderate elevation, such as that of Kalb-Lauzeh in Central
Syria (Vlth century) raised seven steps above the pavement of the church, were never
erected above a Confessio. In Syria the origin of these elevated platforms may be
traced to the raised sanctuaries of temples of the Roman epoch. At Baalbeck, in
the temple of Bacchus, the platform on which, looking west, the image of the titular
divinity was erected, with its annular barrel vaulted crypt beneath, is raised as many
as sixteen steps above the floor of the temple.

The walls are constructed of courses of brick separated by layers"of mortar of

Fig. 132. Classis (near Ravenna). Sant' Apollinare (533-549J

varying thickness. The side walls of the nave and aisles (Fig. 132) are decorated
with blank arcading corresponding to the arcades of the interior. These are pierced
by very large round-headed windows. The arcades of the aisles springing from
pilasters which project about 8 in., thus allowing the wall itself to be reduced
to the moderate thickness of about 2 ft, describe an unbroken curve round the
windows without any indication of capitals. Those of the nave, on the other hand,
have impost cornices formed of three projecting courses of bricks, and rest on bases
of the same material.

The apses are decorated with a saw-tooth cornice. The principal one, at its
junction with the eastern wall of the church, is flanked by massive buttresses. The
end walls of the aisles are raised above the line of the roofs so as to form two strong
abutments corresponding to the sanctuary arch. The eastern front exhibits, at the
lower ends of the gable and of the two half-gables, the characteristic graduated

THE SCHOOL OF RAVENNA

bracket-like projections which we saw for the first time in the eastern front of San
Vitale. The western front, strengthened by pilasters at the outer angles, is pierced by
a three-light window with small shafts carrying pulvins carved with crosses on their
faces ; and this, with the round-headed windows in the side walls and the apse,
provided all the light for the church. Originally this front was approached through
a square atrium or cloistered fore-court, the only traces of which above ground belong
to the side from which the church was entered. It was closed at either end by a
squat tower. The one to the left still exists, and measures internally about 28 ft. by
31 ft. Close to the left aisle rises the majestic campanile. It has a wooden roof, and
its construction, in which shorter and thicker bricks are used than those in the walls
of the church, shows that it is an addition of a later date than that of the
basilica itself.

The architect of Sant' Apollinare in Classe, as we have already observed in the
account of San Vitale, was Julianus Argentarius. And it was the work of builders of
Ravenna, as is indicated by the construction of the walls, and also by the decorative
motives of blank arcades and saw-tooth cornices.

The most notable thing about this building is the method adopted by its creator,
chiefly with a view to lightness of construction and therefore economy, of
compensating the thinness of the outer walls by facing them on the outside
with blank arcades, and strengthening them at the most important points by
buttresses of greater or less substance as the occasion demanded. The principle
of making the elements of resistance depend on their distribution
rather than on their bulk had been already applied by the
same architect in San Vitale, where he displayed a marvellous
grasp of the principles of scientific construction such as had
not been seen, so far as we can judge from existing monuments,
since the erection of the Baths of Diocletian.

THE CATHEDRAL OF PARENZO was the work of its first
bishop, Euphrasius (about 521 or 522-553), and was erected
between 535 and about 543- 1 The founder with a model of
the church in his hand is represented in the semi-dome of the
apse, at the base of which may be read his dedicatory inscription.
Restored again and again in mediaeval and modern times, enough
of the orginal structure still survives to make it one of the most
valuable monuments of the early centuries of Christianity that the
Italian peninsula can boast.

It is a basilica with a nave and two aisles, the former ending
in a deep apse, internally semicircular and decorated with precious
mosaics, while the exterior presents the form of a semi-dodecagon.
The side apses are merely niches sunk in the outer walls (Fig.
133). The nave is separated from the aisles by twenty round
arches, ten on either side, supported by marble columns, on
which are set capitals carrying pulvins of the Ravennate type
bearing the monogram of Euphrasius (Fig. 134).

These capitals do not in every case fit their columns, which

makes one think that they were not wrought on the spot but were imported from
Constantinople, where it seems that, in the V I th century, marble capitals were prepared

1 Jackson, Dalmatia, the Quarnero, and I stria.

Fig. 133. I'arenzo.
Plan of Duomo
(about 535-543)-

LOMBARDIC ARCHITECTURE

and carved for exportation. It was from this source, in all probability (unless
indeed they imported them from Salonica), that the craftsmen of Ravenna, capable
architects and builders, and excellent mosaic workers as they were, but not such
skilful carvers as the Greeks, procured the capitals of the marble of Proconnesus
(the kind in common use in the capital of the Eastern Empire), worked in the
Byzantine style, which they used not only in the nave of the cathedral of
Parenzo, but also in those of the cathedral and the church of Santa Maria at
Grado (571-586), of Sant' Apollinare in Classe near Ravenna (533-549), and of
the abbey church of Pcmposa (VI th century), as also in the lower colonnade and
the upper arcades in the presbytery of San Vitale at Ravenna (526-547).

Fig. 134. Parenzo. Duomo (about 535-543).

These capitals belong to three Byzantine types : the cubical funnel-shaped
(Fig. 135); the Composite bird and basket (Fig. 136); and a Byzantine version
of the Composite capital, with the body shaped like an expanded calyx. Both
in design and execution, all of them, with those of the atrium, are clearly the
work of Byzantine hands.

With the exception of the apses, the building is entirely roofed with timber.
The outer walls are of " opus incertum " of broken stone and brick. On the
outside they' are strengthened by buttresses at the angles. The walls of both
nave and aisles are decorated with blank arcades. The exterior of the apse, on
the other hand, is plain.

Opposite to the basilica opens the door of the octagonal baptistery, which
is contemporary with the atrium and basilica. This arrangement had been
employed as long ago as the IVth century in the cathedral of Aquileia (rebuilt
by the patriarch Poppo, 1017 or 1019 -1042 or 1045), and was probably suggested
by that of the cantkams which stood in the middle of the atrium in the oldest

THE SCHOOL OF RAVENNA

Kg. 135. Parenro. Capital in the Duomo (about
535-543)-

Christian basilicas. The Byzantines did not introduce it in any church of earlier
date than the one with which we are dealing. For instance, in the pre-Justinianean
St. Sophia at Constantinople the bap-
tistery was circular, and placed at the
side ; and there it may still be seen
standing to the north of the present
church, exactly like the baptistery of
the Constantinian basilica of St. John
Lateran at Rome. For, as has been
suggested before now, l this large circular
building, with its rectangular niches
recessed in the thickness of the wall,
suits the time of Constantine I (306-337)
and Constantius II (337-361), who ap-
pear respectively as the founder and
rebuilder of the original St. Sophia,
and is in all probability that baptistery
which was large enough to accom-
modate the Sixth Council of Con-
stantinople (394). 2 Later it was turned
into a sacristy, after the erection of the

new baptistery (now the tomb of Mustapha I) by Justinian near the Horologium,
with the dedication of St. John the Forerunner. 3 The latter building in form and

construction exactly suits the time of
Justinian, with its octagon planned like
the two lateral rooms of the inner west
front of the " Domus Augustana " on the
Palatine at Rome (about 85), though the
central space passes into the circle of the
dome by means of eight spherical pen-
dentives of the Romano-Ravennate type,
continuous with the dome itself.

The upper part of the facade of the
cathedral of Parenzo is pierced by three
large, round-headed windows.

The church of Euphrasius, possibly
designed by Julianus Argentarius, was, so
far as its construction is concerned, ap-
parently the work of builders from Ravenna.
Their presence is revealed by the plan,
taken from the Roman basilica, with the
modifications introduced by the School of

apses.

Ravenna ; that is to say, the apse with its
polygonal exterior, flanked by sacristies,
which in this case are reduced to minor
Another feature is the external decoration of blank arcades, which we

Fig. 136. Parenzo. Capital in the Duomo (about
535-543)-

1 Lethaby and Swainson, The Church of Santta Sophia, Constantinople.

* Du Cange, Hist. Byz. Constantinopolis Christiana.

* Banduri, fmferium Orientate sive antiquitates Constantinopolitanae, A -tony inns, de Saiicta Sophia.

90 LOMBARDIC ARCHITECTURE

find in the contemporary as well as in the older basilicas of Ravenna. Then,
there is the strengthening of the external angles ; and, lastly, the tri-apsidal
arrangement.

CHURCH OF SAN VITTORE AT RAVENNA. We know that it was in existence
in 564. 1 Alterations in later times reduced it to the state which it presented before
the recent restoration.

Originally it consisted of a nave and two aisles. Part of the nave survived, with
brick piers of "f" shape carrying an arcade ; and it ended in an apse, semicircular
internally and five-sided externally. Both nave and apse were lighted by very
narrow, round-headed windows, splayed on the inside. The pilasters attached
to the piers were carried up and merged in an arched corbel course. There is the
usual saw-tooth cornice at the top.

The late restoration was based on indications of the portions which had
vanished. It was then made clear, among other things, what was the size of
the original building, the nave of which had a width of about 21 ft. from pillar to
pillar, while the aisles measured respectively, about 13 ft. from the pillars to the outer
wall in the case of the left aisle or women's side, while the right was under 9 ft. wide.
There were also discovered on the front of the church, which was pierced high up by
three round windows, two buttresses marking the internal distribution of nave and
aisles.

The church of San Vittore tells us. that, in the second half of the Vlth century,
or, to be more precise, after the erection of the basilica of Sant' Apollinare in Classe
(533-549) and the cathedral of Parenzo (535-543), and about the year 564, when
we know that our church was already in existence, the builders of Ravenna were
beginning to light their basilicas with narrow windows instead of the spacious ones
which they had employed previously. We cannot say whether this was due to the
spirit of the age, or to the fact that from lack of pecuniary means the new basilicas
no longer displayed the same splendour of gold and gleam of precious marbles as the
churches of Ravenna had commonly done in the past, and therefore no longer
demanded a superabundance of light.

At the same date another church of Ravenna was provided with narrow windows
Sant' Andrea, of which only some poor relics survive. Founded by Archbishop
Peter Chrysologus (433 or 439-449 or 458), it was certainly restored by Archbishop
Maximian (546-556), when, as Agnellus relates, marble columns replaced the
wooden supports. In the course of this restoration it seems that the inscription
referring to the original foundation, and also the portrait of the founder, were
preserved. Agnellus gives an account of them.

San Vittore and Sant' Andrea provide a test by which the age of other buildings
in the Ravennate style, which have no certified dates, may be fixed approximately.

CHURCH OF SANTA MARIA DI POMPOSA. Its erection is generally referred
to the Vlth century, and there is proof that it was already standing in 59O. 2 I
believe that it may be ascribed to the years which followed the consecration of
Sant' Apollinare in Classe, and before the building of San Vittore at Ravenna.
The reason is that, in the interior, marble columns were still used, as in the church
at Classis, whereas brick piers were employed in San Vittore. And again, the nave

1 Fabri, op. cil 2 Bottom, Pompom al tempo di Guido suo monaco.

THE SCHOOL OF RAVENNA 91

walls at Pomposa are not pierced with the narrow windows of San Vittore and
Sant' Andrea.

The interior is divided into a nave and aisles, with nine arches on each side
supported by marble columns, on which may be seen, among others, capitals in the
Byzantine style carrying the usual Ravennate pulvins and suggestive of the Vlth
century. It terminates in three apses, the principal one being polygonal externally
and semicircular internally, while the subordinate ones are semicircular both inside
and out.

The exterior of the nave is decorated on its northern face, which is the least

*'g- '37- Pomposa. Santa Maria (Vlth Century).

restored, by blank arcading, the openings of which, as in the nave arcade in the
interior, are of various sizes. Each of the arches contains a very large round-headed
window. The aisle walls were evidently originally decorated on the outside with
lesenas, and pierced by narrow windows. The upper part of the front (Fig. 137)
exhibits in the gable the two bracket-like projections characteristic of Ravenna, and
is strengthened by two buttresses which divide it into three parts. This division had
been already applied to the front (now destroyed) of Santa Croce at Ravenna. 1
Against the lower part a narthex was added at the time when various decorative
works were carried out, and the church was consecrated in 1026. The date may
be read in the centre of the mosaic pavement of the nave. The exterior of this
narthex is decorated with interesting carvings, on which instructive comparisons may
be based. Near to the left aisle rises the imposing campanile, about 163 ft.
high, built in the Lombardic style in 1063 to replace a massive lighthouse
tower. 2

1 Bull, di arch, (ristiana, 1866. F. Lanciani, Scopertc neg li edifizi cristiani di Rcevenna.
" Federici, Rerum Pomposianarum historia.

92 LOMBARDIC ARCHITECTURE

CHURCH OF SAN PIETRO IN SILVIS, OR PARISH CHURCH OF BAGNACAVALLO. -
The construction is evidence for assigning it to about the same date as San Vittore
at Ravenna. Of previous writers, Graziani ' considered it to belong to the Vlth or
preceding century ; Cattaneo 2 placed it in the Vlth.

The interior contains a nave and aisles, separated by plain piers of X form
supporting the round arches which carry the walls of the nave. The pilasters which
project from these piers and increase their solidity stop before reaching the line of
the aisle roofs. Both nave and aisles have wooden roofs, and the former ends in an

Fig. 138. -Bagnacavallo. Parish Church (Vlth Century).

apse, semicircular internally and polygonal externally. One of the aisles is wider
than the other.

The walls, as well as the piers, are built of regular courses of brick set in
mortar made of lime, sand, grit, and pounded pottery. The side walls of the nave
(Fig. 138) are decorated with a large arched corbel course marked off in pairs by
lesenas, which rise from a stringcourse of brick, and are crowned by a saw-
tooth cornice. Windows open in it at regular intervals -no longer the large ones
of the older basilicas of Ravenna, but of restricted dimensions. They have round
heads, and are splayed inside. The side walls of the aisles are on the outside
divided into compartments by lesenas, and have very narrow round-headed windows,
mere loopholes in fact, splayed both inside and out. They do not correspond to the
windows which light the nave.

1 Notizie istoriche delta chiesa arcipretale di San Pietro in Sylvis di Bagnacavallo.

Op. cit.

THE SCHOOL OF RAVENNA

The front of the church presents a uniform surface, but the divisions of the
interior are indicated by two buttresses which correspond to the lines of the nave

Fig. 139. Bagnacavallo. Parish Church. Arch of Ciborium (783-826).

arcades. The outer angles are strengthened by returned lesenas. The pediment of

the nave, below which is a two-light window with stilted arches resting on a marble

shaft carrying a pulvin, and the half-pediments of the aisles, are ornamented with

the characteristic graduated bracket-like projections so often met with. The eastern

pediment has similar graduated projections,

and is pierced by a small window in the

form of a cross, now blocked up, but origin-

ally intended to provide ventilation for the

timbers of the nave roof. A round tower

formed an adjunct to the church in former

times, but it is believed to have collapsed in

the earthquake of 1688, and has now com-

pletely disappeared.

In the church are preserved two arched
tops of an altar ciborium (Fig. 139) given by
one John, who was the parish priest at the
time when Deus Dedit was bishop of Faenza
(783-826). 1 It is Ravennate work.

The church of Bagnacavallo, which by a
fortunate chance has kept its original form
almost untouched, claims our special attention.
It is, in fact, the ecclesiastical building which
provides the oldest surviving specimens of

. . i . ,i

narrow round-headed windows with double

splays. The Romans sometimes used openings

of this kind in sepulchral chambers, for they are to be seen in two such structures

illustrated by Montano, 2 which formerly stood, the one outside the Porta Salaria, the

140 Rome, \illacalled " Sette Bassi.

Double-splayed window (100-155).

1 Graziani, op. fit.

3 Op. cit.

94 LOMBARDIC ARCHITECTURE

* other on the Via Labicana near Rome. The only existing example which I can
point out, and it has not been previously noted, is to be found in a cryptoporticus (a
good deal of which is buried by the accumulation of soil) of the villa known as the

"SetteBassi" on the Via Latina (100-155), where may be
seen a row of loopholes with double splays (Figs. 140, 141),
still preserving their brick facing. The builders of Ravenna
were, however, the first to use them in churches, just as may
be seen in the one at Bagnacavallo, which unquestionably
Fig. 141. Rome. Villa called came from their hands. Later they were appropriated by

wtndow the Comacine gilds. In the present case the employment
(100-155). of these openings, about which there has been so much

fanciful writing, is easily explained by the conditions of the

locality in which the church was situated. The reasons were : first, the necessity of
preventing ill-intentioned persons from obtaining an entry into churches in remote
and unprotected situations, during the hours when they were closed, by means of the
windows nearest to the ground, and consequently easiest of access. In the second
place, the need of compensating for the loss of light resulting from the restriction of
the window opening to the smallest possible compass ; for it is known that a double
splay, as compared with a single one, has the advantage of admitting a greater amount
of both direct and diffused light. Thirdly, the convenience of making the windows,
by means of the double splay, appear larger than they really are, and so contributing
to the decorative treatment of the walls, as well as to the monumental aspect of the
building.

The church also provides an interesting example of an unbroken fa9ade divided
into compartments by buttresses which fulfil the triple purpose of clearly indicating
the divisions of the interior, of decorating the front of the church, and of providing
additional support.

CATHEDRAL OF GRADO. The present building is the work of the patriarch of
Aquileia, Elias (571 586). 1 This is confirmed by the mosaic inscription existing in
the pavement of the church.

It is a basilica with nave and aisles. The former terminates in a deep apse,
semicircular internally, and polygonal externally, and is separated from the aisles by
marble columns crowned with capitals, some of which are. Roman ones brought, in all
probability, like the columns and their bases, from the neighbouring Aquileia ; while
others are cased in stucco, and others, again, are of the same date as the building of
the church. These last are in some cases Composite, with the body shaped like a bell,
and turn-over leaves of the acanthus spinosus laboriously worked with the drill.
Others, again, are of the cubical Byzantine type with foliage carved on the faces, and
sometimes crosses made in the sides of the basket, and deeply undercut. They are all
the work of Greek chisels, as is clear from the execution.

The arches, above which rise the walls of the nave carrying the open
timbered roof, spring directly from the abacus of the capitals, thereby securing the
advantage of admitting more light and increasing the elegance of the building. This
very simple method of springing arches from columns, previously adopted in the
palace of Diocletian at Spalato, built between about 300 and 3O5, 2 had its origin at
Pompeii. Arches (of course earlier than the catastrophe of 79) with this peculiarity
have been found there, and are mentioned by Choisy. 3

1 Gams, op. cit. - Jelic, Bulic, and Rutar, Guida di Spalato e Salona. 3 Histoire de t Architecture.

THE SCHOOL OF RAVENNA

These nave walls are strengthened on the outside, at points corresponding to
the columns of the interior, with lesenas of slight projection finished off by a
moulding. They now stop short of the cornice under the roof, but originally must
have carried blank arcades. To each pair of these lesenas correspond single ones on
the inside, starting between the arches of the nave and rising as high as the tie-
beams of the roof. The original windows were small and round-headed. The gable
of the front exhibits the usual Ravennate bracketed projections.

The church is approached through an atrium of the same date, part of which was
taken up later by a square campanile. Close by, on the north side, as in the case of
the original Basilica Ursiana at Ravenna (370-384), stands the octagonal baptistery,
with a deep apse of the Ravennate type projecting on the east. This has been
recently restored, and is quite devoid of ornament.

The cathedral of Grado is probably a work of the School of Ravenna, with
contributory help from Greek carvers. For though, as we saw, the capitals of the nave
columns, wrought expressly for this building, are
to be ascribed to the Byzantine School, on the
other hand the design and the construction of the
church belong to that of Ravenna. This is made
clear by the form of the apse ; by the introduction
of the typical decorative blank arcades, and the
characteristic graduated bracket-like projections ;
and, lastly," by the use of narrow windows which,
as our examination of the buildings has shown,
were preferred by the builders of Ravenna in the
second half of the VI th century for lighting their
basilicas. I have not come across their application
to an entire building in any Eastern church which
I have seen, and ascribed, or possibly ascribed, to
the same century as that which saw the erection
of the cathedral of Grado.

Granted the presence of Ravennate builders
at Grado, we may reasonably attribute the con-
struction of the small church of Santa Maria delle
Grazie (Fig. 142), close by, to the same two
sources. And this in spite of its having an apse
flanked by two lateral chambers, and included with
them in the rectangular end of the church.

There was no occasion to get Byzantine builders, as Cattaneo 1 imagines, to import
this arrangement into Italy, where, not to cite other instances, the Xenodochium of
Pammachius at Porto, and the large Basilica of Santa Sinforosa on the Via
Tiburtina near Rome, provided early examples of it.

Fig. 142. Grado. Santa Maria delle
Grazie. Capital in the Nave (Vlth
Century).

When Narses was replaced as viceroy of Italy by the unwarlike Longinus (568),
the misgovernment of the latter and the other exarchs who succeeded him, the
religious strife which raged between the Church of Rome and the Church of
Ravenna, the archbishops of which, strong in the Emperor's protection, had

1 Op. dt.

LOMBARDIC ARCHITECTURE

assumed the title of Pope, and finally the dreaded invasions of the Lombards in the
P^xarchate, gradually deprived Ravenna of all power and splendour. The last blow
was given by the Donation of Pippin (755), which resulted in the definitive
subjection of the Church of Ravenna to that of Rome. The School of Ravenna, too,
following the fortunes of the State, gradually fell into a condition of decay,
until it finally disappeared to make way for the Lombardic School which was coming

into being.

Owing to the
distressof the times
very few buildings
exist which can be
ascribed to the
agency of Raven-
nate builders. All
the same these
buildings possess
no small interest,
both on account of
certain new ele-
ments of construc-
tion and decoration
which they contain,
and also for the con-
siderations which
may be based upon
them.

GUARD-HOUSE
OF THE PALACE
OF THEODORIC AT
RAVENNA. Re-
cent operations
have freed this
building from later
accretions and
made it clear that
the reputed re-
mains of the palace
of the great Gothic
king are a later

addition to the palace itself, made perhaps at the beginning of the VI Ilth century
by the exarchs, who, frightened by the spread of the Lombard power, and dreading a
surprise attack, fortified themselves in the palace of Theodoric which had become
their residence (Fig. 143).

The facade, constructed of materials taken from older buildings, is finished
at either end by a massive angle-buttress crowned by a cornice forming a pediment.
At the top they merge into arcades which form two blank hanging loggias. Between
these the building advances in the centre, and in the lower part of this projection the
entrance is formed, flanked on either side by an arcade of two arches supported by

Fig. 143. Ravenna. Guard-house of the Palace of Theodoric (VHIth Century).

THE SCHOOL OF RAVENNA

a column. The marble jambs of the door, made up from various sources, though the
carving on the imposts shows differences of execution between the two sides, still
clearly belong to the same date, and are contemporary with the carvings in the
neighbouring Sant' Apollinare Nuovo of the time of Theodoric. A round-headed
niche is recessed above the door. The internal passage of the ground floor has
cross vaulting supported by arches which spring from corbels. The upper story
was reached by two corkscrew staircases formed in the two towers which flank
the inner door.

The peculiarities worth notice, presented by this building, are the following:

(1) The advancing centre of the facade, with the entrance door surmounted
by an arched niche an

anticipation of the project-
ing porch of the door of a
church with an open loggia
above it, which we find in
some Lombardic churches,
e.g. the Cathedral' of Mo-
dena (1099-1106) which
had Lanfrancus for its
architect. 12

(2) The decorative
feature of hanging loggias,
probably borrowed from

the pensile arcades which .. .^ -^ ^_

ornament the upper row 3$^2 WtT

of niches in the Golden g^l % A fc

Gate of Diocletian's palace

at Spalato (Fig. 144).

(3) The vaulting sup-
ported by prominent
transverse arches spring-
ing from corbels. This
device, perhaps suggested
by the arches springing
from brackets which, as
far back as the Vlth cen-
tury, the craftsmen of Ravenna had used for a decorative purpose on a sarcophagus
in Sant' Apollinare in Classe, is an entirely new idea. For though, long before this,
use had been made of transverse arches supported by corbels projecting from the
main walls, and sometimes decorated on the outer face, as, for instance, those in
the narthex of the basilica of Eski-Djuma at Salonica, on which a simple cross, or
a cross in a wreath with a dove on either side, are carved, still such arches were only
intended to carry flat ceilings.

Fig. 144. Spalato. Palace of Diocletian. Golden Gate (about 30x3-305).

THE CHURCH OF SANTA MARIA IN VALLE AT CIVIDALE IN FRIULI consists
of a square chamber under 20 ft. wide, with a cross vault (Fig. 145). At one end
of this is the presbytery, divided into three small chapels by four columns and two

1 Muratori, Rerum Ital. script. Translatio corporis S. Geminiani,

2 Bortolotti, Antiche vile di San Gcminiano.

VOL. I

9 8

LOMBARDIC ARCHITECTURE

pillars carrying architraves, from which spring three barrel vaults. The capitals
of these columns are Byzantine Corinthian, showing both in design and execution
a certain reversidn to the classical manner. This tendency appears in the treatment
of their upper part, and also in the manner in which the wild acanthus foliage which
decorates them is carved.

Fig. 145. Cividale. Santa Maria in Valle (762-776).

The sanctuary is separated from the church by a low marble screen, and a wooden
beam supported by two small pillars with Byzantine Corinthian capitals showing two
rows of leaves of the acanthus spinosus, treated in the same style of carving, midway
between the Roman and Byzantine, as the larger capitals in the sanctuary.

The exterior of the walls of the church is decorated in their upper part with
blank arcading, every section of which contains a round-headed window. The
sanctuary is lighted by three arched windows smaller than the others.

THE SCHOOL OF RAVENNA 99

As to the date of Santa Maria in Valle, very different views have been put for-
ward by archaeologists and art-historians. So great is the discrepancy that, while some
think that we have to deal with a classical building, probably a temple, to which the
sanctuary was added in the time of Pertrudis, wife of the Duke of Friuli, to whose
piety also are to be ascribed the stucco decorations which are one of the attractions
of the interior of the church, others, on the contrary, argue that the building which
we see was entirely built and decorated by order of this lady, while another view is
that it is the result of a rebuilding in the Xlth or Xllth century.

I am unable to give my adhesion to any of these theories, for a careful examina-
tion of the structure, and comparison with a number of other works of art, have led
me to quite different conclusions. My view is that the existing building was erected
by order of Pertrudis (762-776), and is the work of Ravennate builders. The unbroken
continuity of the walls of the nave and sanctuary proves that they are of the same
date. And the decorative arcading on the exterior, with windows in each division,
confined, however, to the side walls, as we have often seen in older Ravennate buildings,
betrays the presence of craftsmen belonging to that School. The marble capitals,
too, all of which, especially the smaller ones, are clearly the work of one and the same
hand, reveal the decadent Byzantine manner modified by classical reminiscences,
which is characteristic of the Ravennate carvers of the Vlllth century.

As to the vaulting of the nave, no plausible reason can be given why it should
not be regarded, like the stilted barrel vaulting in the sanctuary, as the work
of Italian builders of that period. With the sharp edges of its groins only
maintained about halfway from the angle corbels on which they rest, and then
growing flatter as they gradually rise to their intersecting point, the cross vaulting
is, on the one hand, manifestly earlier than the Xlth and Xllth centuries ; while,
on the other, it does not exclude the presence of Italian workmen, since, as we
shall have occasion to see when dealing with the ecclesiastical buildings of the
Carolingian epoch, Italian builders did not in every case betray that want of
technical experience which many writers like to fancy that they displayed.

Later, possibly in the Xllth century, the front was rebuilt. In its construction
the marble fragments were used which have now been removed, and are to be seen
attached to the walls of the narthex. At the same time the stucco decorations
were carried out which form one of the treasures of Santa Maria in Valle. In
the Vlllth century there is absolutely no place for this rich, graceful, attractive
decoration, with its bold modelling, its correct and natural outlines, completely
underworked, which could not even have been produced in the Vth or Vlth centuries,
though this description of plastic decoration was highly esteemed by the artists
both of Ravenna and Constantinople in that period. An example of a beautiful
underworked moulding produced by the latter in that age may be seen in SS.
Sergius and Bacchus at Constantinople (about 527).

CHURCH OF SANTA MARIA DELLE CACCIE AT PAVIA. Of the original structure,
founded by King Ratchis (744-749),! nothing was left a few years ago but a fragment
of an aisle wall, decorated with blank arcading (Fig. 146) corresponding to the arcade
of the interior. Above, it was finished off by a plain brick stringcourse which,
with another course of the same kind, must originally have enclosed a saw-tooth
cornice.

In one section of the arcading a large, recessed, unsplayed window opened,

1 Romualdo, f'/avia Papla sacra.

H 2

IOO

LOMBARDIC ARCHITECTURE

not however of such wide dimensions as had been
in use at Ravenna before the second half of the
Vlth century. The architect of Santa Maria delle
Caccie in all probability came from Ravenna, inas-
much as the church was Ravennate in style, a fact
which is shown by the decorative use of blank
arcading on the outer walls of the aisles.

As we shall see presently, the Comacine gilds
of the VHIth century used to relieve the exterior
of the side walls of the aisles in their churches, not
with blank arcades, but with arched corbel courses
divided into groups by lesenas. And the windows
which they constructed in these side walls were in-
variably of very small dimensions, and splayed both
inside and out.

Fig. 146. Pavia. Santa Maria delle
Caccie (744-749).

We will conclude this chapter by dealing briefly

with three very well known monuments of the Lombard period, the origin of
which has given rise to very diverse opinions, but which I ascribe to craftsmen
of the School of Ravenna.

THE TOMB OF THEODOTA. The date of this sarcophagus (Fig. 147), which once
contained the mortal remains of the Theodota who fell a victim to the passion of the
Lombard king, Cunibert (688-700), is to be placed in the first half of the VII Ith
century, or, more precisely, about the year 720. The two sides and one of the ends
are preserved in the Museum at Pavia.

Its carvings, among which the two peacocks drinking at a two-handled vase
surmounted by a cross may be compared with a similar subject carved on the

G-i

I -rrT VI

r x^

Fig. 147. Pavia. Museum. Side of the Tomb of Theodota (about 720).

sarcophagus of John V, archbishop of Ravenna (about 725 or 742-752), in Sant'
Apollinare in Classe (Fig. 148), are not to be set down as a work of the Comacine
masters, or even the best of them, though executed in the capital of the kingdom of
Lombardy. For the Comacine artists of the Lombard period, in their sculptured
panels, show all the want of spontaneity of a craft learnt in the " laborerii," with the
engraved lines of triangular section made by the chisel, and used indiscriminately in

THE SCHOOL OF RAVENNA 101

all the ornamental and geometrical treatment of the parts not left plain, characteristics
which are not found in the tomb of Theodota. Really the carvings are to be ascribed
to artists of the School of Ravenna, among whom the decorative carving of panels in
the Vllth and VHIth centuries has nearly always a languid manner, and does not so
often become mere cutting without any roundness of modelling ; and even when it is
so, there is not that clear-cut effect produced by the Comacine masters. The truth of
this may be easily verified by anyone who understands the subject, if he will examine
the carved sarcophagi of that period preserved in Sant' Apollinare in Classe. They
may have come from the same hand as that which designed and executed the
archivolts of the baptistery of Callistus at Cividale (VHIth century).

Fig. 148. Classis. Sant' Apollinare. Sarcophagus of John V (725 or 742-752).

With reference to this I may observe that we must not be surprised at finding
in the carving of this tomb an art decidedly superior to that which we find in the
carvings of the contemporary sarcophagus of Archbishop Felix of Ravenna (708-724),
also preserved at Sant' Apollinare in Classe. The latter sculptures are regarded by
many as evidence of the serious artistic decadence of the Vlllth century. But it is
more than likely that, in the days of the famous Liutprand, the best carvers, as well as
the best mosaic workers of Ravenna, emigrated to the adjacent kingdom of Lombardy,
whither they were attracted by the considerable number of works in course of execu-
tion, some of them of an important character.

THE BAPTISTERY OF CALLISTUS IN THE CATHEDRAL OF CIVIDALE IN
FRIULI was erected by Callistus, patriarch of Aquileia, after he had moved the see, in
the year 73O, 1 to Cividale. It was rebuilt after 1000, as is shown by the spurred base
of one of the columns. Before this, it seems that it had been restored by another
patriarch, Sigualdus (772-776).

Of the structure of Callistus there remain unquestionably the seven carved
archivolts, as well as the eight capitals on which they rest, and perhaps some of the
fragments of plutei with which the base of the erection is partly constructed (Fig. 149).
Both in execution and design these archivolts are so close to the carvings of the tomb
of Theodota at Pavia, that we might suppose them to come from the same hand.

1 Gams, op. fie.

IO2

LOMBARDIC ARCHITECTURE

The capitals, midway between Corinthian and Composite, have two rows of acanthus
leaves rather accentuated. These capitals, though very coarse in design and
execution, nevertheless in the form of the leaves, and the way in which they are
defined, point to the school which produced those in Santa Maria in Valle
(762-776).

Of the period of the patriarch Sigualdus there survives a plutens which
forms one side of the base of the baptistery (Fig. 150). On this slab we find

a design (a cross be-
tween two candle-
sticks, with palms
and roses in the
unoccupied space)
which had been
familiar to the
artists of Ravenna
from the Vlth cen-
tury onwards. We
also meet with the
other motive of a
conventional tree
ending in a kind of
lily flower, with
lions' heads issuing
fr o m its side
branches. It had
been already used,
in the early years
of the Vlllth cen-
tury, in one of the
long sides of the
tomb of Theodota
at Pavia (about
720). So that we
shall not be far from
the truth if we refer
this pluteus to the
same school.

There may be
observed in it a

marked difference between the treatment of the conventional ornament, which is not
without a certain grace, and the elementary way in which the animals are modelled,
with the exception perhaps of the doves, and, still more, the manner in which the
angel is represented. It would be impossible to imagine anything more clumsy
and barbarous.

To the time of the same patriarch may be assigned the fragment of a
pluteus showing two square compartments occupied by symbols of the Evangelists,
and also another exhibiting a wheel of lilies, closely related to some marble
carvings in Santa Maria in Valle, though the latter are the expression of a better
ornamental design.

Fig. 149. Cividale. Cathedral. Baptistery of Callistus (Vlllth Century).

THE SCHOOL OF RAVENNA

103

It must clearly have been a carver of the same school who produced the
altar executed for the Lombard king, Ratchis (744-749), and still to be seen in

T
/

Fig. 150. Cividale. Cathedral. Baptistery of Callistus (Vlllth Century).

Fig. 151. Cividale. San Martino. Altar of Ratchis (744-749).

the church of San Martino (formerly San Giovanni Evangelista) at Cividale
(Fig. 151).

The carvings at Pavia and Cividale which we have just examined suggest
a few comments. Many fancy that they are the productions of Greek chisels.

io 4

LOiMBARDIC ARCHITECTURE

In formulating, however, this opinion, they have not only taken no account,
in the case of carved panels, of the style of composition and technical execution
characteristic of the Byzantine School at that period (a subject that will occupy
us when we come to deal with Pre-Lombardic carving, and endeavour to
dissipate another myth about Byzantine art and artists in Italy), but they have
also disregarded various reasons pointing in an opposite direction, which I will
forthwith state.

First and foremost, it is inconceivable that the Lombards should have
availed themselves of the services of Greeks for the buildings which they erected,
or indeed have entrusted any kind of work to them, because, as Cordero 1 rightly

Fig. 152 Corneto Tarquinia. Archaic Etruscan Carving.

observes, the Greeks were the sworn enemies of the Lombards ; they were hardly
ever at peace with one another ; and the Lombards always preferred to employ
the artistic services of their own subjects rather than those of a hostile people.
By this preference they secured two things. They made it impossible for the
Eastern Empire to employ Greek artists as political emissaries, and they demon-
strated to their Italian subjects that the rule of their Northern masters was
not only less cruel and rapacious than that of the Greeks, but that even the arts
prospered under it.

On the other hand, it was quite natural that the Lombards, anxious as they
were to get possession of Ravenna, especially in the time of Liutprand, who
captured and held it for a short period, should, with a certain amount of worldfy
wisdom, engage the services of the craftsmen of the place. And it is easy to
understand that, after Aistulf (749-756 : like Ratchis, a son of Pemmo, Duke of
Friuli) had once more taken Ravenna (752), thus putting an end to Greek
rule in the Exarchate, and up to the day when the Lombards were compelled by
King Pippin to abandon for ever their newly conquered territory (755), the artists of
Ravenna contributed their skill to the execution of the works ordered by the

1 Op. tit.

THE SCHOOL OF RAVENNA

105

Lombards within their dominions. Again, it is reasonable to suppose that these
artists, brought into contact with the members of the Comacine gilds, became
familiar with them, and were associated with them for some time in the execution of
important works. Lastly, it must be realised that the features which these writers
regard as constituting the
Byzantine style, and there-
fore indicating the presence
of Greek carvers, were no
foreign importation, but
rather a new creation of the
artists of Italy, suggested by
models' provided by the an-
cient monuments in the
peninsula, and adapted to
new times and new needs,
if indeed they were not due
to their unaided invention.

Thus, for instance, the
motive of squares enclosing
figures of saints, symbolical
animals, birds, fish, &c., ob-
viously owes its origin to the
design of squares formed by
cable mouldings containing
goats, horses, lions, telamons,
flowers, winged sphinxes,
stags with animals on their
backs biting their necks, ani-
mals pursuing one another,
various kinds of birds, gor-
gons, sea-horses, minotaurs,
&c., used by the Etruscans
for decorative purposes, and
to be seen among the archaic
sculptures at Corneto Tar-
quinia, either preserved on
the spot (Fig. 152) or else
transferred to the Archaeo-
logical Museum at Florence. Fig. 153. Rome. Laternn Museum. Mosaic (Illrd Century).
Another source may be the

scheme of compartments with human figures, birds, fish, knots, &c., sometimes
employed by the Romans in mosaic work (Fig. 153).

The motive, again, of griffons and animals biting themselves, and of large fishes
attacking small ones, which Cattaneo l would bring to Italy as an importation from
the East in the Vlllth century, was really imitated from Roman work. In particular,
precisely the latter form may be seen represented in the spandrels of a shrine from
Todi in the Galleria Lapidaria of the Vatican Museum 2 (Fig. 154).

The most striking characteristic of Vlllth century carving, interlacing, had been

Op. dt.

Amelung, Die Sculpturcn des Vatieanischen Museums.

io6

LOMBARDIC ARCHITECTURE

used by the Romans not only on vases and domestic utensils, but also in architectural
decoration, as also, and more particularly, in mosaics. This may be verified by any-

Fig. 154. Rome. Vatican Museum. Aedicula.

one in museums, in the early Christian Catacombs, and in buildings of the Imperial
age. And before the Romans it had been used by the Etruscans.

Again, for such ornamental motives as roses, rosettes, whorls, stars of six or eight
points, lilies, pellets, round arches intersecting so as to produce pointed arches, the

Fig. 155. Ravenna. San Francesco. Sarcophagus (IVth Century).

bead and reel ornament, vine branches laden with grapes and birds pecking at them,
&c., it is Roman monuments which provide the earliest models in Italy.

The favourite scheme of the Ravennate sculptors, a colonnade, or isolated arches,
very interesting specimens of which are found on a IVth century sarcophagus in San

THE SCHOOL OF RAVENNA 107

Francesco at Ravenna (Fig. 155), and on some of the sarcophagi in Sant' Apollinare
in Classe, framing at first figures of the Redeemer and the Apostles, and later, crosses,
wreaths, palms, sheep, doves, &c., was suggested by the sarcophagi with colonnaded
fronts, which made their appearance in Italy in the first centuries of the Christian era.
Finally, the motive of "cauliculi " or slender volutes, which the Comacine masters
were the first to use at the tops of ciboriums and arcaded altar frontals, is nothing but
a decadent reminiscence of the Etruscan and Roman recurring volute (" corridietro ").

CHAPTER II

THE COMACINE MASTERS

expression " magistri Comacini " appears for the first time in the
code of the Lombard king, Rotharis (636-652), where, in the laws num-
bered CXLIIII and CXLV, 1 they figure as master masons with full
and unlimited powers to make contracts and sub-contracts for build-
ing works ; to have their collegantes or " colleagues " partners, members of the
gild or fraternity, call them what you will and lastly, their serfs (servi) or workmen
and labourers. 2

Many and various are the views of writers, both in Italy and outside it, with
regard to the etymology of the name. The most plausible theory is still that which
derives it from the diocese of Como, including, as it did in those days, the districts
of Mendrisio, Lugano, Bellinzona, and Magadino.

This corporation of architects, builders, carvers, and workmen, rather less than
a century later, forms the subject of the " Memoratorio de mercedes Comacinorum '"
or schedule of pay of King Liutprand (7I2-744), 3 which provides some interesting
data for the history of architecture in Italy owing to certain enactments contained
in articles CLVIII, CLX, and CLXII, and relating, not only to architecture, but
also to carving, as the last mentioned article shows.

The origin of the Comacine masters in the diocese of Como is explained quite
naturally, according to De Dartein, 4 Merzario, 5 and others, by the custom, which has
always existed, among the craftsmen and workmen of that region, of leaving their
native places in order to betake themselves in gangs wherever building works are
about to be or have been begun, urged thereto by their barren mountain soil,
pecuniary gain, their innate ability and enterprising character. Another explana-
tion is to be found in the presence on the shores of the lakes of Como, Lugano,
and the Maggiore, of numerous stone, marble, and timber yards, which furnish building
material for the cities of the plains. These yards gave scope for the practice of the
crafts of carver, carpenter, builder, &c. ; and these, in their turn, by constant practice
and continuous progress, ultimately developed architects and sculptors.

And here we may naturally feel surprise at the appearance, amid the darkness
of the early centuries of the Middle Ages, of a corporation of craftsmen who, though
of Roman origin, none the less enjoyed Lombard citizenship and the rights belonging
to it ; while the Roman or Italian subjects of Lombard rule were, if not slaves, nothing
better than "aldi," that is to say midway between freedmen and serfs manumitted

1 Historiae patriae monutnenta edita jussu regis Caroli Alberli EJicta regum Langobaraorum Edictunt
Kotharis regis.

2 Troya, Codice diplomatic/) longobardo. 3 Hist, patriae man. Edicttim Liutfrattdi regis.
* Op. fit. 6 I maestri comae ini.

1 08

THE COMACINE MASTERS 109

on the condition of performing the manual tasks assigned them by the manumittor.
A corporation too, which had a legal monopoly of public and private building work
within the territories occupied by the Lombards, as the Code of Rotharis proves, and
can claim the honour of filling up the gap which for so long was believed, especially
by non-Italian writers, to exist between the incorporated artisans of the Roman
epoch, supposed to have vanished with the fall of the Empire, and the gilds of crafts-
men which sprang up so luxuriantly in the Xlllth and XlVth centuries.

Such surprise, however, may easily be allayed if we consider that in reality the
fraternity of craftsmen, in Italy at least, by no means came to an end with the
Barbarian invasions, 1 and particularly that of the Lombards, who actually preserved
those Roman institutions which best fulfilled their aim of keeping the conquered
people in subjection. Accordingly, they would have maintained the corporations of
artisans, in order to make the exaction of tribute easier, and at the same time to' be
able to keep a hold over the individuals composing them. 2

It has been pointed out, on the strength of a passage in Cassiodorus, that, under
the Goths, there existed magistrates attached to the corporations connected with the
supply of corn, a fact which suggests that the Roman system was preserved under
the Barbarian dominion. Two letters of Gregory the Great (590-604) prove the
existence, at the close of the Vlth century and the beginning of the Vllth, of
a corporation of soap-makers at Naples, and of another of bakers at Otranto. Grego-
rovius 3 states that, in the time of Pope Hadrian I (772-795), not only did there
exist in Rome the associations of milites.peregrini, notaries, and the Papal singers,
but that there must also have been others of doctors, craftsmen, traders, and workmen
of every description.

Hence we have good grounds for inferring that the corporation of "Comacini,"
who apparently were neither more nor less than the successors of the master masons
who in the days of the Empire had directed the operations of the collegia specially
devoted to building, survived the barbarian invasions which were so disastrous to
Italy in the centuries preceding the accession of Rotharis to the Lombard throne.
This view is confirmed by the undoubted fact that from this time onwards the
" Comacini " formed a very important gild, as is shown by the need which he felt of
making regulations for it in his laws. This gild cannot have sprung into existence
full grown, and, as it were, by magic, just when the Code of Rotharis made its
appearance in 643. It must have already been in existence, and have attained some
degree of importance, well before Alboin's descent on Italy (568). Troya, 4 in fact,
remarks that when the Lombards of the time of Autharis (583-590) and of Agilulf
and Theodelinda (590-625) wanted to erect buildings, they must have made use of
it ; and that everything leads one to think that before the promulgation of the Code
of Rotharis some of the members (i.e. those of the highest capacity and reputation)
had already been enfranchised by " impans " or express grace of the king. However
that may be, the mention of the associations of Comacini in the reign of Rotharis and
Liutprand is one of the earliest in the Barbarian world, and earlier than that of any
gild of architects or builders belonging to the Middle Ages.

We know nothing about the organisation of these associations, and any state-
ments made by writers with reference thereto are mere conjectures. The same may
be said of various terms connected with them, such as the "laborerium," the
" schola," and the " loggia," " loya," or " loia " ; for these names only made their

1 Leo, Storia degli Stati Italiani. * OrlanJo, Delle fratellanze artigianein Italia.

8 History of the City of Rome in the Middle Ages. * Op. fit.

no LOMBARDIC ARCHITECTURE

appearance after the year looo. 1 Nor have we any documents to show whether
Charles the Great (768-814), after becoming king of the Lombards (774), maintained,
diminished, or abrogated the rights or privileges bestowed on the Comacini by the
Code of Rotharis and the " memoratorio " of Liutprand. Amico Ricci, 2 indeed, states
that the Popes, after the removal of the fears inspired by the Lombard dominion, not
only confirmed to the Comacine masters the privileges which they had obtained in
their own country from the Italian kings, but further secured those privileges for them
in all the Catholic countries whither they were led by the objects of their associations.
He goes on to assert that in the Empire of Charles the Great these associations were
exempted from obedience to every local law, statute, and obligation, and were also
empowered to fix the scale of payments, and in their chapters-general to settle
without interference everything connected with their internal government. But these
assertions find no confirmation either in Papal Bulls, the Acts of the Carolingian
kings, or in the best known annalists.

Another piece of mere hypothesis is the idea of those who, with the same writer,
argue that in the days of Charles the Comacine masters formed themselves into
closer unions, with their own peculiar regulations and ceremonies kept as a profound
secret ; that they began to be called " free " or " frank masons," and that from these
associations were derived the societies specially known as Freemasons, who spread
through Italy, Germany, Switzerland, Provence, Spain, England, and Scotland, and
were the origin of the Masonic Lodges, at first composed solely of architects, builders,
and the workmen associated with them.

Nor can anything better be alleged for the theory that, after the fall of
the Lombards, the Comacini founded a " School " at Rome with the object of sending
not only their younger, but also the older members there, to study the surviving
monuments of the ancient world. A confirmation of this is supposed 3 to exist
in the fact that the "Liber Pontificalis" describes how, when Pope Leo III
(795-816) returned to Rome after taking refuge with the Duke of Spoleto, there came
forth to meet him at the Milvian Bridge the " Schools " of foreigners in Rome :
" scole peregrinorum, videlicet Francorum, Frisonorum, Saxonorum, atque Lango-
bardorum." 4 It is true that Gregorovius 5 finds that, at the end of the VHIth
century, there were in existence at Rome, besides various local associations,
the " Schools " of foreigners " Scholae Peregrinorum " an institution of a different
kind. He also notes that the oldest of these foreign corporations was that of the
jews " Schola Judaeorum" in the Trastevere ; next in order came that of the
Greeks" Schola Graecorum " which had its centre near Santa Maria in Cosmedin ;
and, lastly, came the " Schola Saxonum," the " Schola Francorum," the " Schola
Frisonum," and the " Schola Langobardorum." But we are also told that the
" Schola Saxonum " founded by the King of Wessex, when he came on pilgrimage
to Rome in 728, 6 had as its object the instruction of Saxon chiefs and people in the
Catholic faith, from which it may be inferred that the " Schola Langobardorum,"
which is believed not to have been instituted till after the fall of King Desiderius
(774), had a similar purpose. The same view is taken by Dyer, 7 who thinks that
the Scholae Francorum, Frisonum, Langobardorum, and also the Schola Saxonum

1 Merzario, op. fit. 2 Sloria delF architettura in Italia dal secolo IV al XVIII.

3 Merzario, op. cit. * Duchesne, Le liber pontificalis.

5 Op. cit.

8 Rerum Britannicarum medii aevi scriptores. Rolls Series The Anglo-Saxon Chronicle.

7 A History of the City of Rome.

THE COMACINE MASTERS in

were originally intended solely for the religious instruction of the nations to which
they belonged, but that they disappeared between the IXth and Xlth centuries, or
at least shrank into mere hostels for the reception of poor pilgrims, and burying
places for the respective nations. On these " Scholae " and their churches an
important study has lately appeared. 1

Whatever may have been the organisation of the Comacine or Lombard gilds,
and however these may have been affected by outward events, they did not
cease to exist in consequence of the fall of the Lombard kingdom. With the first
breath of municipal freedom, and with the rise of the new brotherhoods of artisans,
they too, perhaps, may have reformed themselves like the latter, who were nothing
but the continuation of the " collegium " of Roman times preserving its existence
through the barbarian ages, and transformed little by little into the mediaeval
corporation. The members may have found themselves constrained to enter into
a more perfect unity of thought and sentiment, to bind themselves into a more
compact body, and thus put themselves in a condition to maintain their ancient
supremacy in carrying out the most important building works in Italy. But we
cannot say anything more. And even putting aside all tradition, the monuments
themselves are there to confirm what we have said.

Finally, towards the end of the Xlth century, the Comacine brotherhoods began
to relax their bonds of union, to make room gradually for personality, and for
artistic and scientific individuality, till at length they vanish at the close of the XVth
century with the disappearance of the Lombardic style which they had created, and
the rise of the architecture of the Renaissance. 2

1 P. Ehrlc, Ricerche su alaine antiche chiese del Sorgo di S. Pietro.

2 Archivio storico delF arte, Anno II. Carotti, Viceiide del diiomo di Milano.

CHAPTER III

THE PRE-LOMBARDIC STYLE

FROM THE REIGN OF AUTHARIS TO THE FALL OF THE KINGDOM OF LOMBARDY

AS late as the reign of Autharis (583-590), the Lombards and other
Northern barbarians who had descended upon Italy in the year 568, did
nothing but plunder the churches belonging to the conquered people whose
cities they destroyed. Paulus Diaconus 1 supplies us with the evidence of
the fact, and the comments of Troya 2 on the CCXLVIIIth law in the Code of
Rotharis (636-652) vouch for its truth. Autharis, indeed, began some constructive
work ; but he was prematurely carried off by poison, and his church at Fara Berga-
masca, erected for Arian as opposed to Catholic worship, is the only building which
history records as erected by his order. Brighter days, however, were in store for art.
On the death of Autharis, his widow Theodelinda (590-625) imposed her own
form of creed on the Court, induced her second husband Agilulf (590-615) to embrace
it, and their son Adaload was also baptised in it. Thereupon the whole Lombard
nation, following the example of its rulers, was received into the Roman Church, and
religious zeal soon multiplied the number of places of worship and monasteries. The
queen took the lead in these works of piety, and may be truly said to have rekindled
in the districts subject to Lombard rule the dying embers of the Fine Arts. In fact,
though it is impossible to believe that all the ecclesiastical or even secular buildings
attributed by tradition to her or to Agilulf were actually due to them, still the
number which may be accepted as such is considerable. And after Theodelinda
there was not a single Lombard sovereign, whether Arian or Catholic, that did not
help by means of some work or other to keep alive, so far as was possible in that age
of barbarism, the spirit of the Fine Arts, and more particularly architecture, for the
practice of which they provided frequent opportunities.

It is true that the Liber Pontificalis 9 describes the Lombard kings as " protervi,"
" perfidi," " pestiferi," " atrocissimi," " sceleratissimi," " crudelissimi," and so forth. Yet
one of the kings, Liutprand, had natural instincts of piety and virtue which were not,
according to some historians, Prof. Oman 4 among them, the least efficient cause of the
evils which Rome was at that time on the point of bringing upon Italy. For the
Papal biographer forgot that it was he who, by first taking the town of Sutri and
then presenting it to the Pope, laid the foundation of the Temporal Power.
However, epithets such as we have quoted are powerless to obliterate the memory of
the numerous religious buildings due to the piety of the kings themselves or of their
officials.

1 Man. Germ. Hist. Pauli historia Langobardorum. 8 Op. fit.

3 Duchesne, Le liber pontifaalis. 4 The Dark Ages.

112

THE PRE-LOMBARDIC STYLE 113

This piety perhaps had its origin in policy. Indeed, not a few students of
Lombard history hold that many of their princes regarded the restoration of old
churches, and still more the building of new ones, as an instrument of government,
the intention being to satisfy the people by these displays, and make them see that, if
their new masters were steadily pressing the Papacy closer and closer, they were at
the same time indefatigable supporters of the religion of their conquered subjects.
But if this is so, it must be also conceded that it was not religious devotion but cool
political calculation that led Charles the Great (" the most benignant," " the most
excellent," " the most Christian king," as he is called by the Papal biographer cited
above) and his heirs to make donations and grant privileges to the clergy and
monasteries.

Of the buildings erected by the Lombards during their sway in Italy, and definitely
recorded by Paulus Diaconus, as well as of those which are or can be assigned to that
period on the strength of documentary evidence, or historical notices, or tradition,
either not one stone remains upon another, or else a remorseless criticism and recent
discoveries have disposed of their claims in such a summary manner that hardly one
has survived the ordeal.

For example, among the instances of buildings, religious as well as secular,
brough tforward by Cordero, 1 who, we may remark, was the first to demonstrate by his
fearless criticism the untenability of the dates freely assigned in his day to structures
belonging to the Middle Ages, the only one that has not been struck off the
list is the church of San Salvatore at Brescia. Again, of those cited by De
Dartein 2 two only, the churches of Santa Maria delle Caccie at Pavia, and San
Salvatore at Brescia, are at present recognised as belonging to the same period. To
them I have now added a third, Santa Maria in Valle at Cividale. Later, I shall add
two more : the parish church of Arliano near Lucca, and the basilica of San Pietro at
Toscanella.

More fortunate have been the buildings dealt with by Cattaneo, 3 viz. the churches
of Santa Maria delle Caccie at Pavia (744-749), San Salvatore at Brescia (753), and
the church at San Giorgio in Valpolicella (712-740). And the same may be said of
the church of Santa Teuteria at Verona, consecrated in 751 4 and remodelled in 1 160
when it was re-consecrated, with the addition of the present cupola, a fact which I
was able to verify when the masonry was recently laid bare. The antiquity of the
first two has now been admitted, and the others have not yet fallen under the blows
of criticism, or, what is more important, the logic of facts.

Of the buildings accepted by all critics alike, San Salvatore at Brescia is still the
one which, by common consent of the best authorities, is regarded as the most
important, and for the following reasons. It remains very nearly in its original
condition ; it is the only one, informed by a single idea, which exhibits the workman-
ship and the style of the Lombard age ; and, lastly, it would not be easy to discover
other buildings presenting these characteristics.

Nevertheless, the results of my own studies and researches are not in complete
agreement with these views. Indeed, I believe that San Salvatore at Brescia has
been given an importance which it does not really possess, and that there are two
other buildings in which archaeologists and art-historians might have found a much
safer guide for information as to the architectural characteristics of the period. These
are the basilica of San Pietro at Toscanella, and the parish church of Arliano near

1 Op. cit. " Op. cit. ' Op. cit.

4 Biancolini, Notizie storiche delle ehiese di Verona.
VOL. I I

114

LOMBARDIC ARCHITECTURE

Lucca. They mutually supply one another's deficiencies, for the one can show those
parts of the original structure which are wanting in the other, where they have been
destroyed or tampered with ; and they provide a reliable example of the architectural
style in general use in the districts subject to Lombard rule.

To this style I give the name of " Pre-Lombardic " ; and I do so, not because I
want to invent a new appellation for the family which includes the monuments not
only of the Lombard period, but also those erected in the regions occupied
by the Lombards during the two centuries preceding the year 1000. The reason
rather is that these monuments, while owing something to the Roman and Romano-
Ravennate styles for we find in them the organic construction of the former and the
decorative motives of the latter at the same time present features both constructive
and decorative which are foreign to those styles. These features are absolutely new,
and form marked characteristics of the style of the later Lombardic basilica
which afterwards influenced all the Christian architecture of central and northern
Europe.

This Pre-Lombardic style originated under the Lombard rule. Slowly but
surely, through the influence of Roman, Rbmano-Ravennate, and Byzantino-
Ravennate architecture on the Comacine or Lombard masters, with the addition of
certain new elements which formed part of their natural inheritance, it advanced
toward the " Lombardic" style, properly so called, of which it was the precursor, and

for which it prepared the way. And in
all its phases it represents the develop-
ment of the style which, after attaining
its completion in Lombardy in the
course of the Xlth century, spread over
so many regions of Europe, where it
exercised undisputed sway until the
" Pointed Style " came to supplant it.

Having said so much by way of
preface, we will now turn to examine,
in chronological order, the three build-
ings referred to. Only, there is a
fourth which we ought to take before
them, the crypt of the church of Sant'
Eusebio at Pavia ; for though it is not
so old as the Lombard period, it con-
tains valuable evidence about the
carving of that age.

CRYPT OF THE CHURCH OF SANT'
EUSEBIO AT PAVIA. We know from
Paulus Diaconus that the original
basilica of Sant' Eusebio, the founda-
tion of which is assigned to a time
earlier than the Lombard dominion, 1
was in existence in the days of
Rotharis (636-652) and dedicated to Arian worship. The latter circumstance is
an argument for the rebuilding or radical restoration of the church in the reign

1 Romualdo, op. cit.

Fig. 156. Pavia. Crypt of Sant' Eusebio
(Vlth or Vllth Century).

THE PRE-LOMBARDIC STYLE

of Autharis (583-590), the unflinching champion of Arianism. What is certain
is that the building of the time of Rotharis was not the original one, for it is impossible
to believe that, previous to the descent of Alboin (568), the art of carving in Italy had
fallen to the degraded level which produced the results to be seen in the crypt, the
only part of the church which escaped the rebuilding in the early years of the
XVII Ith century.

This crypt (Fig. 156) is a small basilica, properly orientated, below the apse and
presbytery. It has cross vaulting with visible arches. Two of the bays at the end,

Fig- 157. Pavia. Crypt of Sant' Eusebio.
Capital (Vlth or Vllth Century).

Fig. 158. Pavia. Crypt of Sant' Eusebio.
Capital (Vlth or VI Ith Century).

beneath the apse, have ribs, so that they must be later than the year 1000. The
vaulting springs from wall piers and six isolated columns, some of tufa, the others of
marble taken from older buildings, with four others which .have been made for their
present position. The latter are square in section with the corners rounded off", and
form one piece with the capita! which has the shape of an inverted truncated pyramid.
In every instance the base is buried beneath the surface.

All the columns, those that have been brought from elsewhere as well as those
specially made with capitals in one piece, carry very barbarous marble capitals with, at
each angle and on each face, an unribbed leaf, rude and stiff (Fig. 157); or else a
simple hollowing out at the angles ; or, thirdly, a row of leaves like those first
described, with a similar row below them, those at the angles being inverted
(Fig. 158).

These capitals, two of which, viz. those with the lower leaves at the angles
inverted, have in their design no counterparts among the many and varied capitals of
the Pre-Lombardic style which are known to me, proclaim themselves as the work of
one hand, and were obviously made for the crypt of the first church and then used

I 2

LOMBARDIC ARCHITECTURE

over again when, in consequence of the famous earthquake of 1117, so many churches
in Pavia were, as I believe, restored or rebuilt. Or the reconstruction may have taken
place when, at some time after the epoch of 1000, it was desired to raise the level
of the presbytery, thus giving more space to the crypt ; the result being obtained by
rebuilding the vaulting on stilted arches and, presumably, raising the pavement of the
crypt.

Their design and execution are so rude that they seem archaic beside the
barbarous but still superior Pre-Lombardic capitals of the VHIth century ; so that I
think we shall not go far wrong if we assign them to the period between 583, the year
in which Autharis ascended the throne, and the reign of Rotharis (636-652). They
tell us how carving had degenerated during the first half of the Vllth century in
the lands subject to Lombard rule, and how rude and unskilful were the artists
produced by the Comacine gilds at that period. At the same time they reveal the
fact that a new art was coming into being, showing itself at first in a somewhat timid
and barbarous guise, but always original. It was the mission of this new art to replace
the Ravennate and Byzantine styles in Italy.

PARISH CHURCH OF ARLIANO NEAR LUCCA. The church of San Martino
at Arliano has not, so far as I know, been hitherto noticed in the history of art.

Fig. 159. Arliano. Parish Church (712-744)-

The precise date of its erection is not known. It is mentioned, however, as early
as 892 in a document which speaks of it as a parish church existing from an indefinite
period. 1 At the same time, its construction and architectural decoration, compared

1 Metnorie e docitmenti per servire all' Istoria del Ducato di Lucca. R, Accademia Liiccliese di Sciense Lettere
ed Arli.

THE PRE-LOMBARDIC STYLE

117

with those of the oldest churches of Lucca, the date of which is certain, point unques-
tionably to the time before the epoch of 1000. On the other hand, they appear archaic
when compared with those of churches of known date erected in North Italy in the
course of the IXth and Xth centuries, such as the parish churches at Agliate and San
Leo, and the basilicas of
San Vincenzo in Prato
and San Celso at Milan.
Hence we may place the
church of Arliano in the
Lombard period, and, by
a process of elimination,
after the erection of
Sant' Eusebio at Pavia
probably in the VHIth
century and the reign
of that great church
builder, Liutprand (712-
744): "Hie gloriosissi-
mus rex multas in Christi
honore per singula loca
ubi degere solebat
basilicas construxit." l
But it must be earlier
than San Pietro at Tos-
canella, the architectural
decoration of which
shows an advance be-
yond that at Arliano.
It is a basilica with nave
and two aisles separated
by four rectangular piers,
from which spring round
arches. At the eastern
end of the nave, and
starting immediately
from its termination, is
the semicircular apse. It
is worth while mention- Fig. 160. Constantinople. Inner face of the Golden Gate (408-450).

ing that the plan has

not the oblong shape common to churches of the old Latin type, but rather takes the
form of a square, each side measuring about 55 ft. In the next place, the left aisle,
like that of San Vittore at Ravenna, is wider than the right. The origin of this
inequality is perhaps to be found in the fact that, the former being assigned to the
women, it was found necessary to give them more room than the men, who
according to the Roman rite, had their places in the south aisle, known as the
" pars virorum."

Originally both nave and aisles had open timber roofs, but these have been
replaced by more recent vaulting. At the same time, I suppose, the existing piers

1 Man. Germ. Hist. Pauli historia Langobardorum.

LOMBARDIC ARCHITECTURE

were constructed by encasing the original supports, probably columns taken from
older buildings, and thus giving them the form of piers.

Externally, the facing of the walls consists for the most part of coursed and
worked stones of various dimensions, evidently taken from some older building.

Fig. 161. Arliano. Parish Church. Apse (712-744).

The front (Fig. 159), almost untouched in spite of its age a condition presented
by no other church of the Lombard period, is turned towards the west, and
has three openings for the doors corresponding to the nave and aisles, which are,
moreover, indicated on the outside by two lesenas projecting from the facade. The

THE PRE-LOMBARDIC STYLE

119

middle door, the jambs of which have been rebuilt, is strictly rectangular, and the lintel
is relieved by a round arch in which is sunk a lunette wider than the opening of
the door itself. This arrangement, containing in itself the germ which, when
developed in course of time, produced the typical Lombardic portal, suggests a few
comments.

The rectangular doorway with its lintel relieved by an open arch is a fairly ancient
invention. Examples of it are to be found in the Forum of Augustus at Rome,
finished in 2 B.C. 1 The idea, too, of a rectangular doorway surmounted by a monolithic
lunette flush with the wall, an instance of which is offered by an Etruscan tomb at
Cortona, 2 is also old. Of Roman date is the arrangement of a doorway surmounted by
a recessed lunette. Early instances of this, occurring in decorative forms, are afforded
by certain sepulchral monuments in Phrygia, believed to belong to the age of the Anto-
nines 3 or even earlier. 4 Very early specimens in actual construction are to be seen on the
inner face of the Golden Gate (Fig. 160), and in the Gate of Rhegium, in the Theodosian
Walls of Constantinople, which are dated by Van Millingen 5 in the reign of

Fig. 162. Arliano. Parish Church. Corbels (712-744).

Theodosius II (408-450). Each has a sunk lunette intended to hold an icon. In
Italy, on the other hand, the oldest instance that I can cite of a square-headed
opening crowned by a recessed lunette, is to be found in the windows of the
mausoleum of Galla Placidia at Ravenna (about 440). So that its invention must be
credited to the builders of the East.

The doorway with its lunette is set in a projection beyond the external face of the
nave wall. The eaves cornice of the fa$ade is composed of a continuous stepped
arched corbel course, while lower down, at the sides, the walls are decorated with a
similar course broken by lesenas. The latter form of decoration is also applied to the
side walls, and to the east end of the church and its apse (Fig. 161). Some of the
arches of these courses spring from corbels rudely carved with projections, striations,
diamond faceting, and barbarous heads of living beings (Fig. 162). The walls of the
church were originally pierced by very narrow round-headed windows splayed on both
sides, and also by round openings and luminous crosses.

These figure corbels at Arliano show that the fashion, prevalent in the decadence
of classical Roman art, of representing real or imaginary beings on the face of consoles
supporting the topmost cornice of a building or the architrave of a door, in the manner,

1 Lanciani, The Ruins and Excavations of Ancient Rome.
3 Texier, Description de FAsie Atineure.
5 Byzantine Constantinople.

2 Martha,

* Perrot et Chipiez, of. cit.

I2O

LOMBARDIC ARCHITECTURE

for instance, to be seen in the baptistery of the cathedral at Spalato (Fig. 163), which
is believed to have been originally a temple dedicated to Jupiter Capitolinus, 1 but in

Fig. 163. Spalato. Door of Baptistery (about 300-305).

any case formed part of the palace of Diocletian, and also on the Golden Gate of that
palace, did not originate, as some think, with the birth of the Lombardic style, but
was inherited by it from the Pre-Lombardic.

The use of consoles of this kind seems to have begun in the reign of Diocletian.
From an anonymous drawing published by Hulsen 2 it appears that the travertine

Fig. 164. Rome. House of Nicola Crescenzio (Xllth Century).

consoles belonging to the pediment of the Curia of Diocletian at Rome, dating from
the first years of the IVth century, were decorated with acanthus leaves and dolphins

1 Jelid, Bulic e Rutar, op. cit.

Die Ausgrabwigen atif dein Forum Romanum.

THE PRE-LOMBARDIC STYLE 121

with intertwining tails, modelled in stucco. Into the front and south side of the well-
known house of Nicola Crescenzio (Xllth century) at Rome, popularly known as the
. house of Cola di Rienzo, or of Pilate, are built a large number of consoles taken from
older buildings, and to be assigned to the last years of the Illrd century and the early
ones of the IVth (Fig. 164). Their faces are carved with figures, with or without
wings, single or in pairs, and in some cases supporting an animal or a basket filled
with fruit.

Secondly, the luminous crosses show that the Lombard builders had adopted
this form as early as the VHIth century. They borrowed it from Ravenna, where it
had been used as far back as the first years of the Vlth century in the mausoleum of
Theodoric.

BASILICA OF SAN PIETRO AT TOSCANELLA. With regard to the date of this
church, the views of writers differ very widely. Thus, while Turriozzi l states that it is
possible that the existing structure may go back to the middle of the Vllth century,
Campanari, 2 on the other hand, thinks that it was erected in the IXth, and that
towards the close of the next century it was enlarged by two bays and embellished
with a facade. And while Promis 3 believes that it was not built before the Xlth
century, and Dehio 4 also considers that it belongs to that century, though the front
was perhaps not finished till the Xllth, Rohault de Fleury 5 asserts that it was in
existence by the IXth, Lenoir 6 thinks that it was built about that time, and Gaily 7
regards it as a work of about the middle of the Vllth century with the exception of
the front, which he would place in the first half of the Xlth. Amico Ricci, 8 again,
while not committing himself to any definite statement about the foundation of the
basilica, which may belong to the Xlth century, inclines to believe that it was finished
in the closing years of that century or, more probably, in the course of the next.
Lastly, Gentile 9 argues that it was built at the end of the Vllth century or the early
years of the VHIth, and was then enlarged and decorated with a front at different
dates between the end of the Xth century and the course of the Xllth.

This striking divergence of opinion is due to the fact that most of those who
have dealt with this instructive monument have based their opinion on the convenient
but fallacious evidence of the ritual of the Church. Campanari, for instance, decides
the date of the building by its orientation ; though in the IXth century the orienta-
tion of churches had become a matter of mere convenience. Or else they depend on
arbitrary statements, or on the mistaken belief that Lombardic architecture was
already fully developed in the period between the end of the Vllth century and the
early years of the VHIth; or on merely general or even erroneous historical con-
siderations, or on that enthusiasm which sometimes stands in the way of scrupulous
veracity ; or, lastly, on the opinions of others, without taking the trouble to verify
them.

So far as I can see, the history of this church if not the true, at least the
conscientious history has still to be written. Let us attempt it.

The exact dates of its foundation and of its later additions are not known.

1 Afemorie istoriche della cittti Tuscania eke ora volgarmente dicesi Toscanella.

3 Tuscania e i siioi monumenti.

3 Trattalo di archiiettura civile e mi/ilare di Francesco di Giorgio Martini.

* Dehio und von Bezold, Die kirchliche Baukunst ties Abcndlandes.

8 La A/esst. tudes archJologiques sur ses monuments. * Architecture monasli</ue.

7 The ecclesiastical architecture of Italy from the time of Constantine to the XV century.

* Op. cit. * San Pietro di Toscanella in Archivio storico tielf arte, Anno II.

122 LOMBARDIC ARCHITECTURE

Researches made by me on the spot were fruitless, for the local archives are silent
on this point ; and the documents, which might have thrown some light on the
question, belonging to the abbey of San Giusto, the imposing ruins of which are to
be seen near the town, have disappeared. Nor was I more fortunate in my perusal
of the documents of the abbey of Monte Amiata, 1 or of the Register of Farfa. 2

It is true that one of the two historians of Toscanella, Turriozzi, 3 states that,
about the middle of the Vllth century, the bishop's throne was moved from the
church of Santa Maria Maggiore to San Pietro. Further, he tells how " there could
be seen the leaden seal authenticating the translation of the relics of the martyr
saints, Secundianus, Verianus, and Marcellianus, in the year of our Lord 648 :
f Anno Domini CCCCCCXL VIII. Ind. VI. corpora sanctorum mar ty rum Secundiani,
Marcelliani, Viriani, & Deodati a domo sanctorum traslata sunt in civitatem
Tuscanam" These facts, if true, would be of first-class importance for us because,
although I am not one of those who think that a building must belong to a particular
date merely on account of certain events directly connected with it, they might
be brought into connection with the foundation of the oldest part of the church.

Unfortunately the date of the transfer of the bishop's chair is not certified by
any authentic document. The bare fact is only known by what can be gathered
from the well-known Bull of Leo IV (845-857), recorded by the two historians
referred to, which confirms to Virobono, bishop of Toscanella, jurisdiction over all
places subject to that diocese. That is to say, in 852, Santa Maria Maggiore, which
in the past had been the principal church of the see, had ceased to be the cathedral,
and was now a " pieve " or parish church (" . . . ecclesiam S. Dei genitricis semperque
Virginis Mariae, quae olim caput episcopii extitit, et nunc plebs facta est . . ."). The
document, too, relating to the translation from Cencelli to Toscanella of the bodies of
SS. Secundianus, Verianus, and Marcellianus, which misled Turriozzi, is considered to
be apocryphal. All that is known about the translation is that it took place in early
times, as we read in the " Acta Sanctorum" " Utut est, possessio Tuscaniensium,
quandocumque sit adita, certo antiqua reputari debet." *

There is, however, one last clue, and that is the presence at Toscanella in the time
of the famous Liutprand (712-744), of the Comacine master Rodpertus, which is
established by the well-known deed of sale (739) of property belonging to him in the
Vicus Dianus and other places within the territory of the city. 58 Of this fact we must
lay hold, remembering that the reign of Liutprand was long and prosperous, that it
marked the zenith of the Lombard dominion, and that it was the most productive of
buildings. Another reason is that San Pietro is in the same Pre-Lombardic style as
the church of Arliano (VHIth century); and lastly, various decorative details in the
basilica unquestionably point to the first half of the VHIth century.

In my belief the basilica of San Pietro as it stands to-day is the result of four
distinct periods. To the earliest, that is to say the time of Liutprand, belongs the
original church, comprising the east end and the three adjoining bays of the nave of
the present building, with a crypt or " confessio " beneath, which has been rebuilt and
enlarged at some later date. To the second period, i.e. the last years of the Xlth cen-
tury, are to be assigned the existing crypt below the presbytery, the raising of the
floor of the original chancel, and the erection of the present ciborium over the high

1 Archimo delta R. Societb Romano, di Storia patria, Vol. XVI. Calisse, Documenti del monastero di San
Salvatore sul Monte Amiata rigitardanti il territorio romano (Secoli VIII-XI).
" 11 Regesto di Farfa di Gregorio di Catino (Giorgi e Balzani).

3 Op. at. * De SS. Secundiano et Socc. MM. Die Nona August!.

5 Troya, op. cit. 6 Brunetti, Codice diplomatico toscano.

THE PRE-LOMBARD1C STYLE

123

altar. To the third, or middle of the Xllth century, are to be ascribed the extension
in length of the primitive church, and the construction of the oldest portion of the
facade. Lastly, to the fourth period, or end of the Xllth century, must be attributed
the central portion of the front, and also the
tessellated mosaic pavement in the central
part of the chancel and in the nave.

Let us now examine the original church,
which must be regarded as the best repre-
sentative of the ecclesiastical architecture of
Italy in the VHIth century, and one of the
most remarkable existing churches of the
three centuries preceding the epoch of the
year 1000 to be found not only in Italy but
also in the countries beyond the Alps. It
consists of a nave and aisles (Figs. 165, 167)
separated by four columns and two piers with
engaged columns surmounted, with one ex-
ception, by capitals which in their rudi-
mentary form show the Pre-Lombardic
cubical type. The columns themselves, one
of which has been renewed, have been taken
from older buildings like their capitals, which
carry heavy pulvins. The columns are con-
nected by a low wall with a continuous seat,
forming the division between the nave and
the aisles. The nave arches are of various

dimensions. They

are ornamented with

dentils in the form

of parallelepipeds : a

decorative motive

which, though rude, is

none the less original

and very effective. I

have not met with

it in any church

older than the period about the year 1000 that I have

seen.

The walls of the nave are finished off on the inside by a

range of blank arches, with shafts carrying small cubical

capitals. These arches recall the range of shafts resting on
Fig. 1 66. Rome. Baths of consoles and intended, besides providing supports for the

Titus (79-81). (From Pal- ,

ladio, "Letermedi Kama.") beams of the roof, to decorate the upper story of the walls
in the nave of the basilica at Kalb-Lauzeh (Vlth century).

These latter shafts, in their turn, recall the colonnettes supported on corbels which
were used to decorate some of the halls in the Baths of ancient Rome : for instance
the Baths of Titus (79-81) (Fig. 166).

This blank arcading ought to be noticed, for later, i.e. in the course of the
Xlth century, passages were made in it, and so it became the source of the internal

H VHIth Century.
mst Xllth Century.
mt Xllth Century.
Fig. 165. Toscanella. Plan of San Pietro.

I2 4

LOMBARDIC ARCHITECTURE

arcaded wall galleries which are a characteristic feature, and an original one, of the
Lombardo-Norman style. This is the earliest specimen that I can adduce of this
form of decoration. There is, indeed, the basilica of Eski-Djuma at Salonica
(Vth century), which orginally showed, high up, corresponding to the women's gallery,
arcades with recessed arches (Fig. 168) supported (i) by small piers with engaged
columns which carry low capitals elliptical in section, ornamented on the exterior with
plain, stiff, hollowed-out leaves, and on the face looking into the church with a cross ;
and (2) by a single massive pier built about the middle of the length of the wall in

Fig. 167. Toscanella. San Pietro (VHIth and Xllth Centuries).

order to strengthen it. These arcades, however, which are now built up and in places
destroyed altogether, were filled by transennae intended to transmit a modified
light to the gallery, and fixed against the smooth strips which separate the two
halves of the ellipse of the capitals. The basilica also of St. Demetrius (Vth
century) in the same place has the upper part of the nave embellished with an
arcade with engaged columns ; but these too, before they were walled up, were filled
by transennae intended to light the nave.

The upper end of the nave opens into a spacious presbytery bounded at the
further extremity by the apse (which is flanked by two niches taken out of the
thickness of the outer wall), and in front by the piers of the chancel arch which
support both the transverse and longitudinal arches of the presbytery, and also the
two nearest arches of the nave.

The presbytery of the original church was divided from the nave and aisles

THE PRE-LOMBARDIC STYLE

125

by a screen. Of this screen, and also of the altar and other ritual fittings, numerous

marble fragments survive, now built up in the roughly constructed dwarf walls,

furnished with seats on the inside, which were intended to form an outer partition for

the actual presbytery, and also to separate its middle portion from the sides. These

fragments, consisting

of entire plntei, parts

of plutei, cornices,

and uprights, at first

sight appear to be of

the same date, but

to a trained eye

reveal work of two

distinct periods.

To the earlier
belong, for example,
the altar frontal
(Fig. 169) and the
fragment of a pluteus
(Fig. 170) here illus-
trated. In the carv-
ing of this group we
find, in the first
place, the Pre-Lom-
bardic characteristics
of the first half of
the Vlllth century ;
e.g. the motive of
crosses framed in
pairs of arches which,
with the pillars from
which they spring,
are sometimes com-
posed of inter-
lacings ; a kind of
ornamental cresting
of cauliculi arranged
symmetrically ; and
thirdly, the design
of squares formed
by interlacing bands
containing flowers,
crosses, bunches of
grapes, leaves, con-
ventional plants,

birds, &c. But the art displayed is less advanced than that of two fragments
of plutei built into the wall of the portico of the SS. Apostoli at Rome, which we may
believe were set up by order of Pope Hadrian I after the fall of the Lombards (774-795),
as is shown by the less barbarous treatment of the leaves and birds pecking (the

Fig. 168. Salonica Eski-Djuma (Vth Century).

126

LOMBARDIC ARCHITECTURE

feathers being represented by irregular triangular notches made by the hammer)
which are carved on them, and also by the presence of conventional lilies, not single

^i rp>

tit

feifeS

Fig. 169. Toscanella. San Pietro. Altar frontal (739).

but combined in heads. So that I have no hesitation in assigning this group to the
first half of the VHIth century and to the handiwork of Master Rodpertus or one of
his " colleagues."

To a later period, on the other hand, belong the carvings illustrated by Fig. 171,

Fig. 170. Toscanella. San Pietro. Fragment of phileus (739).

and all the others of the same type and workmanship to be found in the screens of the
church. They exhibit, both in design and execution, a striking improvement on

THE PRE-LOMBARDIC STYLE

127

those of the earlier period. This
group shows more regular and
accurate design than the marble
fragments of the time of Hadrian I,
which came to light in the recent
restoration of Santa Maria in
Cosmedin at Rome (Fig I/2). 1
These fragments are to be regarded
as representative of the best work
in carving ordered by Hadrian I ;
that eminent restorer and rebuilder
having, as we know, decorated the
church in such a way that it should
deserve its title of " cosmedin."
They must also be considered to
be the result of Comacine chisels
on account of the designs they
show, which were at that time a
novelty at Rome, and in view of
the characteristic sharp edge of the
carving, and the typical crudity
and rudeness which mark their
productions.

At Rome, in the Vlth, Vllth,
and VHIth centuries, up to the
fall of the Lombards, so hated
and dreaded by the Popes, the

Fig. 171. Toscanella. San I'ietro. Plutetis (IXth Century).

local artists, for the decorative treatment of panels, kept to the motives of the old

Fit,-. 172. Rome.

riutius (774-795).

1 Giovenale, Z.a basilica di Santa Maria in CosmeJin.Anniiario aelf Assodaiione artislica fra i cultori tit
architettura in Roma, anno V.

128

LOMBARDIC ARCHITECTURE

Roman School of the classical period and of the decadence, supplemented by those
of the Schools of Constantinople and Ravenna. So that the panels which may

Fig. 173. Pxomc. Santa Sabina. Pluteus (795-816).

be regarded as their work merely show the usual characteristic framing, lozenges, eight-
pointed stars, flowers, rounds, and crosses ; and, later, small sunk panels, roses, whorls,

knots, small trees, and
geometrical interlacings.

The group in ques-
tion, moreover, shows an
art superior to that of the
carving seen in the remains
of an altar and chancel
screen (Fig. 173) belonging
to the time of Pope Leo III
(795-816), which a few
years ago were degraded
to serve as steps up to the
high altar in the basilica of
Santa Sabina at Rome
(Vth century), but are now
exhibited in the left aisle.
The execution of the
bunches of grapes and the
form of the leaves do not
suffer by comparison with
the carvings of the choir
enclosure (Fig. 174) of Pope
Eugenius II (824-827), now

Fig. 174. Rome. Santa Sabina. Pluteits (824-827).

shown with them.

THE PRE-LOMBARDIC STYLE

129

These carvings in Santa Sabina are to be regarded as the work of local chisels,
because we find in them, especially in the later ones, a method of cutting which is not
so uniformly triangular as that of contemporary Comacine work,
and also a certain grace and elegance a breath, as it were, of
classic art, which we should look for in vain among the produc-
tions of the Lombard gilds of the IXth century.

Hence we shall not go far wrong if we assign the second
group of carvings in San Pietro to the time of Virobono, bishop
of Tuscania, whom we have already come across, or rather of his
successor John, who filled the episcopal chair for many years
with so much honour that Pope John VIII (872-882) sent him
as Apostolic Legate to preside at a
Council held in 876 in France, where he
sat at the right hand of the Emperor
Charles the Bald.

With these fragments go the two
cubical capitals (Fig. 175) clumsily in-
serted below the impost line of the pres-
bytery arches ; another small cubical
capital which may be noticed at the
entrance to the presbytery from the right
aisle, where it supports the damaged
archivolt of a ciborium ; and, thirdly,
three small capitals closely allied to the
preceding, built into the vestibule of the
crypt and the adjacent passage. All
these capitals, which must have belonged
to the chancel screens and ciborium of
the original church, are cubes with the
lower angles bevelled off, and the faces
ornamented by cauliculi and rude leaves

packed into shells or else free. They are carved in low relief
without undercutting, and the design is as barbarous as the
execution is coarse. The one at the entrance to the presbytery
shows a family likeness to the small cubical capital -{Fig. 176)
made for the iconostasis of Pope Hadrian I in Santa Maria
in Cosmedin at Rome (774-795), recently discovered, with the
shaft and base belonging to it, in the campanile of that church
for which it had been used in the Xllth century. But it is not
of the same date as the latter, for the greater crudeness of both
design and execution, and the greater poverty of composition
which it shows, make it evident that it is the work of a period
earlier than Pope Hadrian, in other words that it belongs to
the first half of the VHIth century.

There is no evidence to show whether, originally, the
presbytery was only slightly elevated above the floor of the
nave, like that, for instance, of Santa Petronilla near Rome
(IVth century), which is raised above the choir merely the depth of the threshold
between them, while the choir is raised by only a single step above the floor of the
VOL. I K

Fig. '75' Toscanella.
San I'ietro. Capital
(739).

ftl

Fig. 176. Rome. Santa
Maria in Cosmedin.
Column in the Cam-
panile( VI 1th Century).

130

LOMBARDIC ARCHITECTURE

nave. On the other hand, it may have stood at the top of some kind of a flight of
steps, as in the church of San Protaso in the suburbs of Como (391-420), where the
floor of the presbytery was raised about li feet above the level of the nave, 1 or the
primitive basilica of San Valentino on the Via Flaminia near Rome (337-352), where
the bema, including the scliola cantorum or choir, stood, some what above the floor of
the church, and the apse was raised several steps higher than the berna? Another
instance was the church immediately preceding in date the basilica of Euphrasius at
Parenzo, and going back to the IVth century, 3 where the presbytery was raised
about 2 feet above the floor of the nave. The fact is that at Toscanella the bases
of the piers, columns, and half-columns in the presbytery, were left buried when the
floor was raised, as may still be seen.

Beneath the presbytery and apse a crypt or confessio was constructed, but above

Fig. 177. Toscanella. San Pietro (Vlllth and Xllth Centuries).

ground owing to the abrupt fall of the site. That a crypt was built at the same time
as the church is confirmed by the existence of three windows, one in the middle of
the apse and two in the aisles, intended to light it, for they are evidently contemporary
with the building of the church.

The outer walls of the church (Fig. 177), of rubble concrete faced with regular
courses of dressed tufa, are embellished with blank arcades and arched corbel courses
divided by small lesenas. The latter show the erroneousness of the view of
archaeologists 4 who would postpone the introduction of such miniature arcades to the
epoch of about the year 1000. In this connection I may remark that, if arched

1 Barelli, Chiesa di San Protaso net sobborghi di Como. Rivista archeologica delta provincia di Como,
fasc. 25.

^ Marucchi, // cimilero e la basilica di San Valentino,

8 N. Bullettino di archeologia Cristiatia, 1895. Marucchi, Le recenti scoperte del ditomo di Parenzo.

4 Rivista arch, della provincia di Como, fasc. 10. Barelli, Battistero di Lcnno.

THE PRE-LOMBARDIC STYLE

corbel courses broken by lesenas are a decorative feature invented at Ravenna, it was
the Lombard gilds which gave them refinement, used them more freely, combined
them in greater variety, and thereby imparted to wall surfaces an air of grace and
elegance which it would be vain to look for in buildings of the Romano- Ravennate
style.

The walls are finished by a cornice carried on consoles. The apse (Fig. 178)
is embellished by an arched corbel course divided by vertical rolls, by a band of tiles
arranged lozenge-wise, and by two courses of rectangular cavities. The latter,
three in each division, are a new idea which may have suggested the invention of
the arched niches grouped in threes by lesenas, the earliest example of which is

Fig. 178. Toscanella. San Pietro. Apse (739).

presented by the apse of Sant' Ambrogio at Milan (789-824). The church was
lighted (i) by narrow, and sometimes very narrow, round-headed windows splayed
both inside and out ; (2) by loops splayed inside ; (3) by luminous crosses ; and (4)
by rectangular openings recessed in steps. Of the latter, by the way, a number of
examples are to be found in ancient Roman tombs. This kind of opening was in
course of time widened, arched at the top, and moulded in the jambs and archivolts ;
the result being the characteristic Lombardic recessed and moulded window.

The narrow double-splayed windows, the lighting capacity of which steadily
diminishes as they get nearer the ground so that at last they become mere loops, those
in the presbytery being even narrower than those in the nave, might have been made
on purpose to create serious difficulties for archaeologists. It is not easy to decide
whether our Comacine master, Rodpertus, with the problem before him of providing

K 2

132

LOMBARDIC ARCHITECTURE

Fig. 179 Toscanella. San Pietro. Crypt (Xlth Century).

light for his church, adopted this arrangement in order to increase the solidity of the
building ; or to make it difficult for ill-intentioned persons to penetrate into the
sacred precincts unobserved ; or with the idea of creating, especially in the sanctuary,

Fig. 180. Toscanella. San Pietro. Capital
in the nave (Xllth Century).

Fig. 181. Toscanella. San Pietro. Capital in
the nave (Xrith Century).

THE PRE-LOMBARDIC STYLE

133

an atmosphere of reserve and mystery by leaving it rather dark ; or because windows
of this form seemed to him best suited for a building so exposed to the fury of the
winds, especially from the north ; or, in conclusion, for all these reasons taken
together.

The original doors had lunettes above them.

The date of the church is shown by the carving of the marble fragments and the
cubical Corinthianesque capitals already noticed. Additional evidence is provided
by the plain cubical
capitals surmount-
ing thehalf-columns
of the presbytery
and the first arches
of the nave, and,
again, by the plain
cubical capitals
used in the blank
arcading. The lat-
ter are merely a
rudimentary form
of the other worked
capitals. Besides,
that the building is
earlier than the
IXth century is
shown by the intro-
duction of the small
rectangular cavities
in the apse, the
forerunners of the
arched niches which
appear about the
dawn of that cen-
tury. And the apse
also shows that it is
later than the VI I th
century. Before
that time, the apses
of churches in Italy,
with the single ex-
ception of San
Giovanni Evange-
lista at Ravenna

(425), exhibited perfectly bare outer faces, or else were merely finished at the top with
the classical Roman motive of a range of consoles, or, under the decadence of art,
with a band of saw-tooth, the two motives being sometimes combined.

And that it was built by the agency of a Comacine gild, and not by workmen
either local or from the neighbouring Duchy of Rome, is not difficult to prove. First
and foremost, the constructive and artistic quality of the building is plain evidence of
the fact. And then we must remember that, down to the fall of the Lombard kingdom,

Fig. 182. Viterbo. San Sisto Vecchio (Xllth Century).

134

LOMBARDIC ARCHITECTURE

the Code of Rotharis and the " Memoratorio " of Liutprand continued in force, and
the Comacine masters still enjoyed the privileges granted to them by those monarchs.
Moreover, we may take note that the statutes of Tuscania, which are believed to
date from the first half of the XHIth century, 1 that is to say a period when the pros-
perous Ghibelline town had not yet been compelled by force of circumstances to
forgo the right of self-government, 2 do not say that among the twenty crafts into
which the city was divided, was included one of masons, a fact from which it may
reasonably be inferred that for important constructions the people of Tuscania were
obliged to rely at all times, and how much more in the early Middle Ages, on the

Fig. 183. Viterbo. San Giovanni in Zoccoli (1037).

building gilds of other districts. And, finally, we may reflect that the Roman master
masons and " marmorarii " never built in the Pre-Lombardic style, and that the exist-
ing specimens of their handiwork, from the Vllth to the Xlth century, are there to
prove the absolute impossibility that San Pietro at Toscanella could have been pro-
duced by them.

In the course of the Xlth century the interior received its present form. The
crypt was rebuilt (Fig. 179), the approach to it remodelled, the presbytery raised.
These were the changes, I believe, recorded in the inscription on the ciborium of the
high altar bearing the date 1093 I an d this is confirmed by, among other things, the
capitals of simple Roman Composite form, specially made for the church and all of
the same date, to be seen in the crypt, the ante-crypt, and the ciborium just mentioned :

1 Campanari, op. cit. - Bussi, Istoria della ci/tA di Viterbo.

THE PRE-LOMBARDIC STYLE

135

by the small Pre-Lombardic cubical capitals, brought from elsewhere, in the original
church ; by the character of the cross vaulting in the crypt ; and by the buried bases
of the pillars in the presbytery.

Towards the middle of the Xllth century the church was enlarged by adding two
bays to the nave and aisles, and various minor works were also carried out. The date
is made clear both by the capitals in the side doors of the front, and also by those in the
new part of the building (Figs. 180, 181), no longer simple Composite, but Lombardic
of definite Xllth century character, one of them exhibiting the Corinthianesque style
of those in the original crypt of San Pietro in Ciel d'Oro at Pavia, which was re-
consecrated in H32. 12
This capital at Toscan-
ella recalls the style of
various specimens in
San Sisto Vecchio at
Viterbo (Fig. 182),
which is not so early
as is generally thought.
The earliest of the
churches there is San
Giovanni in Zoccoli
(Fig. 183), finished in
1037, as is proved by a
document in its
archives, which shows
that a bell re-cast in
1697 bore that date, 3
and confirmed by its
capitals, some of which
are clearly contempo-
rary with those in the
gallery of San Flaviano
at Montefiascone. San
Sisto Vecchio must
have been built in the
first half of the Xllth
century, very probably in the pontificate of Eugenius III (1145-1150) who resided at
Viterbo for considerable periods.

Finally, at the close of the Xllth century, the front was remodelled. Its central
door is evidently the production of Roman " marmorarii " at the end of that century
or the beginning of the next. The capitals of the open loggia above it are clearly of
the same date as those in a similar position in the church of Santa Maria Maggiore,
also at Toscanella (1206), and those of the portico of San Lorenzo fuori !e mura at
Rome (1216-1227). Thirdly-, the rose window, which, with the pair of two-light
windows flanking it, recalls the front of the SS. Crocifisso at Lugnano in Teverina
(Fig. 184), belonging, as I believe, to the Xllth century, cannot be any earlier,
because rose windows did not make their appearance in church fronts before about

1 Dell' Acqua, Per la solenne riapertttra al culto della basilica di San Pietro in Ciel d'Oro.
Maiocchi e Casacca, Codex diplomat if us ord. S. Augustini Papiae.
3 Bussi, op. cit.

Fig. 184. Lugnano in Teverina. Church of the Crocifisso (Xllth Century)-

LOMBARDIC ARCHITECTURE

the middle of that century. A very early, certainly dated example is the one in
Santa Maria del Vescovado at Assisi, of the year 1163.

THE BASILICA OF SAN SALVATORE AT BRESCIA was, together with the
monastery, erected from the foundations by order of King Desiderius (756-774) and
his wife Ansa, in place of an earlier church dedicated to St. Michael the Archangel
and St. Peter the Apostle, as we learn from two documents of 759 and 760 : " monas-
terii cum ecclesiis, et reliquis edificiis a nobis ibidem constructum " " rnonasterio
. . . quod nos . . . fundavimus et ereximus et superna subveniente misericordia
hedificavimus." 1 The church was begun in 753, the date given by an ancient service

book belonging to the monastery.

As the present condition of the
building (Fig. 185 and Frontispiece)
prevents a complete examination of its
constructive and artistic features, we
will confine ourselves to the following
statements.

The original front was destroyed
when the upper church of Santa Giulia
was built.

The church has lost its apse,
nothing being left of it except the
foundation wall in the crypt below the
presbytery.

The Byzantine cubical funnel-
shaped capitals of the nave arcades,
covered with pierced and undercut
foliage, point to the Vlth century and
the work of Greek carvers. In all prob-
ability they came from the earlier church
of SS. Michael and Peter.

On the other hand, the Corinthian
capitals with Byzantine sharply in-
dented foliage, and others of the same
class but with stiff, plain leaves, which suggest, though in somewhat degraded form,
the Ravennate features of the capitals in the Santo Spirito at Ravenna, point to
the same date as that of the capitals in the original San Giovanni in Fonte (Fig. 186)
and Santa Maria Matricolare at Verona (about 750-760)^ and must be ascribed to
Ravennate chisels of the Vlllth century, and not to Comacine or Byzantine carvers.
For in the Vlllth century the Comacine gilds produced only Pre-Lombardic cubical
capitals, as is proved by one (Fig. 187) preserved in the Museo Civico at Brescia,
coming from the original crypt of San Filastrio in the ancient church of the Virgin
there a church the erection of which is ascribed to the second half of the Vlllth
century. Nor have I been able to discover in the East any trace of capitals similar
in design or execution to those of Corinthian type with plain, stiff leaves in San
Salvatore.

The upper part of the Vlllth century nave was altered at a later date, and
covered with the present barrel vault.

Fig. 185. Brescia. San Salvatore (753).

1 Hist, patriae man. Codex diplomaticits Langobardiae.

2 Biancolini, op. cit.

THE PRE-LOMBARDIC STYLE

'37

All that remains of the walls of the aisles, which may be regarded as belonging
to the original building, is either coated with plaster or concealed by structures
erected against it, so that it is impossible
to ascertain whether or not there is any
architectural decoration on the outer face.

The windows, now blocked up, in these
walls, suggest by their form a date near to
those in the parish church of Arliano (712-
744) and in San Pietro at Toscanella (739).

The cross vaulting in the aisles is later
than the original building, as is shown by
the way in which it is constructed.

The basilica of the time of Desiderius
had a crypt only beneath the apse, divided
into three aisles by four piers supporting
longitudinal arches on which the pavement
of the apse rested. Later, probably in the
Xllth century, the crypt was enlarged by
an extension beneath the presbytery.

Fig. 186. Verona. San Giovanni in Fonte.
Capital (about 750-760).

With so few buildings to judge by, it
becomes difficult to form a precise idea of
the architecture in vogue during the Lom-
bard period ; all the more as the surviving specimens are not only few and far
between, but are also more or less poor in character. They certainly did not
rank, not even San Pietro at Toscanella, the best of them, among the important
buildings of the period, for Paulus Diaconus has not mentioned any one of them.

At the same time, I believe that the Lombard architecture was not that uncouth,
debased, and barbarous product that it is generally held to have been ; and I think
that Muratori l was right when he wrote that some of the better buildings erected
under the Lombard Monarchy, if they had survived the assaults of time, would have
presented no uncomely spectacle, seeing that they excited the admiration of Paulus
Diaconus, who, we must not forget, had been able to contemplate the numerous

important structures of antiquity which in his day were
still in existence at Rome.

In fact, so far as we can judge from San Pietro at
Toscanella, the buildings of this period exhibit merits
of construction of no ordinary character, while the
architectural decoration of their exteriors is superior, in
variety of motives and their intelligent distribution, to
that of any Christian monument erected in Italy before
the IXth century. Not to mention that some of the
churches must have been sumptuously decorated with
mural paintings, among which may be classed the much
extolled embellishments which Queen Rodelinda, wife of
Pertarit(f 686), caused to be executed in the basilica of the Mother of God erected by her
outside the walls of Pavia : " opere mirabili condidit, ornamentisque mirificis decoravit." 2

1 Antiquitates i'.alicae medii aevi. Dissertatio XXIV.
1 Alon. Germ. Hist. Pauli historia Langobardorum .

Fig. 187 Hrescia. Museo Civico.
Capital from the crypt of San
Filastrio (VHIth Century).

138

LOMBARDIC ARCHITECTURE

Having said so much by way of preface, let us turn to the characteristics of an
original nature which differentiate the Pre-Lombardic style in this period, or rather
in its last hundred years, being obliged as we are to derive them from monuments
which, in almost every case, belong to the VHIth century.

I. The use of half-columns in piers or walls, with capitals formed of a cube of
stone simply chamfered off at the angles.

II. The employment for the small columns and piers of the ritual fittings, such
as chancel screens, choir enclosures, and altar ciboriums, of capitals generally
shaped out of a prism of marble or some kind of stone, forming one piece with
the shaft and base. I call them "Pre-Lombardic cubical" (Figs. 188, 189) to
distinguish them from the Byzantine funnel-shaped cubical capitals. Rather free
and debased Corinthian in type, they form cubes chamfered in the lower part of the
angles which are filled with leaves, sometimes of the stiff, plain type, but more often
of the palm. Each of the faces is ornamented with coarse cauliculi, either singly or

in pairs ; rude and stiff leaves, some-
times plain, at others carved, or, again,
with the tips turning over in clusters,
in some cases free, in others packed
into shells ; roses, single and double;
whorls ; volutes, crosses, conventional
trees decorated with flutings ; chan-
nelling, sometimes arranged in groups ;
flutings horizontal, vertical, zigzag,
and radiating, &c. No attempt is
made to reproduce animal figures, and
hardly any to represent human beings.
With the exception of a well-known
cubical capital in the Museo Archeo-
logico at Verona, which shows a
human head framed in a medallion,

and a similar capital of the Vlllth century just discovered in the crypt of San
Giovanni at Asti, with on its face a nimbed head enclosed in a medallion, I do
not know what other example to refer to. The decorative details, too, are clumsy
and inaccurate, and sometimes merely engraved, though as a rule they are carved
in low relief without undercutting, made by sharp, rough indentations, and the
lights and shadows produced simply by furrows and notches.

Notable specimens of the capitals in question are those of the well-known
ciborium in the church at San Giorgio in the Valpolicella, erected under Liutprand
(712-744), and Dominicus, bishop of Verona (712-740). They are carved in tufa, an
easy material to work, and executed entirely with the triangular cutting made by
the chisel. The inscriptions on two of the columns make them a reliable document
of the age of Liutprand, so that they can be used as fixed points for comparison
(Fig. 190).

This ciborium recalls a canopy formerly in the little church of San Prospero
outside the Porta Eburnea at Perugia, now preserved in the University Church. It
appears to me to be a work of the latter part of the Vlllth century, on account of
the capitals which recall the Ravennate manner of two of the same style,
belonging to the time of King Desiderius (756-774), in San Vincenzo in Prato at
Milan.

Fig. 1 88. Cividale.
Capital in the
Museum (Vlllth
Century).

Fig. 189. Cividale.
Capital in the
Museum (Vlllth
Century).

THE PRE-LOMBARDIC STYLE

139

In the Vllth century, in the districts under Lombard rule, taste in architectural
forms remained at a higher level than skill in design which, at least for carving, had
sunk to the lowest depths. The
technique, too, of the art of carving
was in no better plight, being re-
duced to a positively elementary
state, especially in the case of
capitals, as is shown by those in
the crypt of Sant' Eusebio at
Pavia. The Comacine masters
had ceased to produce capitals at
all, as it had become more con-
venient for them, and less costly
for pious founders, to use up old
ones collected from various sources.
But when they found that, in ad-
dition to the decay of their own Fig ' I9a ~ SanGior8 c ib i o r ium P (7 1 , i 2-74o) ChUrcK Cap '' alS f
art and their want of skill, they

were faced by the fact that such capitals became more difficult to procure as time
went on, especially those of small dimensions suited for ritual furniture, they were
obliged to supplement them with the productions of their own chisels. Accordingly
they confined themselves to simple cubes chamfered at the corners in various rude
ways, the only attempt to relieve their coarseness being some barbarous ornament
engraved or carved upon them.

The introduction of capitals of this sort must be referred to the second half of
the Vllth century. In the first half they were not yet in use, as those in the crypt of
Sant' Eusebio at Pavia prove; whereas in the early years of the Vllhh century they
are found, in a not very rudimentary form, in the ciborium of the church at San
Giorgio in the Valpolicella. At first they bear all the marks of that crudity which

characterises an art
which has reached
its lowest level.
Later, however,
during the long and
prosperous reign of
Liutprand, they
steadily improved
both in design and
execution. The
Comacine masters
carried them all over
Italy, whence they
spread along the
eastern coast of the
Adriatic and be-
yond the Alps.

This new type

of capital was the invention of the Comacine masters. In Italy the craftsmen of
Ravenna did not employ them, and the very few specimens to be found in Ravenna

Fig. 191. Classis. Sanl" Apollinare. Ciborium of St. Eleucaclius (IXth Century).

140

LOMBARDIC ARCHITECTURE

and it sterritory are obviously the work of Lombard chisels in the Vlllth or IXth
century. The capitals of the ciborium of Saint Eleucadius (IXth century) in Sant'
Apollinare in Classe (Fig. 191), with their Pre-Lombardic ornamentation, though
of Byzantine cubical form, show us what the cubical type preferred by the artists of
Ravenna in the IXth century was really like.

The Byzantines did not adopt this type of capital till late Among them the
oldest certainly dated examples (or presumably such), with the lower part of the
corners slightly chamfered off, and mostly ornamented with crosses, flowers, leaves,
&c., carved in low relief on their faces, are to be found in the churches of St.
Nicodemus (about 1044), St. Theodore (1049), and Kapnikaraea (Xlth century) at
Athens ; also on the Acropolis there (Figs. 192, 193), in a heap of fragments near the
Parthenon, mixed up with sculpture of every kind and date ; and, again, in the church
of the monastery of Daphni (Xlth century) near Eleusis (Fig. 194). We also meet with

Fig. 193. Athens.
Acropolis. Capital
(Xlth Century).

Fig. 192. Athens.
Acropolis. Capital
(Xlth Century).

Fig. 194. Daphni.
Church. Capital
(Xlth Century).

them, in the form of cubes chamfered the whole way up the angles, in St. Saviour Pante-
poptes at Constantinople (1081-1118). Lastly, we find them in St. Panteleemon at
Salonica (Xlth or Xllth century), which, in spite of its affinity with the Holy
Apostles, St. Elias, and the church of the Virgin (though it is more advanced than
this), is differentiated from them by its decorative niches on the exterior, and must
be placed at a later date, i.e. at the end of the Xlth or beginning of the Xllth
century.

This type of capital was the only one in use during the Vlllth century in the
Kingdom of Lombardy, and, after its fall, in its former territory and in the Duchy
of Rome. The only exceptions are a few productions of Ravennate carvers, such
as those of the Vlllth century in the baptistery of Callistus and Santa Maria in Valle
at Cividale ; in the old church of Santa Maria Matricolare (about 750-760), and
the original San Giovanni in Fonte (about 750-760) at Verona ; and in San Salvatore
at Brescia (753). Cattaneo, 1 indeed, included among those of the Vlllth and early
IXth centuries in use in the aforesaid territories, certain Corinthian and Composite
capitals, in some cases accurate in execution, and of varied form. For instance,
he reckoned among them the Roman Composite capitals with plain stiff leaves,
and the volutes and ovolo left uncarved, in the crypt of Santa Maria in Cosmedin

1 Op. fit.

THE PRE-LOMBARDIC STYLE

141

at Rome (Fig. 195) ; and also those of Corinthian type in the church of Santa Maria
in Domnica, also at Rome, built by Pope Paschal I (817-824). He adduced
them as the earliest efforts of the new birth of
Italian art, but in this view I believe he was mis-
taken. So far as capitals are concerned, the
classical revival of the Xlth century was almost
contemporary with the appearance of the Lom-
bardic style ; and it started in Lombardy, in the
most vital centre of the old Lombard realm, where
the embers of art glowed brightest at that time in
Italy, and thence spread over the whole peninsula.
These capitals, then, wrongly ascribed by Cattaneo
to the VHIth or IXth centuries, must be assigned
to the Xlth, or even the Xllth and XHIth. The
imitation of classical Roman work was a task be-
yond the powers of the Lombard carvers with their
very scanty skill, and still more those of Rome
in the VHIth century, and they certainly had not
the enterprise to attempt it ; while the carvers of

Ravenna never produced in that period capitals of the kind found in the two
churches which we have referred to.

For these reasons I regard as work of the end of the Xlth century or the
beginning of the Xllth the capitals in the basilica of Santa Maria in Domnica,
referred by Cattaneo to the pontificate of Paschal I, who rebuilt the church from its
foundations. 1 I believe that they are a result of the restoration which the building
must have needed after the damage surely suffered by it, like the others on the Celian,
at the hands of Guiscard when he entered Rome in io84. 2 I should also assign

Fig. 195. Rome. Santa Maria in Cos-
medin. Capital in the crypt (Xlth
Century).

Fig. 196. Rome. Coliseum. Capital
(69-80).

Fig. 197. Rome. Sant" Agnese outside
the walls. Capital (1st Century).

to the early Xlth century and to Italian hands (for no such architectural forms
are to be found in the East) the Composite capitals in the crypt of Santa Maria
in Cosmedin. These latter are more or less simple modifications of the initial form
of the Composite which, as I have pointed out, 3 appeared in Rome under the first two
Flavian Emperors (69-81). The archetypes are to be seen in the Flavian

1 Duchesne, Le liber pontificalis. 2 Man. Germ. Hist. Annales Ceccancnses.

3 Rivoira, Delia scoltura ornamcntalc dai tempi di Roma imperial! al Mille. Nwn>a Antologia, anno 1904,
fasc. 790.

I 4 2

LOMBARDIC ARCHITECTURE

Fig. 198. Rome. Lateran Museum. Capital
(795-816).

Amphitheatre at Rome, opened in the year 80 (Fig. 196) ; and some very fine ones in
the nave of Sant' Agnese outside the walls, probably belong to the 1st century

(Fig. 197).

A similar form of the classical Com-
posite was for a considerable period in
favour at Rome. Thus, in the days of Pope
Symmachus (498-514), we find a carver of
that school converting a block of marble
with the epitaph of Celer, the architect of
Nero, into the well-known and damaged
capital set near the entrance to the stairs
leading down to the basilica of Sant' Agnese
outside the walls. Two of the same pattern,
and well preserved, have been used in the
narthex of the basilica. All three have a
special stamp distinguishing them from those
of the same kind belonging to an earlier
period. They have no longer the old graceful form consisting of a double circle
of leaves, marked by a rib down the middle and more or less undercut, with shoots
springing up between them, and plain discs taking the place of volutes, also some-
times undercut, while the flower on the abacus is represented by a mere boss. On
the contrary, these are squat, and formed merely by two rows of leaves stuck on
to the body of the capital without any indication of a central rib, while the volutes
are irregular discs squeezed up and forming one mass with the capital ; and the
abacus is orna-
mented with a
rude ovolo.

The decad-
ence in composi-
tion, design, and
execution, which
we find in the
capitals of the
narthex at Sant'
Agnese, became
so marked in
course of time
that the Roman
carvers of the end
of the VII Ith cen-
tury and the be-
ginning of the
IXth could pro-
duce no better
specimens than
the two very bar-
barous ones

placed against the north wall of the old cloister of San Cosimato at Rome, two
more of rather better style, belonging to the time of Pope Leo III (795-816),

Fig. 199. Yillanova. Pliiteus in San Pietro (Ylllth Century).

THE PRE-LOMBARDIC STYLE

preserved in the Lateran Museum (Fig. 198), and some of the same date as the
last in the Forum. It was only in the first half of the Xlth century that the debased
type of the IXth century Composite capitals appeared at Rome in an improved
form. Interesting examples of this new type are provided by the crypt of Santa
Maria in Cosmedin.

Even though we take no account of the fact that it is not proved that this
church, as built by Hadrian I, was provided with a crypt or confessio, and paying
no attention to the common belief that the present crypt was constructed at the time
of the restorations and embellishments ascribed to the Xlth century, 1 we know what
sort of capitals were specially made for Hadrian's church. One of them has survived,
a solitary specimen which we have to thank the Roman " marmorarii " of the Xllth
century for using in the
campanile, instead of
destroying it, as they
seem to have done with
all the rest that belonged
to the old chancel screen.
And as it is impossible

to believe that such an

I -11 j u- i ^"

unskilled chisel was em-
ployed to carve the capi-
tals of an important
piece of ritual furniture,
while a far superior one
was reserved to produce
those in the crypt, there-
fore it is not unreason-
able to assign the latter
to a period later than
the end of the VHIth
century, or, to be more
precise, to the first half
of the Xlth.

III. The use in the
interior of buildings of
blank arcading.

IV. Lastly, in the
carved panels (Figs. 199,
200), uprights, and archi-
traves of chancel and choir screens, altar fronts, ambons, and archivolts of ciboriums,
the introduction of interlacing bands, generally with a double groove ; palm and vine
leaves ; a very free use of lilies, roses, bunches of grapes, Latin or Greek crosses with
the outer angles often ending in curls ; pairs of SS facing one another ; decorative
arcadings ; semicircular arches interlacing so as to produce pointed arches ; bosses,
bead and reel ornament, wheels, and stars ; birds pecking at fruit, leaves, or flowers,
or drinking from a vase ; fishes, cocks, serpents, lions, stags, bulls, griffons and
chimaeras ; and, lastly, though rarely, human figures and the symbols of the
Evangelists.

: Giovenale, Anmiario &c.

Fig. 200. Sirmione (Lake of Garda). Plutens built into the Casa Comunale
(VHIth Century).

144

LOMBARDIC ARCHITECTURE

These carvings are in low relief, all on one plane but of irregular depth, entirely
worked without undercutting, and the fillets, bands, crosses, leaves, flowers, arcades,

Fig. 201. Civita Castellana. Portico of the Duomo. Front of sarcophagus (VHIth Century).

animal forms, &c., are produced by sharp, rough indentations. In these ornaments,
in spite of the incorrectness of the drawing and the rudeness of the execution, a
certain charm is sometimes to be found. This incorrectness is far more marked in
the animal figures, where the form is often so little realised that it is impossible to
say to what species they belong. This is even more the case in human figures, which
are characterised by an absolute lack of expression in the faces, by an infantile
monotony in the composition, by the absence of any rules of proportion, by excessive
stiffness in attitudes and drapery, and lastly, by an elementary treatment of the

extremities which indicates
the lowest stage of art. This
can be verified by any one
who gives even a passing
glance at Fig. 201, which re-
produces part of a sarcopha-
gus, now built into the left
wall of the portico of the
cathedral at Civita Castellana
(XI I Ith century). Here the
trunks and branches of the
trees, two of which are placed
so as to form a kind of
Fig. 202. San Giorgio in Valpolicelia. arcade, are represented by

Church. Archivolt of cibonum

(712-740). grooved bands recalling,

both in execution and de-
sign, the interlacings on slabs
of the time of Liutprand
which we noticed in San Pietro at Toscanella (739). This similarity in style suggests
that we may ascribe the carving on this sarcophagus front to the same period, and to

THE PRE-LOMBARDIC STYLE

Some writers regard it as the result of a

Fig. 203. San Giorgio in
Valpolicella. Church. Archi-
volt of ciborium (712-740).

a Comacine chisel working to the order of the Lombard duke named in the inscription
cut along the top of the slab.

I am well aware that many people will find it hard to believe that the last
Pre-Lombardic characteristic is original,
mere imitation of Byzantine work, or
else an indirect copy of such work in-
fluenced, however, by Northern ele-
ments ; while others look on it as a
result of the grafting of Northern or
barbarian and Byzantine elements on
the stock of Roman art in the period
of its decline. Others, again, consider
it to be the work of Eastern artists,
or of Italians trained in the Eastern
School ; while, lastly, another class be-
lieves that it was derived, by means of
a transformation, from Roman and Early
Christian art.

We may allow that some of the
elements just mentioned contributed to
the formation of Pre-Lombardic decora-
tive carving, and I mean thereby Roman art both Pagan and Christian, and also
Byzantine, the Northern or barbarian elements being a myth, 1 while the Byzantine
sculptors of the Vlllth century, whether they were refugees from the iconoclastic
persecutions or not, were, if I am right, merely Italian artists. But it is none the less
true that this carving does constitute a new style, the invention of which must be
credited to the Comacine gilds, drawing their inspiration from Etruscan, Roman, and
Raven nate art.

These gilds diffused it throughout Italy and along the eastern coast of the
Adriatic, and thence it spread beyond the Alps, acquiring, as time went on, in the
different countries, a special character derived from the traditions of the School,
local influence, &c. Concrete instances of this can be pointed out in many localities,
especially at Rome, where the Pre-Lombardic carved panels, which had only just
made their appearance in the days of Hadrian I and after the Lombard Kingdom
had come to an end (774-795), had already acquired a character of their own, both
in composition and execution, by the time of Eugenius II (824-827).

In fact, though we may find in previous works the prototypes of the decorative
elements which appear in carvings of this kind the grooved bands, interlaced,
knotted, and twisted in various ways ; the compartments of different shapes
enclosing fanciful objects of all sorts : stars, crosses, lilies, bunches of grapes, leaves,
sun-flowers, daisies, roses, whorls, bosses, birds, &c. ; arcades ; intersecting arches ;
cauliculi and the bead and reel ; doves generally pecking at something ; peacocks
drinking at a vase or fountain, with sometimes a serpent biting the crest on their
heads ; fishes, animals, birds, griffons, &c., following, facing, attacking, and biting one
another these elements are very often combined in such a way as to form
absolutely new motives and compositions which, though not very refined, are still
original, rich, varied (sometimes even restless to the eye), and actually pleasing.

That the Comacine masters were the authors of this carving is proved by the

1 Archivio storico lombardo, 1896. Fontana, Suit origine delt arte longobarda.
VOL. I L

146

LOMBARDIC ARCHITECTURE

fact that the earliest specimens made their appearance at the beginning of the VHIth
century, in the districts subject to the Lombards, in company with the characteristic
Pre-Lombardic cubical capitals.

That it was not a creation of the East, imported into Italy by the Greeks, can
easily be proved. The

Fig. 204. Constantinople. Imperial Museum.
Archivolt of ciborium (Vlth Century).

of the ciborium in the

church at San Giorgio

in the Valpolicella

(about 712-740) pre-
sent the oldest cer-
tainly dated example of this kind of

ritual furniture ornamented with

interlacing basket work and animal

forms, and edged, though only

partially, with a cornice of cauliculi ;

features which form three of the chief

elements in Pre-Lombardic carving

of the VHIth century. The whole is

treated in low relief, the cutting

being of very acute triangular section,

left as the chisel produced it on all

the parts which stand out from the

background.

Before the appearance of this

ciborium (a fixed point of reference

for the character of Pre-Lombardic decorative carving in the early part of the Vlllth

century) there was no specimen of this kind of ritual furniture in the East presenting

such a wealth of interlacing as the one at San Giorgio, or decorated with figures, even

symbolical ones, of animals, or finished with a cornice of cauliculi. The Comacine

masters applied the
last motive to
gabled archivolts,
and finally sub-
stituted for it the
design of leaves fol-
lowing the slope of
the gable, a notable
instance of which is
the canopy of the
ciborium in Sant'
Ambrogio at Milan,
belonging to the
XHIth century. 1
Before Leo III the Isaurian (717-740) issued the first decree (726) against the

worship and production of images, and before the substitution of figures of animals

for those of saints during the reign of Constantine V Copronymus (740-775)1 the

1 Biscaro, Archivio storico lombardo, 1904, 1905. Note e aocutnenti Santanibrosiani.

Fig. 205. Rome. Santa Maria Antica.
Archivolt of ciborium (705-707).

THE PRE-LOMBARDIC STYLE

147

Byzantines were fond of decorating the archivolts of ciboriums with figures of Christ,
the Apostles, and angels, framed by foliage, interlacing, bands of indentations, bead
and reel ornament, &c. (Fig. 204) ; or else with foliage alone ; or, thirdly, with

Fig. 206. Athens. Church of the Virgin Gorgoepekoos. Slab in the fa9ade.

leaves, scroll work, flowers, and whorls, as may be seen in the archivolt found not
long ago in Santa Maria Antica in the Forum at Rome (Fig. 205), a church which
has been well described by Mr. Rushforth. 1 This archivolt must be assigned to the

1 Papers of the British School at Kome, Vol. \.-The Church of S. Maria Antiqua.

L 2

148

LOMBARDIC ARCHITECTURE

time of Pope John VII (705-707) and to a Greek chisel, for we know that the Pope,
who was a Greek by origin, embellished the church with mural paintings, and erected
an ambon. Hence it is more than probable that he also provided it with a ciborium,
and entrusted its execution to some carver of the Greek School, an origin which is
indicated by both the composition and the technique of the archivolt. On the

Fig. 207. Athens. Church of the Virgin Gorgoepekoos. Slab in the facade.

one hand, it does not show the facility and variety of grouping which characterize
the productions of the Pre-Lombardic School in the Vlllth century ; and, on the
other, it presents a design which is less faulty, and cutting not so irregular, though
at the same time not so deep, and therefore poorer in light and shade, than that
to be found in the works of the contemporary Italian School.

Then, if we pass from the archivolts of ciboriums to carved slabs or panels, we
find that, if we do not notice in the Eastern examples so marked a want of balance

THE PRE-LOMBARDIC STYLE

149

between the hand with its lack of skill and the imagination with its wealth of fancy,
as strikes us in those of the Pre-Lombardic School ; neither do we come across that rich
variety of motives, and those interlacings, with their bands crossing and recrossing in
circles and knots of ingenious and wonderful complexity, two features which are
characteristic of that School and exhibit a power of fancy which we should look for
in vain in the works of the Byzantine School. And we also find that in the East, I
mean of course in religious sculpture, animal representations are preponderantly
derived from Early Christian symbolism, or else reproduce motives from Graeco-
Roman art, and are hardly ever original. When they are, the animals, whether real or
imaginary, are not grouped in such a fanciful way, or engaged in such strange
conflicts, as occur in the animal subjects of Pre-Lombardic carving which later
became the most striking feature of decorative sculpture in the Lombardic style.

We have to reckon, however, with the five slabs at Athens, covered in low
relief with figures of griffons pecking at pine-cones, birds fighting with dog-headed
snakes, lions biting themselves in the back, winged sphinxes flanking a conventional
tree, with wingless sphinxes in one case appearing above them, and, lastly, a lion in
the act of tearing in pieces a lamb, used as building material in the front (Figs. 206,
207) and back of
the church of the
Virgin Gorgoe-
pekoos (Xlth or
XI Ith century),
wrongly called the
old cathedral. It
is on the strength
of these that Cat-
taneo 1 and others
with him assert
that in the VI I Ith
century figures of
this kind were
common in the

churches of the East and absolute novelties in the West. But this statement is very
far from the truth, for it seems that the five slabs in question came, like another
preserved in the National Museum with two lions flanking a conventional tree, from
a Graeco- Egyptian temple close by, possibly dedicated to Serapis. Very different
were the sculptured panels in Greek churches, not only in the VI I Ith, but in the
centuries before and after, right down to the XI Ith. And without hunting over the
whole of Greece for specimens and there are plenty of them : I confine myself
to reproducing here the fragment of a plutens (Fig. 208) lying near the threshold of
the door leading to the graveyard of the round church of St. George at Salonica
(Vth century) the same church of the Virgin Gorgoepekoos tells us what they
were like.

The fact is that the sculptures collected from various sources, which form so large
a part of the outer facing of that very interesting church (Fig. 209), include, together
with specimens of Pagan times, a large number which belong to the Christian ages,
apparently from the I Vth to the XI Ith century. Now, among the latter we observe
lozenges, either alone or connected by knots, lilies, roses, palmetto leaves, crosses of

1 op. dt.

Fig. 208. Salonica. Church of St. George. Fragment of plitteus (IXth
or Xlh Century).

5 o

LOMBARDIC ARCHITECTURE

various forms with the foot often ending in foliage, interlacings, bead and reel orna-
ment, intersecting arcades, a small arcade with pairs of columns tied together with a
cable, and crosses flanked by lions or griffons. All are carved in low relief of
varying depth, which sometimes is rounded off, while in other cases it is treated in
cuts of triangular section left as the chisel produced them, not, however, so sharply
defined as in works of the Pre-Lombardic style. But we never find the strange and
fantastic animal representations which characterize the productions of that School.

Fig. 209. Athens. Church of the Virgin Gorgoepekoos (Xlth or Xllth .
Century).

CHAPTER IV

ARCHITECTURE IN ITALY AND DALMATIA IN THE TIME OF

CHARLES THE GREAT

FROM Ciampin^to d'Agincourt, 2 from d'Agincourt to Cordero, 3 and right up
to our own time, writers have always been lavish in their praise of Charles the
Great (768-814) for having, as if by a touch of the enchanter's wand, raised
up art which had utterly fallen from its high station. And this is supposed
to be specially true of architecture, for the improvement of which he took peculiar
care, guided therein by his personal taste. And as if this were not enough, some of
them have further suggested that the buildings erected by the Emperor's orders were
designed by that master mind to serve as a universal standard in architecture ; and
that it was only through the incapacity of hi? degenerate successors that this vast
conception was never carried out. In any case, the buildings in question had a
considerable influence on the architecture which preceded the Pointed style.

I am, however, afraid that, hitherto in the history of art there has been some
exaggeration not only of the artistic capacities of the Emperor himself, but also of
the influence exercised, especially in Italy, by the buildings erected in his time.
The truth is that Charles had far more at heart the diffusion of civilisation in the
vast Empire which he founded than the promotion of the fine arts. At the right
time we shall discuss the buildings actually erected in his dominions to the north of
the Alps by his order and with the aid of Imperial funds, or shortly after his death
and under his influence ; and it will be seen that they are not of such a character as
to make my suspicions groundless.

In the territories taken from the Lombards he confined himself to the restoration
of an ecclesiastical or civil building here or there. The structures which De Dartein
and other writers believe to have been erected in the Emperor's lifetime under the
immediate influence of the Palatine Chapel at Aachen (796-804), viz. the
" Rotonda" at Brescia and Santa Sofia at Padua, were really buildings erected after
the Xth century.

The old, so-called " winter " cathedral dedicated to the Mother of God, at Brescia,
which must have been built in the second half of the VHIth century, a date confirmed
by the capital brought from its original crypt (San Filastrio), and now in the Museo
Civico of the town, was an edifice of basilican form, the plan of which was brought to
light in the course of the recent restoration, and has been traced out on the floor of
the present Rotonda. It consisted of a nave without aisles, the outer walls being

1 Op. dt.

2 Storia delt arte dimostrata cot monument i delta sua decadenza tul IV secolo fino al suo rinnovamenta nel
XVI. ' Op. cit.

IS'

152

LOMBARDIC ARCHITECTURE

only about 2 ft. 4 in. thick, and consequently strengthened on the outside by
buttresses corresponding to the tie beams of the roof. However, it was evidently
rebuilt; for during the restoration of 1883 there was discovered a marble slab with
the date 897, which had been used in the construction of one of the vaulting piers. It
has been suggested 1 that this happened about the year 1000. But the freedom of the
groined vaulting, and the capitals wrought expressly to carry it, indicate a date not
earlier than the Xlth century. It is my belief that the Xlth century or the first half
of the Xllth is the period which best suits the Rotonda of Brescia.

Santa Sofia at Padua, again, was never a round building, and it is certainly not
so old as the Carolingian period. The well-known Charter (1123) of Bishop
Sinibaldus (1106-1124) does not refer to a complete rebuilding of the church, but
rather, as Cattaneo 2 suspected, to the completion of the grand external apse forming
an ambulatory to the internal apse. The Lombardic figures of animals on the lower
story of the exterior of the apse point to the Xlth century.

When the Lombard king, Aistulf, had, by the capture of Ravenna, put an end to
Greek rule in the Exarchate (752), the Imperial prefects of the Adriatic transferred
themselves and their fleet to Zara.

The very disturbed period through which, not only Zara itself, but the whole of
Dalmatia passed after that event, was certainly not of a nature to attract the
craftsmen of Ravenna to the new seat of the Byzantine governors. They found much
more profitable ways of employing head and hand, first in the Lombard Kingdom,
and later in the new Prankish Empire. Then came the conclusion of peace between
Charles the Great and the Emperor Nicephorus I. Zara became the capital of
Byzantine Dalmatia, and the regular seat of the Proconsul or Strategos of the whole
region ; and it was then that plans were formed to embellish it with buildings
corresponding to its new dignity. What better opportunity could present itself for
the craftsmen of Ravenna to return in numbers to their own country, and, still
associated with the Comacine or Lombard masters, betake themselves to work in
Dalmatia, close at hand as it was and well known to them, where, among other
buildings, one was to be erected, the most splendid
of all, suggested by the famous chapel at Aachen ?
And I believe that this is exactly what happened.
And so we are able to add some other names to the
list, now almost complete, of buildings produced by the
School of Ravenna.

THE CHURCH OF SAN DONATO AT ZARA, already
existing in the reign of Constantine VII Porphyrogenitus
(912-958), originally bore the name of the Trinity.
Later, according to tradition, it was called after the
bishop who built it. Some 34 hold that it was founded
by Donatus, bishop of Zara, who, according to Gams, 5
occupied the see between 801 and 806. And with this Fig. 210. Zara. Ground plan of

T San Donate (about 801-806).

view I am in agreement.

1 Emporium, 1898. Arcioni, La Rotonda di Brescia. 2 Op. cit.

3 Hauser e Bulic, // tempio di San Donate in Zara.

4 Emporium, 1901. Smirid, 11 tempio di San Donato in Zara. 5 Op. cit.

ARCHITECTURE IN ITALY AND DALMATIA

The church of San Donate is of concentric plan (Fig. 210). The interior is
encircled by an aisle with a gallery above it, having piers and columns brought from
older buildings, and carrying
capitals of the Roman period.
One of these has been altered,
and on it are carved rude palm
leaves, lilies, shells, and ovolos.
The annular aisle on the ground
floor is covered by a barrel vault
with visible transverse arches.
The upper floor had also origin-
ally a barrel vault, but is now
covered by a wooden roof. Ac-
cess to it was given originally
by a small door on the north,
the archivolt of which is deco-
rated withcauliculi and the bead
and reel ornament.

Above this floor rose the

^ .

<J> ->

r^ "*l

V , '

f .1

Fig. 211. Zara. San Donato (about 801-806).

rose

drum of the cupola which fell at

some date which we cannot fix.

It has been replaced by an open timber roof. The traces of the dome suggest that

it was conical, i.e. of the same form as that of San Vitale.

From the outer wall, which on the south side is strengthened by buttresses, three

curvilinear apses project towards the east. They are continued in the upper story,

and on the outside are decorated with very
lofty blank arcading (Fig. 211).

The form of San Donato, though of the
circular type, is still, like the palace chapel
at Aachen, inspired directly by San Vitale
at Ravenna. Probably this was due to that
law of imitation which at times has such in-
fluence in the history of art. Or else Bishop
Donatus wanted to bequeath to the people
of Zara a striking memorial of the peace, to
the restoration of which he had so largely
contributed, and in consequence of which
their city had attained to the dignity of the
Capital of Byzantine Dalmatia, in the form
of a monument whose plan was derived from
the same source as that of the most famous
building of that age in Western Europe.
Another reason might have been that the
bishop had the idea of adopting for his
church the model chosen by Charles the
Great for the chapel of his own palace, in

order to win the favour of the powerful monarch, and make him the chief contributor

to the erection of the new building.

In the next place, the constructive idea informing the building is the same

Fig. 212. Constantinople. St. Mary Panachrantos.
Apses of the northern church (IXth or Xth and
XHIth or XlVth Centuries).

154

LOMBARDIC ARCHITECTURE

Fig. 213. Constantinople. St. Mary Pana-
chrantos. Apses of the southern church
(XHIth or XlVth Century).

as that of the Palatine Chapel, viz. that of
making the massive outer walls provide the
main element of stability for the internal
vaulting. So much is this the case that in
places where the thickness of the walls has
been diminished by the recesses hollowed
out on the inside, care has been taken to
compensate for it by external buttresses.

These considerations would lead one to
imagine that San Donate was designed by
the architect of the Palatine Chapel. But
such an idea cannot be maintained, for the
arrangement of the constructive elements is
too rude to allow of it. A glance at the
mammoth piers, set without regard to making
the intervals between them equal, will carry
conviction on this point. The Byzantine
School could not have produced an architect
so little sure of himself in the field of scien-
tific vaulting as to be obliged to sacrifice the
most elementary rules of proportion in order
to carry out his design with safety.

My belief is that the architect came from
Ravenna. The -presence of that school of
craftsmen is indicated by three character-
istics : the conical shape of the dome, the peculiar plan of the narthex, and the

blank arcading.

In the present state of things one cannot

say what idea guided the creator of San Donato

in his choice of so abnormal a plan for the

narthex of the church he was designing ; but

it is clear that it was connected with that of a

building well known to him, San Vitale at

Ravenna, from which he also derived the form

of his cupola.

As for the blank arcading which decorates

the exterior of the apses, only an architect of

the School of Ravenna can be credited with it.

With the Greeks apses had hitherto been pro-
vided : (i) with buttresses set on a high plinth,

with the object of keeping the wall firm against

the thrust of the vault, as in St. George (Vth

century) and Eski-Djuma at Salonica (Vth

century) ; (2) with arcades filled by transeniiae,

like St. Demetrius in the same place (Vth

century) ; (3) with small arched niches, as may

be seen in St. Sophia, also at Salonica (about

495). Of course the latter only applies when

... ' * Fig. 215. Constantinople. St. Saviour

such niches are not the result of insertions in Pantocrator. Apses (Xilth Century).

ARCHITECTURE IN ITALY AND DALMATIA

'55

the form of small windows intended to provide more light for the mosaics of the
apse, as in the case in St. Sophia at Constantinople (532-537).

It was not till later that the Greeks embellished their apses with blank
arcading. It is true that the east end of the church of St. Mary Panachrantos at
Constantinople shows this ornamental feature. But the building first erected by
Constantine Lips in the reign of Leo VI the Philosopher (886-912) was restored
by Theodora the mother of Andronicus II Palaeologus (i 282-1 328). 1 And owing
to the thick coat of plaster with which the structure has been covered it is impossible
to make even a cursory examination of the masonry with the object of distinguishing
different periods in it, and the alterations it has undergone, so as to be able to
decide how far it has
been tampered with.

The study which I
have made, and repeated
in the course of the last
few months, of this
and the other ancient
churches of Constantin-
ople, inclines me to be-
lieve that, of the two
small basilicas which
compose St. Mary Pana-
chrantos, the northern
one (Fig. 212) still pre-
serves the skeleton of the
original structure, and is
differentiated from the
southern one (Fig. 213)
by its apse and subordi-
nate apses, unequal in
height and dissimilar in
shape, round the exterior
of which runs a cornice
with a damaged inscrip-
tion mentioning Lips.
The alterations, how-
ever, which have taken

place in it are such that no argument can be based upon it, all the more when we
consider that the church belonging to the convent of Myrelaion (919-945), erected
shortly after Lips's work, has apses which are absolutely plain.

Interesting examples of this form of wall decoration are presented by the
apses of the churches of St. Mary Pammacaristos (Fig. 214), built by Michael
Ducas and his wife Maria, sister of the Emperor Alexius I Comnenus (io8l-iil8), 2
and St. Saviour Pantocrator (Fig. 215), erected by the Emperor John II Comnenus
(i i iS-i 143), or rather by his wife Irene, 3 both at Constantinople; by that of the
Holy Apostles at Salonica (Fig. 216), assigned by Bayet 4 to about the year 1012,
but certainly later than the church of St. Elias in the same place, as is shown by

Fig. 214. Constantinople. St. Mary Pammacaristos (Xlth or Xllth
Century).

128 Du Cange, Hist. By?. Constantinopolis Christiana.
* L'art by:anlin.

i 5 6

LOMBARDIC ARCHITECTURE

the more advanced decorative treatment of the exterior. It may be dated in the
second half of the Xlth century. Texier and Pullan J ascribed this church of
St. Elias to the Xth century. But the presence of pulvins and Lombardic cubical
capitals recalling those set up by Veremundus in the cathedral at Ivrea
(973-1001 or 1002), indicates a date not earlier than the Xlth. And the more
advanced character of the exterior decoration compared with that of the church
of the Virgin, also at Salonica (1028), will not allow us to place it before the

middle of that cen-
tury.

In Italy the
apses of churches in
the Pre - Lombardic
style up to the fall of
the Lombards (774)
were embellished
with arched corbel
courses divided into
groups by vertical
rolls, or by lesenas
as in San Pietro at
Toscanella (739) and
the parish church of
Arliano (712-744).
Later they began to
be also adorned with
deep arched niches
grouped by lesenas,
and of this form the
apse of Sant' Am-
brogio at Milan
affords the earliest
known example.

San Donato at
Zara was therefore
the first church to
exhibit the motive
of blank arcading
applied as decora-
tion to the exterior
of apses. The idea

of treating curvilinear walls in this way is a very old one. The vestibule leading to
the Piazza d'Oro in Hadrian's Villa at Tivoli (125-135) still displays on its outer face
a range of tall arches supported by rectangular pilasters. And as the School of
Ravenna had been alone in keeping an honoured place for this motive for several
centuries past, it is but logical and natural to credit the same School with the idea of
applying it to the apses as well as the side walls of churches.

As for the builders of the church at Zara, in all probability, if not all, at least
some of them were the same as those who worked on the Palatine Chapel at Aachen,

1 Op. dt.

Fig. 216. Salonica. Church of the Holy Apostles (Xlth Century).

ARCHITECTURE IN ITALY AND DALMATIA

as is indicated by the masonry. That is to say they were Ravennate and Comacine

masters, but not working under the direction of Byzantine master masons, as is

proved by the timidity and want of

finish in the construction of the

vaulting. Their presence is also

indicated by the few specimens of

carving still surviving, executed ex-
pressly for the church. Thus the

outer archivolt (Fig. 217) of the door

leading to the staircase of the gallery

(in which for the first time outside

the easternmost geographical boun-
dary of Italy we find the motive of

cauliculi used as a cresting) reveals

the hand of a Ravennate carver and

not one of the best, a fact which

can be verified by a glance at the

archivolts of the ciborium of St.

Eleucadius in Sant' Apollinare in

Classe (IXth century), which also

come from a Ravennate hand. On

the other hand, the rude carving on

a capital in the gallery betrays a

Comacine chisel.

The presence of Comacine and

Ravennate carvers at Zara in the IXth century is no less clearly suggested by some

carvings in the Pre-Lombardic style, to be ascribed to that source and date, which are

preserved, with others of the Middle Ages, in the Museum attached to San Donate.

Specimens of the Pre-
Lombardic style may be
seen in other Dalmatian
towns, alike in com-
position, design, and
execution, revealing a
Ravennate or Comacine
hand of the same date.
Among them may be
mentioned the ciborium
archivolt built into the
wall over the sacristy
door in the cathedral at
Cattaro (Fig. 218). It
was executed for An-
dreaccio Saraceni, the

Fig. 217. Zara. San Donate. External door of staircase
(about 801-806).

Fig. 218. Cattaro. Duomo. Archivolt of ciborium (809).

founder of the cathe-
dral (809). 1

Before leaving the Dalmatian coast, a few remarks may be made. Much has
been written about the very interesting buildings of this region ; but their true origin

1 Jackson, of. fit.

158 LOMBARDIC ARCHITECTURE

and real, not imaginary, merits have never yet been stated. When that is done, and
when those origins and merits have been freed from the mists of prejudice which have
hitherto involved them, I believe that not a few surprises will come to light. Thus,
for instance, it will be found that, with the exception of two new elements of decora-
tion, viz. blank corbel arcades and zigzag bands in relief used for a cornice (a feature
which had its origin in the bands of painting or ornament used by the Etruscans,
Greeks, and other early peoples), the buildings of Diocletian at Spalato, which are
merely an echo of older Roman Imperial structures of much vaster proportions, do not
exhibit a single original motive. I will not even make an exception in favour of the
spurred column bases, of which Choisy 1 gives a specimen ; for in spite of the most
minute search which I made in the palace, I was unable to discover in the ancient

Fig. 219. Zara. San Grisogono (n/

portions a single column base with these adjuncts, even of a merely geometrical
kind.

In the same way it will be discovered that not a few of the Dalmatian ecclesi-
astical buildings to which a great antiquity has been ascribed, either because they are
mentioned by Constantine Porphyrogenitus, or for other reasons, will be found to be
less ancient than is generally supposed. It will also come out that the style of carving
which I call Pre-Lombardic only made its appearance in Dalmatia some time after the
end of the VHIth century, that is to say, after it had been already created, practised,
and diffused through Italy by the Comacine gilds. Another discovery will be that
the Pre-Lombardic carvings preserved in museums, or still remaining in Dalmatian
churches, were the work either of Ravennate or of Lombard masters. And in the
latter case this will be a proof of the statement 2 that the IXth century Bans of Croatia,

1 Histoire de r architecture.

- Bulic, 7 niommtnti croati nel circondario di Knin ed altri contemporanei trovati altrcn<e in Dalmazia
cielF epoca nazionale croa/a, in Opera academiae scientiantm et arliitm Slavoritm iiieridionalium.

ARCHITECTURE IN ITALY AND DALMAT1A

'59

who went on pilgrimage to Cividale (a fact of which we have unquestionable docu-
mentary evidence), brought back thence, or from Lombard/ generally, Comacine
masters to undertake the construction of churches in the districts under their
jurisdiction. In other cases they will be found to be late imitations of the work of
those masters, rude even to the very lowest degree as being the productions of local
carvers. And accordingly, some of these carvings attributed to the VHIth or IXth
century will find their proper place in the IXth and Xth, or even in the Xlth century.

Further, it will be
made clear that none
of the leading charac-
teristics of the Lom-
bardic style appeared
in Dalmatia before
they had been in vogue
in Italy. Finally, it
will be found that the
finest Xllth and
XHIth century Dal-
matian churches, such
as the cathedral of
Zara, rebuilt by Arch-
bishop Lorenzo Peri-
andro (1245-1287),
and consecrated in
I285, 1 San Grisogono,
in the same town (Fig.
2 1 9), dedicated in 1175
by Archbishop Lam-
pridio (1146- 1 179), 2
and the cathedral at
Trau (Fig. 220) which
the Florentine Bishop
Treguano (1206-1254
or 1255) in 1213 3 had
built as high as the
roof and covered in,
are one and all pure

imitations of older Italian churches in the Lombardic (Fig. 22 1) or Lombardo-Tuscan
styles. Or else, while being imitations of such churches, they are carried out with a
different distribution in some parts of the building of the decorative architectural
elements, so as to give them a certain air of novelty, and allow them to be regarded
as one of the many varieties of the Lombardic style. And to this variety we may
give the name Lombardo-Dalmatian.

We will not waste our time over the astounding statement made by some writers
that the Croats of the Vllth and VHIth centuries, because they were neighbours of
the Byzantines, first learned, and then made their own, the art which I call Pre-
Lombardic, so that they were able to introduce it into Friuli, whence it spread

1 Karlali, fllyrifum Sacrum. Ecclesia laderlina. 2 Bianchi, Zara Cristiana.

J Farlati, lllyricum Sacrum. Episcopi Tragurienses.

Fig. 220. Trail. Duomo (1206-1254 or 1255).

i6o

LOMBARDIC ARCHITECTURE

throughout Italy. A barbarous people like the Croats was incapable of working such
a miracle, either at that date, or even at the end of the Xlth century, as may be
gathered from William of Tyre : 1 " Dalmatia .... populo ferocissimo, rapinis et
caedibus assueto inhabitata .... exceptis paucis qui in oris maritimis habitant, qui
ab aliis et moribus et lingua dissimiles, Latinum habent idioma ; reliquis Sclavonico
sermone utentibus et habitu barbarorum." Moreover, how could they acquire from

the Byzantines an art to
which the latter were
strangers, while they
were adepts in another
which in composition,
design, and technique is
so different from the Pre-
Lombardic ? On the
other hand, the records
of the life of a national
art are to be found, not
in the imaginative asser-
tions of writers, but in
monuments of ascer-
tained date ; and such
are entirely wanting to
the proof of the exist-
ence of the supposed
Croatian School.

Nor is there any
more substantial found-
ation for the belief in an
imaginary influence on
Dalmatia of Northern
and Transalpine styles
of art through the me-
dium of Hungary: 2 an
influence specially mani-
fested in the cathedral of
Trau through the links
connecting that building
with the church of Jak,
which is dated, without a single fact to support the statement, in the middle of the
XHIth century. For, above all, we must remember that the cathedral of Trau is
indissolubly connected with the name of an Italian, its second founder, Bishop Treguano ;
without taking into account that it may well have been that the undated church at
Jak was erected by Dalmatian hands, and was derived from Trau, and not vice versa.
In any case, all the most salient features of the cathedral of Trau may have been
derived by its architect from Italy, where they had been invented and practised
long before they were to be seen beyond the mountains and beyond the seas.

1 Migne, Pair. Lat., Vol. 201. Guillelmus Tyrensis archiepiscopus Historia rerum in partibus trans-

Fig. 221. Bergamo. Santa Maria Maggiore (1137).

mannis gestantm.

Jackson, op. cit.

CHAPTER V

PRE-LOMBARDIC ARCHITECTURE

FROM THE CONQUEST BY CHARLES THE GREAT DOWN TO THE APPEARANCE OF
THE LOMBAKDIC STYLE, AND ITS COMPLETION

NEW styles of architecture are never new creations of the human intellect,
but are always evolved from older forms. Art never dies ; and though
we sometimes see it lying prostrate like a corpse, yet it is always
ready to rise up again when the breath of life has once more been
breathed into its frame.

This bringing back to life, when art re-awakens to mark a new epoch of culture,
does not happen instantaneously nor on a single occasion, but more or less slowly
and by stages, in which efforts to climb the rugged ascent, under conditions of peace
and plenty, alternate with intervals of suspense, preparation, renewal.

And so, the architectural movement which took place, first in Italy, and then
beyond the Alps, in the course of the Xlth century, was preceded by a long period
during which the monuments of Rome and of Ravenna were studied, and laborious
experiments were made again and again which were destined, by means of a fusion
of the Pre-Lombardic with the Romano-Ravennate and Byzantino-Ravennate
elements, to transform Roman into Lombardic art.

We have already seen how, as far back as the time of Theodelinda (590-625)
and Agilulf (590-615), art had awakened from the lethargy into which it had sunk
under the barbarian invasions, pestilence, famine, and flood, from which Italy had
suffered, and how from that period down to the fall of the Lombard Kingdom it had
had opportunities, thanks to the piety of the Lombard princes, both for its exercise
and for making some progress along the path which was to lead it to the new forms
which made their appearance after the Xth century. And we shall see, when the
time comes to describe the monuments of Germany, how the short-lived
resurrection of Byzantino-Ravennate architecture which followed Charles the Great's
conquest (774) contained a new germ of life for the Pre-Lombardic style, which was
destined to hasten its development.

Let us now see when and where this germ had the opportunity not merely of
showing the bud, but also of producing first the flower and then the fruit of
Lombardic architecture.

It is well known that Angilbert II, archbishop of Milan (824-860), as early as
the reign of Lothair I (840-855) and Pope Sergius II (844-845), shook off the
authority of both king and pope. It is also clear that his work, interrupted under the
weak Tado (861-869), was continued by the proud prelate Anspert (869-882) ; and
it is equally well known that from Angilbert II to Aribert (1018-1045), who closed
VOL. I '61 M

1 62 LOMBARDIC ARCHITECTURE

the epoch of the archiepiscopal regime at Milan, the occupants of that see stood forth
as the emancipators of North Italy from foreign rule only to make it the slave of
their own ambition. What wonder, then, if the archbishops, mindful of the fact that
great buildings are intimately connected both with pride and politics, affirmed their
wealth, their greatness, their authority, by rebuilding the fallen walls of the city, and
by the erection of sumptuous buildings both in it and in the districts subject to their
spiritual jurisdiction ?

It is, in fact, just at the period when the chair of St. Ambrose was filled by the
two prelates who were the founders of the supremacy of the archbishops of Milan,
that I place the beginning of the compromises, the experiments, the search for
elements which, when developed, would enable ecclesiastical architecture to expand
in new directions destined to mark the victory of the Christian religion in the West,
just as Byzantine architecture had already marked its triumph in the East, only that
the latter had had the advantage of being formed and developed under the protective
shadow of the new capital of the Empire, after Christianity had become the official
religion.

The stages of development which were traversed between this starting-point
and the first appearance of the Lombardic style may be, to all intents and purposes,
summed up as follows.

The primary interest of the Lombard gilds was the study of vaulting
construction and the art of counterbalancing its thrust. This study had been
till then neglected by them, as they were too timid either to emancipate themselves
from the regular forms of the Latin Christian basilica as it existed in the
centuries following the publication of the Edict of Milan (313), with its exclusive
use of wooden roofs for nave and aisles, or to attack problems to which their
constructive or statical attainments were unequal. So that, down to Carolingian
times, they confined the application of vaulting to the apses and rather limited crypts
of their basilicas. Possibly, too, it was used in baptisteries, none of which, however,
have survived ; but it is probable that both their forms and dimensions were of a
modest character.

And so, their first attempts to put these studies into practice were devoted to
vaulting the whole of chapels and baptisteries which, in contrast to those just
mentioned (sometimes without even vaulting to cover them), exhibited forms which
grew steadily more complex and developed. At the same time they began to vault
the bays of a church in front of the apse, in cases where it became necessary to
increase the chancel space, with the result that it was only the body of the church
which continued to be covered by an open timber roof or ceiling.

In these buildings, too, they began to use external buttresses corresponding to
the arcades of the interior, with the object of increasing the stability of the structure,
and also of indicating its internal constructive arrangement. Later they proceeded
to throw across the aisles transverse arches springing from compound piers and wall
piers, again with the idea of stability. Next they threw arches across the nave,
supporting them on substantial cruciform piers, alternating with columns or smaller
piers, and on wall piers, thus binding the whole structure together in an organic
unity.

The next task was to cover the naves of their churches, and this they did
first with barrel vaults, strengthened by transverse arches ; then with cross combined
with barrel vaulting of the previous kind ; lastly with raised cross vaults, with the
groins strongly emphasized thoughout their whole course, or else ribbed to serve

THE RISE OF THE LOMBARDIC STYLE 163

the triple purpose of facilitating the construction of the vaulting, and of increasing
both its strength and decorative effect. From these ribs were developed the vertical
shafts which resulted in giving a more complex and at the same time more varied
form to the compound piers. Groined vaulting of this kind was at first only used
for the aisles. Later it was extended to the nave, and this increased the variety and
complexity of the forms of pier support.

Such was the evolution of the vaulted Lombardic basilica. Each of the phases
we have described was accompanied by motives of decoration and ornamentation
which were either original, or borrowed from the Pre-Lombardic architecture of the
Lombard age, or else from the Romano-Ravennate and Byzantino-Ravennate styles
which preceded it.

The buildings which we shall pass in review, forming a series of types arranged in
chronological order, will furnish the evidence for this evolution. At the same time
they will provide ample materials for following out the various phases in the creation
of the Lombardic style, and for determining the characteristics which distinguish
the Pre-Lombardic buildings of the IXth and Xth centuries, providing thus
considerable assistance for future chronological classification. In this examination
we must include some monuments which, though they throw no new light on the
origins of Lombardic architecture, will serve to illuminate our path when we come
to the second part of our work.

THE CHURCH OF AGLIATE. Giulini, 1 on the authority of a biographer of the
archbishops of Milan, whose work, some three centuries old, he possessed
in manuscript, refers to the foundation of the collegiate church of Agliate by
Anspert (869-882), remarking that, although we know nothing of the evidence
on which this statement is based, at the same time it is not to be treated with
contempt. I think it more likely that the church was built under Angilbert II
(824-860), to take the place of an older structure believed by some to have been
erected by St. Datius in the middle of the Vlth century, and that Anspert afterwards
endowed it with a college of canons. This theory turns out to be a fact when
we come to examine the building, and subject it to that comparison with others
of the same period which is always so fruitful of results.

It consists of a nave and aisles with wooden roofs, separated by stone columns
surmounted, with one exception, by ancient altars, inverted bases, sepulchral cippi,
turned into capitals which support tall misshapen abaci. The nave and aisles
terminate in three semicircular apses, the bays in front of which are vaulted and
separated from one another by walls.

The material of which the rough masonry is composed is mainly large pebbles
worn smooth in the bed of the Lambro, and this accounts for the prevalence of
" opus spicatum " in it. The building is finished with a simple cornice of tufa
on which the roof rests. It is lighted by numerous narrow round-headed windows
of double or single splay, and also by two luminous crosses, one in each gable of the
nave, intended to ventilate the timbers of the roof. It should be mentioned that
there are no windows in the wall of the north aisle.

The doors are crowned by relieving arches. The principal door in the west
front, before the modern sculpture was put up, had curvilinear interlacing decoration
of IXth century character.

The crypt underneath the apse and chancel is not original : the form of its

1 Op. cit.

M 2

164

LOMBARDIC ARCHITECTURE

vaulting indicates a reconstruction which probably took place about the year 1000.
But the capitals in the Pre-Lombardic manner of the VHIth and IXth centuries are
original. One of them, a cube bevelled off at the angles and decorated with leaves
carved out of a niche, cauliculi, and interlacing a skilful piece of work for the time,
and certainly not the deplorable production of some rude rustic carver, as it has been
described, may be compared both in design and execution with the original capitals
in TheodulPs church at Germigny des Prs (801-806), and with the one which
we have already noticed in the old crypt of San Filastrio at Brescia (Vlllth century).
This circumstance, taken in connection with the likeness between the capital
in question and the original ones in the church of San Satire at Milan (876),
and also with the more advanced character of the latter, gives good ground for
fixing the date of the basilica of Agliate in the IXth century ; earlier, however,

Fig. 222. Church of Agliate (824-860).

than the episcopate of Anspert (869-882), and, to be precise, in the time of
Archbishop Angilbert 1 1 (824860).

In the arches of the nave, and the archivolts of the windows and doors (with
the exception of the great west door), are interpolated voussoirs of tufa and bricks
framed by a ring of bricks laid horizontally. This makes a simple but elegant
decoration, contrasting with the rough surface of the wall. It is quite a new feature,
and does not occur, so far as I know, in any other building earlier than the IXth
century. It was derived from the arches of tufa alternating with bricks, in use among
the Romans from the Ilnd century onwards. Their appearance may be seen in the
ruins of the Villa known as the " Sette Bassi " on the Via Latina near Rome.

The outer wall of the apse (Fig. 222) is decorated at the top by a range of
small arched niches, grouped in threes by lesenas which carry the eaves cornice,
just as in the central apse of Sant' Ambrogio at Milan. The two chapels flanking
the apse have on the outside buttresses corresponding to the arcades of the
interior, so that this is another reason for placing the church in the same period
as that in which the eastern end of the Ambrosian Basilica was given its present
form, that is to say the time of Angilbert II.

THE RISE OF THE LOMBARDIC STYLE

165

THE BAPTISTERY OF AGLIATE has the plan of an almost regular nine-sided
figure, two sides being taken up by an apse for the altar. The outer facing of
its rough walls shows that it belongs to the same age as the neighbouring church.
This is proved both by the character of the materials and the way in which they
are employed.

Each side of the polygon, except the one to the north, is pierced high up
by a rather narrow double splayed window. Above the windows runs an arched
corbel course, and above this a range of arched niches surmounted by a plain
tufa cornice on which the roof rests. The building is covered by a cupola.

Its form recalls that of the baptistery belonging to the cathedral of Grado
(571-586), but it is distinguished from it by the number of the sides of the
polygon, by the shape of the apse, and by its relative position with respect to the
church which it was intended to
serve.

The extensive use of "opus
spicatum " in both buildings at
Agliate demands some notice. It
is a method of construction which
goes back to a remote period. Em-
ployed in brick floors as long ago
as the time of Augustus (29 B.C.-
14 A.D.) (" item testacea spicata tibur-
tina sunt diligenter exigenda "J, 1 in
the decadence of the Roman Empire
it passed from the pavement to walls,
as may be seen from the town walls
of Susa erected between the IVth
and Vlth centuries (Fig. 223). North
of the Alps, an ancient instance of
walling material laid herring-bone
wise is afforded by the remains of a
Roman villa discovered at Littleton
in Somerset, and the same thing
could be seen not long ago in a

Roman building at Castor in Northamptonshire. Both belonged to the period
between the Illrd and Vth centuries.

In Italy the use of "opus spicatum" for walls had a very varied history. At
Rome it never took hold. There are only very scanty traces of it in the walls of
Rome, that mosaic of constructive methods beginning with the reigns of Aurelian
(270-275) and Probus (276-282) and coming down to our own times, which, if
ever they are made the subject of a careful study, will be found to provide a
chronological picture of the conditions and practice of the art of building at Rome
during some sixteen hundred years. Nor was it more successful in establishing
itself at Ravenna. It is rare, again, at Pavia. On the other hand, it was very
popular at Verona and in North Italy generally, especially in Lombardy, where we
not only find it frequently employed (particularly in buildings from the VHIth to
the XVth century), but also more firmly rooted than in any other region of Europe.
And this leads me to suspect that its birthplace in Italy was Cisalpine Gaul.

1 Vitruvius de archileetura, VII.

Fig. 223. Susa. " Opus spicatum " in the town walls
(IVth-VIth Centuries).

i66

LOMBARDIC ARCHITECTURE

The suggestion for the method probably comes from the fact that builders in
the districts near the Alps were obliged to make use of the large pebbles which had
been worn smooth in the beds of the rivers. Given the use of such materials, the
arrangement of it which was adopted was the more intelligible because, compared with
the irregular method known as "opus incertum " generally used for pebble con-
struction by Roman builders, it offers the double advantage of not requiring so
much mortar and of producing a more pleasing effect for the eye.

It is probably the Comacine gilds who are to be credited with the transmission
of " opus spicatum " through the darkest ages of barbarism, for they were the largest
users of it throughout the Middle Ages : above all, the gild of Campione, which I
believe to be responsible for the finest-herring bone work known.

Its employment was sometimes due to a mere whim of the builders (perhaps
with the object of indicating the participation of a particular gild in the work), or else
to the convenience of using up chippings, as at Milan in Sant' Ambrogio, San
Vincenzo in Prato (835-859), San Calimero, the apse of which was ascertained by
Cattaneo 1 to belong to the IXth or Xth century, San Celso, rebuilt in 996 by Arch-
bishop Landulf (980-998)^ and Sant' Eustorgio which, as we shall see, must go back
to the Xth century, and others. In these churches the "opus spicatum" is generally
formed of bricks smaller than those used for building, but also sometimes of broken
tiles shaped for the purpose, laid end-wise in courses, or in two rows separated by a
course of bricks or merely a bed of mortar.

In other cases its use depended on constructive needs, as in the outer wall of the
Castle of the Visconti at Agrate (XlVth century) and the numerous private buildings
grouped round it. They are all built with the large pebbles which abound in the
district, combined with some fragments of stone or brick. Another instance is the
wall of the Castle of Bramafam near Aosta (Xlth century). In a third class it was
due to a combination of decorative and constructive purposes.

BASILICA OF SAN VINCENZO IN PRATO AT MILAN. The monastery was origin-
ally founded by the Lombard king, Desiderius (770). When Archbishop Angilbert II

in 835 took its abbot Gau-
dentius (f 842) to transfer
him to Sant' Ambrogio, the
convent of San Vincenzo
merely possessed a chapel,
as it did during Odelpert's
tenure of the see of Milan
(8o5-8i4). 3 Some have been
surprised that a monastery
of sufficient importance to
provide an abbot for Sant'
Ambrogio should have
nothing better than a chapel,

1 Op. at.

- Giulini, op. cit.

3 7'hesaunis antiqttitaliitn et
liistoriarum Italiae transpadanae et
Alpibus vicinae. Puricelli, Am-
brosianae Mediolani basilicas ac
Fig. 225. Milan. San Vincenzo in Prato. Capital in the crypt (770). monasterii &c.

THE RISE OF THE LOMBARDIC STYLE

167

and an imaginary basilica has been created. But, to take one instance, the monastery
of Saint Guilhem du Desert, originally at Gellone, founded by William, Duke of
Aquitaine, after he became a monk, possessed at first only a chapel. 1

The erection of the present San Vincenzo is consequently later than the year
835. With this occasion we may connect the notices 2 of donations to the monastery
made by Abbot
Giselbert, the succes-
sor of Gaudentius,
and of legacies be-
queathed to him. In
859 the structure
must have been
finished, as the
bodies of Saints
Quirinus and Nico-
medes were trans-
lated into it. 3 Cat-
taneo 4 thinks that
this took place at
the consecration of
the church.

The exterior
still retains its origi-
nal features. The
apses (Fig. 224), with
the range of arched
niches and their
arched corbel courses
divided into groups
by lesenas, copy the
decorative forms of
those in Sant* Am-
brogio. The side and
front walls are plain.
The pediment, how-
ever, of the latter,
like the correspond-
ing one at the east

end of the church, is decorated by an arched corbel course following the slope of the
gable, and by a small cruciform window enclosed in a frame, an original motive to be
seen in some churches of the Lombard ic and derived styles.

The interior, divided by columns into a nave and aisles with wooden roofs, shows
the alterations made in the early years of the Xlth century, or, to be more accurate,
about the year 1023, when Archbishop Aribert (1018-1045) deposited the body of
St. Abundius in the church. 6 The changes which then took place are seen in, among
other things, the spacious and lofty crypt beneath the chancel (the old crypt was

1 Mabillon, Acta sanctorum ordinis S. Benedict! Vita S. Willelmi duds ac monachi Gellonensis in Gallia.
a K. htituto lombardo di scienze e lettere, 1868. Belgioioso, La basilica Milanese di San Vincenzo in Praia.
3 Giulini, op. cit. * Op. (it. 5 Giulini, op. fit.

Fig. 224. Milan. San Vincenzo in Prato (835-859).

i68

LOMBARDIC ARCHITECTURE

apparently confined to the space under the apse, the floor of which would be only a
little higher than that of the church), where the cross vaulting, with its strongly
marked groins and visible arches, exactly corresponds to the date just mentioned.
In this crypt may be seen a capital (Fig. 225) belonging to the same period as

Fig. 226. Milan. San Vincenzo in Prato.
Capital in the nave (about 1023).

Fig. 227. Como. Sant' Abondio.
Capital (1013-1095).

another in the left-hand arcade in the church, and exhibiting the Ravennate style of
the Vlllth century. Both belonged to the original foundation of 770.

Another indication of the changes made is the presence in the nave of capitals
(Fig. 226) recalling the type of one (Fig. 227) in the interior of Sant' Abondio just
outside Como (1013-1095), and consistent with the suggested date.

THE CHURCH OF SAN SATIRO AT MILAN was
erected by order of Archbishop Anspert, together with
the adjacent bell-
tower, in 876. 1 The
fact is mentioned
in the prelate's epi-
taph, and in his
will which may be
found in Muratori. 2

The sketch
of the original
ground-plan, pre-
served in the Am-
brosian Library at
Milan, 3 recalls in its
internal arrange-
ments that of the
church of Germigny

Fig. 228. Milan. San Satiro.
Capital (876).

desPre"s (801-806).
The exterior origi-
nally exhibited a
succession of exe-
dras. It was, and is, vaulted in all parts.

Fig. 231. Milan. Capital in tower of the
Monastero Maggiore (869-882).

The domed

Fig. 229. Milan. San Satiro.
Capital (876).

1 Giulini, op. cit.

2 Antiq. Ital. medii aevi Dissertatio L VI.

3 fiassegna cfarte, 1901. Beltrami, Bramante e Milano.

THE RISE OF THE LOMBARDIC STYLE

169

tower which it must have originally possessed has been replaced by the present
octagonal cupola.

Two Pre-Lombardic cubical capitals (Figs. 228, 229) are to be noticed, belonging
to the age of the founder. They recall one which we observed in the church at Agliate,
and show the forms and ornamental motives followed by the Lombard carvers for
capitals in the second half of the IXth century. These artists further left the mark
of their chisels on five of the eight capitals in the square tower of the Monastero
Maggiore at Milan (Fig. 230) belonging to the time of Archbishop Anspert (869-882).
I mean those of cubical form (Fig. 231) carved out of the same block of stone as the
shaft and base, and carry-
ing pulvins from which
spring arches showing a
reconstruction later than
the period about the
year 1000.

By itself, San Satiro
would possess only a
small amount of interest
for us were it not for
the adjoining campanile
(Fig. 232). This is a
massive square tower,
almost entirely built of
brick, the highest stage,
i.e. the bell-chamber, be-
ing evidently a later
addition. Its date is very
important, and confers
on it the claim to be
considered the prototype
of the Lombardic cam-
panile. Previously, bell-
towers on a large scale
did not present an archi-
tectural scheme like that
which characterizes the
campaniles of the Lom-
bardic and derived styles,
and that of San Satiro

Fig. 230. Milan. Tower of the Monastero Maggiore (869-882).

among them. The two oldest examples that I can adduce the "Monks Tower' of
Sant' Ambrogio at Milan (789-824^ and that of Santa Maria della Cella at Viterbo
(IXth century), are sufficient evidence of the statement.

The former of these, which is recognized as having been erected after the
Benedictines had been settled in the new monastery of Sant' Ambrogio (789),
was increased in height by one story, the present bell-chamber : this has been made
clear recently when the walls were stripped. In the original part the only artistic
elements are the supports of the two-light openings constructed of materials brought
from elsewhere.

The latter, which no one has previously used as evidence, in its topmost stage

170

LOMBARDIC ARCHITECTURE

exhibits (Fig. 233) two-light openings divided by octagonal shafts with cubical
capitals, chamfered at the angles, and supporting very tall pulvins moulded like
a cornice. Its date may be placed in the time of Pope Leo III (795-816), and
for this reason. It was the Lombard king, Desiderius (756-774), who raised Viterbo
from political obscurity by taking up his station there with his forces when he broke
with Pope Hadrian I. And it seems l that it was then that the town was provided

Fig. 232. Milan. San Satire.
Campanile (876).

Fig. 233. Viterbo. Santa Maria della Cella.
Campanile (IXth Century).

with the fortifications mentioned in a document of 808 in the Farfa Register. The
building of a church would naturally follow, viz. San Lorenzo, which has now
disappeared, and the foundation by the powerful abbey of Farfa of a convent
or " cella " as it was then called. As a matter of fact, it is in a document of 775 in
that Register that we find the earliest mention of the " Oratorium S. Mariae de
Cella " with monks and a " praepositus " appointed to administer its property. 2 The
possession of this "cella" of St. Mary in the castle of Viterbo was later, in 801,
confirmed by Charles the Great to the monks of Farfa. 3

1 Pinzi, G/i ospedali medioevali e I'ospedale grande di Viterbo. 2 Pinzi, Storia della citta di Vilerbo.

3 Pinzi, Cenni storici sulla chiesa e confraternita di Santa Maria aetta Cella in Viterbo.

THE RISE OF THE LOMBARDIC STYLE 171

It is with this last occasion that I believe the erection of the campanile is to be
connected. Leo Ill's return to Rome and the coronation of Charles the Great (800)
were the signal for a renewal of the activity in building in the City and Duchy of Rome
which had distinguised the pontificate of Hadrian I (772-795). We need not, then,
be surprised if the abbot of Farfa was also inspired to undertake fresh works in a
church which belonged to and was the offspring of his own abbey, the superior rights
of which had just been confirmed by the Emperor. As a matter of fact, the four
cubical capitals of Pre-Lombardic style in this campanile carry pulvins with the
features which made their appearance in Carolingian times, and of which examples
may be seen in the church of Germigny des Prds. The horse-shoe arches, too, which
spring from these capitals only made their first appearance, so far as the West is
concerned, in buildings of the early years of the IXth century, as we shall see when
we come to deal with Theodulf's church. These two features might almost make one
suspect that the man who designed the tower of Santa Maria della Cella had previously
taken part in the erection of that church before the Benedictines of Farfa employed
him in the works connected with the Cella at Viterbo.

Further, the facing of the walls shows, to an experienced eye, considerable
likeness to that of the walls of San Pietro at Toscanella (739), which is not the case
with the early Xlth century buildings at Viterbo, e.g. the church of San Giovanni in
Zoccoli (1037). This circumstance will put the campanile at a period not far from the
VHIth century, rather than in the Xlth, and at the same time suggests the presence
of Comacine masters.

THE PARISH CHURCH OF SAN LEO. The date is given by the ancient altar
ciborium constructed by order of Ursus, Duke of Ferento, in the time of the
Emperor Charles the Fat (881-887) and Pope John VIII (872-882), as is stated in
the still existing inscription published by Marini. 1 The stones and the four marble

Fig. 234. San Leo. Parish church. Fig. 235. San Leo. Parish church.

Capital of ciborium (881-882). Capital of ciborium (881-882)

columns of this ciborium may be seen worked into the present baptistery : the capitals
(Figs. 234, 235) are set round it.

These capitals, the date of which is certain, show, though with somewhat greater
refinement, the same Pre-Lombardic forms as some others (Fig. 236) used in the
original decoration of the exterior of the church, and evidently belong to the same
date. So that we need have no hesitation in saying that the church was built between
88 1 and 882. They are cubes hollowed out at the lower corners, with the recesses
filled by plain leaves bearing sometimes a rosette, or a lily, or a leaf with indented
outline, or striations. This design of leaves with leaves of a different kind or other

1 Saggio di ragioni della cilia di Sanleo delta gia Monleferetro.

172

LOMBARDIC ARCHITECTURE

ornaments carved upon them was freely used after the epoch of 1000. Among other
churches, the cathedral of Aquileia (Fig. 237), built by the patriarch Poppo (1017 or

1019-1042 or 1045), who consecrated it, according to Gams, 1
in 1027 or 1029, and altered by the later patriarch Marquard
(1365-1381), presents examples of it in the capitals of the
nave. Their faces show cauliculi, leaves, roses, bunches of
grapes, crosses curled at the extremities, interlacings, &c.

In the variety and novelty of motives (some of which
may be seen repeated in the Corinthianesque capitals of
Lombardic style belonging to the Xlth and Xllth centuries)
they surpass those in San Satiro at Milan (876), and are the
best certainly dated IXth century specimens which we can
point to either in Italy or north of the Alps.

The nave and aisles of the church of San Leo must
originally have been separated by columns, some of which
have been encased in the supporting piers constructed when

the present vaulting took the place of the old wooden roof. The nave terminates
in an apse flanked by two minor apses. All three, together with the presbytery,
are somewhat raised up owing to the crypt below, now closed and filled with
rubbish.

The exterior was decorated by arched corbel courses grouped by lesenas (Fig.

Fig. 236. San Leo. Parish
church. Capital from the
exterior (881-882).

Fig. 237. Aquileia. Duomo (Xlth and XlVth Centuries).

238). The original method of lighting the building was by rather narrow double-
splayed windows, and loops splayed inside. Of the three side doors, two are sur-

1 op. dt.

THE RISE OF THE LOMBARDIC STYLE

mounted by blank loggias projecting from the wall, with shafts finished by small
Pre-Lombardic cubical capitals. The front of the church is concealed by a modern
building.

The church of San Leo gives us several important pieces of information. It tells
us that in Italy, at the close of the IXth century, churches still retained the form of
the Latin basilica, and had wooden roofs with or without a ceiling for their naves and
aisles. It also tells us that in that period, though spacious crypts were constructed
under the chancels, they did not result in an excessive or even marked elevation of
the latter. Indeed, the chancel at San Leo, even taking into account the modern
raising of the level of the nave,
can only have been elevated by
a step or so above the floor of
the body of the church.

Further, it informs us that
the Lombard gilds, who alone
can be regarded as responsible
forthe construction of thechurch,
still followed for capitals the
Pre-Lombardic cubical forms of
the VHIth century, and that
the Lombard ic types had not
yet come into existence.

And it teaches us that in
the IXth century these gilds
favoured two types of external
decoration for their churches.

The first was a rather rich
treatment, but applied on prin-
ciple only to the most important
part of the building, i.e. to the
apse. To this type belong the
churches of Agliate (824-860)
and San Vincenzo in Prato at
Milan (835-859); and, in all
probability, Sant 1 Ambrogio at
Milan, as built by Angilbert
(824-860), was another instance.

The second type of decora-
tion was simpler, but extended,

so far as we can judge, at least to the side walls of the building. To this type
are to be referred the parish church of San Leo, and the church of San Pietro al
Monte at Civate (Fig. 239), the original portions of which (belonging unquestionably
to a church with a single apse, the one at the opposite end being a later addition)
are ascribed, 1 and I think rightly, to the IXth century, and not later than the
year 860.

Finally, it tells us that in the IXth century, as before, the Italian artists were
more successful in carving slabs than capitals. For proof one has only to compare
the carving of the capitals belonging to the ciborium at San Leo with that of the

1 Archivio storico hmbardo, 1896. Magistretti, San Pietro al Monte di Civate.

Fig. 238. San Leo. Parish church (881-882).

174

LOMBARDIC ARCHITECTURE

marble tympanum (Fig. 240), the work, I should say, of a Ravennate hand, built into
the exterior of the south aisle of the cathedral of Pola, which was consecrated in 858

Fig. 239. Civate. San Pietro al Monte (IXth Century).

by Bishop Andegisus (854-859). 1 The movement of the birds flanking the bishop's
monogram, and the technique shown by them and also by the peacocks on either

Fig. 240. Pola. Duomo. Tympanum on the exterior (857).

side of the inscription, recall the well-known pluteus in Santa Maria degli Angeli,
near Assisi, which is believed to be a production of the IXth century.

THE BASILICA OF SANT' EUSTORGIO AT MILAN was erected on the ruins of the
old basilica of the same name belonging to the first half of the IVth century.
Cattaneo 2 believes that portions of it, viz. the apse and the two plain arches at the
extremity of the nave supported by piers, are work of the end of the IXth century or

1 Gams, op. cit. * op. cit.

THE RISE OF THE LOMBARDIC STYLE

the beginning of the Xth ; while he regards the rest as a reconstruction of the Xlth
and following centuries. And he seems to me to have almost exactly hit the mark.

The decorative treatment of the exterior of the apse (Fig. 241), where the brick
facing shows a free use of regular courses of " opus spicatum," gives ground for
placing the rebuilding of the ancient church between the period in which San
Vincenzo in Prato was erected (835-859) and the year 996, when the reconstruction of
San Celso took place,
the apse of the latter
belonging to this oc-
casion. The reason
is that, on the one
hand, the niches
round the apse of
Sant' Eustorgio are
still low, and the
lesenas project very
little from the face of
the wall and merge
in the arches which
frame the niches, as
in the apse of San
Vincenzo in Prato
where the lesenas
measure about i ft.
X 3 in. On the other
hand, the niches are
not as yet so elon-
gated as those at
San Celso, and the
lesenas, though they
no longer group the
niches into threes,
the scheme followed
both in the apse of
San Vincenzo in
Prato and in those
of Sant' Ambrogio
(/89-824),thechurch
of Agliate (824-860),
and San Calimero at

Milan (IXth or Xth century), are not strengthened by sturdy buttresses as in the
apse of Landulfs church. The latter method of giving support was suggested by
Roman work, as may be seen from Montano, 1 and from the ruins of a building with
an apse in the Villa of the " Sette Bassi "on the Via Latina near Rome (Ilnd century).
Here the apse is kept up by three massive buttresses (Fig. 242).

The large brick piers of T form belonging to the two easternmost arches of the
nave, discovered during the last restoration, and found to belong to the original
church which was altered after the year 1000, enable us to fix more precisely the date

1 Op. eit.

Fig. 241. Milan. Sant' Eustorgio. Apse (Xth Century).

i 7 6

LOMBARDIC ARCHITECTURE

at which it was rebuilt, and place it definitely in the first half of the Xth century.
We may reasonably infer that a complete series of such piers originally divided the
nave from the aisles ; and they were intended to carry, not only the longitudinal
arches of the nave, but also the transverse ones which spanned the aisles. The
nave, apparently, was not so treated on account of its considerable width about
40 ft. Though churches of Roman basilica type were to be seen in Italy before
the Xth century with nave and aisles separated by piers of "]~ shape, e.g. San Vittore
at Ravenna (Vlth century) and the parish church of Bagnacavallo (Vlth century),
such piers supported only longitudinal arches. Even in the first half of the IXth
century, and at Milan, the nave of San Vincenzo in Prato was separated from the
aisles by columns which carried the longitudinal arches and no others. Nor was the

Fig. 242. Rome. Villa called " Sette Bassi." Apse with buttresses (Ilnd Century).

case different with the church of San Leo (881-882) in the second half of the century.
But in the second half of the Xth century the organic conception of transverse arches,
though confined to the aisles, which was evolved in Sant' Eustorgio, is found in its full
development in SS. Felice e Fortunate near Vicenza, where transverse arches
probably spanned the nave as well. And at the dawn of the Xlth century San
Babila at Milan was able to show Lombardic piers and a complete system of vaulting
for both nave and aisles.

THE PARISH CHURCH OF MONTALINO AT STRADELLA has a nave and aisles
separated by piers, and ends in a central and two minor apses, one of the latter
having been destroyed to make way for the present campanile. The piers consist of
a rectangular block with two half-columns on the smaller sides and two roofing shafts

THE RISE OF THE LOMBARDIC STYLE

Fig. 243. Stradella. Church of
Montalino. Capital (Xth
Century).

on the larger. Each column ends in a large torus, above which is a Pre-Lombardic

cubical capital with the lower corners bevelled off (Fig. 243). From these spring

the semicircular longitudinal arches which carry the walls

of the nave. The roofing shafts originally ran up to the

top of the walls to support the tie-beams of the roof which

extends unbroken over both nave and aisles. To these

four cruciform piers correspond a similar number of wall

piers, which were also carried up to support the beams of

the roof.

Nowadays the walls of nave and aisles show traces
of alteration at the top. The members which supported
the roof have been cut short, and the roof itself recon-
structed without regard to their original functions. The
adoption of piers and wall piers to carry the framework of
the roof has enabled the outer walls and those of the nave
to be reduced in thickness. The original windows were
mere loops, round-headed and splayed on both sides.

The exterior was decorated with arched corbel courses

interrupted by lesenas. In the apses (Fig. 244) the eaves cornice is composed of a
cable moulding bordered by the saw-tooth ornament. This is the first appearance
of this moulding as an element of decoration in Lombardic buildings. Its use was
due to the School of Pavia, which borrowed it from the Romans ; and we may see
it in San Michele, and San Pietro in Ciel d'Oro at Pavia, and other churches.

The date of the church at Montalino is not ascertained. Still, we know that

Stradella or Montalino
was given in 943 by the
Italian kings Hugo and
Lothair to Litifred, bishop
of Pavia, who, according
to Gams, 1 filled the see
from 939 to 967. The
donation was confirmed
by the Emperor Otho II
to another bishop of
Pavia, Peter III (978-
983). 2 And it is more than
likely that the bishops of
Pavia, having acquired the
lordship of Montalino,
undertook the erection of
a parish church. If so, it
must have been built after
the year 943, or at latest
after 977.

The formsof the origi-
nal windows and of the
capitals suggest a date nearer to the IXth than the Xlth century: and the years of
Litifred's episcopate following 943 are probably those which saw the erection of the

1 Op. cii. a Cavagna Sangiuliani, IM basilica di San Marcello in Montalino.

VOL. I N

Kig. 244. Stradella. Church of Montalino. Apses Xth Century).

I 7 8

LOMBARDIC ARCHITECTURE

building. This date would be confirmed by the presence of piers which no longer
follow the early T form of those in Sant' Eustorgio at Milan (first half of the Xth
century), while at the same time they are not yet complete compound supports with
continuous Lombardic capitals, as in SS. Felice e Fortunate near Vicenza (985).

It remains to be noticed that the constructive idea which guided the Lombard

gild in carrying out the work, viz. to make the
piers bear the framework of the roof, and to con-
struct them from the base upwards as though they
were to carry vaulting instead of a timbered roof,
forms a sort of anticipation of the more developed
plan of SS. Felice e Fortunate, and must there-
fore be of earlier date, and form one of the stages
traversed by the Lombard builders in their pro-
gress towards the vaulted Lombardic basilica.

Fig. 245. Biella. Baptistery. Plan of
the two stories and turret (Xth and
Xlth Centuries).

THE BAPTISTERY OF THE CATHEDRAL OF
BIELLA is in two stories (Fig. 245). The semi-
circular recesses of the lower one have domical
vaults ; from the upper the cupola starts. It rests

on rudely formed pendentives, and from its crown rises a turret pierced with two-light

openings (Fig. 246). This turret has

been thought to be a later addition. 1 I

lately tested the masonry, and found

that this was the case. It is possible

that when the primitive church of

Santo Stefano was rebuilt in the

first part of the Xlth century (of

this one campanile remains) and

its bell-tower destroyed, this turret

was erected to serve as a temporary

belfry for the church, the bells being

rung through holes pierced in the

crown of the cupola. This quasi-

lantern of masonry foreshadows the

lanterns surmounting large domes in

churches of a later date. Previously,

lanterns were made of wood, as we

may see from that in the abbey church

of Saint Riquier (793-798).

The roofs rest directly on the

vaulting after the Roman fashion.

The exterior is decorated with ranges

of arched niches divided into groups

by lesenas. Over the door is a lunette.

The windows are very narrow and

splayed on both sides. Herring-bone

work is very frequent in them.

In this baptistery the pendentives of the dome are to be noticed (Fig. 247).

1 Mella, A ntico battistero della catledrale di Biella.

Fig. 246. Biella. Baptistery (Xth Century).

THE RISE OF THE LOMBARDIC STYLE

179

They are not the ordinary triangular spherical pendentives of the Romano- Ravennate
type, merging immediately in the dome above them or in the circumference of the
drum. Rather they consist of irregular spherical surfaces of triangular or quad-
rangular shape built to carry pieces of vertical wall. One of the latter, the uppermost
at each angle, by a gradual and irregular transformation, serves to form the transition
from the square base to the circle of the cupola. Pendentives of this kind are the
rude beginnings of the Campano-Lombardic compound pendentives which will be
dealt with when we come to the baptistery of Galliano (1007) ; and they enable us to
fix with sufficient approximation
the date of the building.

Mella 1 would put it as far
back as the VHIth or IXth cen-
tury, preferably the latter. De
Dartein 2 is willing to assign it
definitely to the IXth. Cattaneo 3
inclines to place it in the IXth
or even the Xth. I, in my turn,
propose the second half of the
Xth century. I am confirmed in
this view by the presence of pen-
dentives which are precursors of
the Campano-Lombardic type,
found in a complete though still
rude form at Galliano. It ap-
pears to me incredible that, be-
tween the tentative results at
Biella, and the solution of the
problem at Galliano, there can
have intervened more than a few
years.

My date is also confirmed
by the constructive idea which
underlies the baptistery. I refer
to the way in which the oblique
thrust of the pendentives and

Fig. 247. Biella. Baptistery. Pendentive of the dome
(Xlh Century).

arches against the outer walls is
met by substantial external but-
tresses, radiating as it were from the centre. This shows that the Lombard gilds had
made sure progress in the science of construction, and it foreshadows the solution
of the problems of equipoise, on which the vaulted Lombardic basilica was based.

Before leaving the baptistery of Biella we may say a few words about another
well-known baptistery, that of Novara ; and we will try to settle the question of
its date.

It consists of an octagon, with alternate rectangular and semicircular niches,
apparently (the fact cannot easily be tested) of Roman construction. The interior is
faced with arches springing from columns and capitals brought from other sources,
which carry rough cubical pulvins. The object of these arches is to support an
octagonal drum, upon which rests a dome crowned by a lantern which is a later

1 Op. fit.

Op.

Op. cit.

N 2

i8o

LOMBARDIC ARCHITECTURE

addition. The drum, which is constructed of re-used bricks, is decorated on the
outside with a rude arched corbel course broken by lesenas at the angles, and finished
off with a range of small niches arched in two orders.

On the one hand, the rudeness shown by the internal facing, by the pulvins, and
by the way in which the transition from the octagon of the drum to the circle of the
dome is managed by the adaptation of the walling ; and, on the other hand, the
relatively refined effect of the building compared with the baptistery at Agliate,
induce me to place it in the first half of the Xth century, before the erection of the
baptistery at Biella, to which it is so inferior in constructive idea.

THE CHURCH OF SS. FELICE E FORTUNATO NEAR VICENZA. Recent
operations have resulted in the disappearance of the few characteristic remains of the
building erected in 985 by the bishop Rudolphus, 1 whose tenure of the see is given by
Gams 2 as from 967 or 968 to 974. We, however, will adhere to the date 985, for
which there is documentary evidence. These remains had survived the mutilations,
transformations, and reconstructions which the church had undergone in the Xllth,
XlVth, and XVIIth centuries, and were sufficient to indicate the organic structure of
the building, and also to provide specimens of its decoration. However, Cattaneo's
description, 3 and my own notes and sketches made on the spot before the operations
referred to were begun, enable me to give a short account of both the constructive and
decorative elements.

The original building was of rectangular plan, with nave and aisles separated by
columns alternating with piers This is a feature that should be noticed, for the

scheme of supports alternately large
and small was, at a later date, adopted
in the earliest Lombardic vaulted
basilicas. The columns only sup-
ported the arches which carried the
walls of the nave, whereas the piers
provided in addition a starting point
for the transverse arches which
spanned both aisles and nave.

What the nature of the roof may
have been it is impossible to say.
But considering that San Babila at
Milan, belonging to the earliest years
of the Xlth century, possessed cross
and barrel vaults springing from piers
of uniform thickness, and that it was
only in the first quarter of the same
century that groined vaulting began
to be carried on piers alternately large and small, I do not think we shall go far
wrong if we suppose that the church of Rudolphus had a wooden roof, and that the
transverse arches were only intended to secure a solid and rational concatenation
of the different parts of the structure, and also to make the timber roof
diaphragmatic, thus simplifying its construction, and rendering less easy the spread

Fig. 248. Vicenza. SS. Felice e Fortunate,
the primitive church (985).

Capital of

1 Grande illustrazione del Loinbardo- Veneto. Cabianca e Lampertico, Vicenza e il sno tcrritorio.

2 Of. (it. :! Op. cit.

THE RISE OF THE LOMBARDIC STYLE

181

Fig. 249. Vicenza. SS. Felice e Fortunate.
Base in the primitive church (985).

of the flames in case of fire. The same thing occurs in San Miniato al Monte near
Florence (1013).

The compound piers showed a decided Lombardic character. The solitary one
which survived in a mutilated form from the alterations of the XlVth century, and
has now perished, had a continuous capital (Fig. 248) showing two rows of palmetto
leaves unexpanded. It rested on an Attic base (Fig. 249), with Lombardic profile,
composed of two rolls of the same thickness and of very nearly equal projection, with

a slightly recessed scotia between them ; and
at the angles of the plinth below the half-
columns it had strengthening spurs, rudely
shaped like leaves.

This pier was noteworthy both on account
of the profile of the base, and the design and
proportions of the capital, the precursor of so
many similar ones in the centuries succeeding
the epoch of 1000 ; and also for the charac-
teristic spurs ornamenting the base, which
afford the earliest example of this familiar
appendage of both Lombardic and Pointed
architecture. Earlier instances brought forward by some writers either are non-
existent, like those already mentioned on the bases of columns in Diocletian's
palace at Spalato, or else they are not really such. This is the case with the
spurred bases in the crypt of the cathedral at Chur, a work of the Xllth century,
or, to be more precise, of the time of Bishop Bruno (1179-1180), who consecrated
the church, and not of the VHIth, as some suppose. A glance at the nine
capitals surmounting the shafts above these bases, with their foliage, striations,
and human heads, will convince anyone of the truth of my statement. The
same is true of the similar bases in the external dome arcading of the Palatine
Chapel at Aachen (796-804), which is the result of a raising of the walls of the drum
carried out in the Xlllth century.

This feature, at first of very simple form, but later the subject of much elabora-
tion, perhaps originated in an attempt to counteract the friable nature of the sand-
stone used. 1 It may equally well be due to the passion for novelty which at that
period stirred the Lombard gilds.

The capitals of the columns were imitations of ancient types, with an infusion of
Lombardic taste. For though the only one which survived the disastrous changes in
the building was an imitation of the most elaborate Ionic pattern, still it was embel-
lished with Pre-Lombardic decorative detail, and its massive abacus exhibited the
interlacing which is to be seen on that member of so many Lombardic capitals of the
Xlth and following centuries.

This surviving capital (now in its turn destroyed) teaches us that towards the
end of the Xth century there was already an idea of abandoning the rude cubical
capitals of Corinthian type and Pre-Lombardic forms, which are so markedly
characteristic of the VHIth and I Xth centuries, and of replacing them by others
which, though more directly imitated from the old Roman models, presented
decorative details which gave them a peculiar, that is to say, Lombardic character.

Further, this capital and the continuous one belonging to the compound pier tell us
that, some years before the epoch of 1000, the representations of human beings,

1 Mella, Elementi ifi arc hitettura lombarda.

182

LOMBARDIC ARCHITECTURE

animals, and monsters, characteristic of the Lombardic style, had not yet begun to be
practised by the gilds among which they originated ; a fact which shows that in
that age the style was still in process of formation and had not yet reached perfection.

BASILICA OF SANTO STEFANO AT VERONA. The oldest and most remarkable
portion of this building is the semicircular apse. The exterior is decorated
with an arched corbel course divided by lesenas. A two-storied ambulatory
encircles it within. The lower story is supported by columns and roofed by two
barrel vaults on either side of a cross vault in the middle, which is a later
addition pointing to a date subsequent to the epoch of 1000. Of the capitals of
these columns, twelve are obviously of the same date. They are Corinthianesque,
surmounted by an abacus.

The upper story (Fig. 250) is also supported by columns, and roofed with
both barrel and intersecting vaults. Seven of the capitals which surmount the
. columns are of the same pattern as those just

described.

None of these capitals have a necking, and
some have been mutilated in order to make them
fit the shafts. They exhibit the Ravennate
manner of the VHIth century; and though
they show more incorrectness in design and
carelessness in execution than the capitals of
the same type and age in San Salvatore at
Brescia (753), San Giovanni in Fonte (about
750-760) and the original Santa Maria Matrico-
Fig. 250. Verona. Santo Stefano. Apse. i are (about 7"5o-76o) at Verona, still they must

Plan of the upper ambulatory (Xth . . ^,

Century). be regarded as contemporary with them, or

nearly so, and consequently productions of the

Vlllth century. To the same period and the same Ravennate artists are to be
ascribed the mutilated capital here illustrated (Fig. 251), which, like the others, has
been used for the upper ambulatory, and also the Corinthianesque specimens with
stiff plain leaves in the crypt, which are contemporary with the nineteen of the
same kind already mentioned.

In the course of a careful examination of the apse ambulatories and the
staircases leading to them in Santo Stefano, I became convinced that the
former were broken off when the nave was rebuilt at a later period, and that
originally they were a continuation of the aisles.

And what are we to say of the date of this apse? Its limited depth,
equal to its radius, following the type derived from the Roman tradition, might
justify us in putting it, as some would do, before the IXth century. But this
reason is superficial and deceptive. The history of the church to which it belongs
does not throw any more light on the question, for all we know about this
first cathedral of Verona, as it is generally believed to have been, is that it was
destroyed by Theodoric (493-526), and that the last bishop buried in the rebuilt
church was Biagius (about 744-750), after whom the episcopal throne was moved
by Hanno to the new cathedral (about 75O-76O). 1 Nor is the problem elucid-
ated by the artistic details presented by the capitals used in the apse, for they are
not contemporary with it.

1 Biancolini, op. cit.

THE RISE OF THE LOMBARDIC STYLE

Fig. 251. Verona. Santo Stcfano. Apse. Capital
in the upper ambulatory (Xth Century).

We are, then, reduced to look for the required date in a decorative archi-
tectural detail, viz. the external lesenas which, being of a substantial character,
point to the second half of the Xth century.
An even safer guide is the vaulting of the
upper story. A careful examination shows
that it is of the same kind as that in the
gallery of the baptistery at Galliano, erected
in the early years of the Xlth century. But
it is ruder in character, so that we shall not
go very far wrong if we regard it as be-
longing to nearly the same period, and
constructed at the end of the Xth century.

The date of the apse of Santo Stefano,
and the feature of the vaulted ambulatories
encircling its interior and corresponding with
the ends of the aisles, give it a place of no
small importance in the history of mediaeval
architecture. Indeed, though semicircular

aisles or ambulatories, continuing the side aisles, were to be seen in Italy in St. John
Lateran at Rome and the cathedral of Ivrea (973-1001 or 1002), two-storied semi-
circular ambulatories, forming a continuation of the aisles and the galleries above
them, did not make their appearance before the Xlth century, and Santo Stefano
is the earliest specimen of such an arrangement.

And though there had been examples of galleries in other Christian
basilicas of the Latin type in Italy previous to this, e.g. San Salvatore at Spoleto
(IVth century), and at Rome San Lorenzo in Agro Verano (579-590), the Santi
Quattro Coronati (625-638), and Sant' Agnese outside the walls (625-638), it is not
clear that these galleries were vaulted. This was not even the case with that in Sant'
Agnese which, contrary to one opinion, had only a wooden ceiling before the XVth
century, when it was replaced by the existing vaulting. So that Santo Stefano
may, with strong probability, boast of being the first Latin Christian church, not
fitted up in a vaulted Roman basilica but built from the beginning as such, to
possess vaulting for both aisles and galleries.

THE CATHEDRAL OF IVREA was rebuilt by Bishop Veremundus, as is stated
in the inscription still preserved in the ambulatory: "f CON-

DIDIT HOC DOMINO PRAESVL WAR-

MVNDVS AB IMO"; and also in the

local Breviary (Breviarium Proprimn

Eporediense), where we read that St.

Veremundus " vetustam aedem Deiparae

sacram novis operibus auxit." Although

we do not know precisely the year in

which he became bishop of Ivrea, as
Fig. 252. Ivrea. Duomo. the date of the death of his predecessor

Sketch plan of the apse T^U j .

(973-1001 or 1002). Eldradus is not ascertained, we may
believe, on the strength of a pastoral

addressed to his people on the occasion of the struggle with F 'g- 253- Ivr ea- Duomo.
, , . . Capital in the apse (973-

the Marquis Ardoin, and still preserved among the MSS. in IQOI or 1002).

fpl.

'!iiiiin!!i !l

LOMBARDIC ARCHITECTURE

Fig. 254. Rome. St John Lateran.
Plan of old apse (844-845).

the Capitular Archives at Ivrea, that his election

took place about 973, in the reign of the Emperor

Otto I, whose commissary and chamberlain he

was in 962. His episcopate is believed to have

lasted from 973 to 1001 or IOO2. 1

Repeated alterations have left nothing surviving

from the primitive building except the semicircular

ambulatory surrounding the choir (Fig. 252), with

the crypt below and the two towers ; and even this

is not in its original condition. The ambulatory has a barrel vault, and its arches,

now walled up, formerly opened round the sanctuary. Among the antique capitals

from which the arches spring is one of cubical shape (Fig. 253), designed and

executed for the build-
ing. It forms a con-
tinuation of the aisles
of the church, which
originally, probably,
also had barrel vault-
ing. This is the oldest
existing dated instance
in Italy of an ambula-
tory corresponding to
the aisles, with the ex-
ception of the one be-
longing to the Lateran
Basilica at Rome (Fig.
254), enlarged by Pope
Sergius II (844-845),
of which I made a
study before it was de-
stroyed to make room
for the new choir. This
arrangement reminds
us of the service am-
bulatory, covered with
a continuous barrel
vault, encircling the
Imperial tribune or
exedra added by Sep-
timius Severus, between
the years 195 and 2O3, 2
to the so-called Sta-
dium of Domitian on
K 'K- ass- Ivrca. Duomo (973-1001 or 1002). the Palatine at Rome.

Outside Italy an early

example is afforded by the crypt of St. Wipertus, near Quedlinburg (936). Some

think that the church of St. Martin at Tours, erected by the bishop Pcrpetuus

1 Savio, Cli antichi vescovi <f Italia dalle origini a! 1300.
* Lanciani, The Kuins and Excavations of Ancient Koine,

THE RISE OF THE LOMBARDIC STYLE

' (460-490^ provided a still earlier one; but, as we shall see at the right time and
place, the fact is very doubtfaL

Beneath the choir is the original crypt, with the plan of a miniature ^"fr^,
ending in a semicircular ambulatory like the one above. This crypt b covered with
cross vaulting in which the arches
are flush with the rest of the
masonry, and are supported by
columns and piers surmounted
by cubical capitals made for the
positions which they occupy.
Not a fen- of these capitals ob-
viously belong to the same set as
the one which we noticed in the
ambulatory above ground, while
others exhibit cubes decorated in
various rude ways.

The original crypt was en-
larged at a later date by extend-
ing it to the space below the
chancel. Or that part of it may
have been rebuilt, for its groined
vaults with visible arches are built
on a different system from that
of the older part, and the capi-
tals on which these arches rest
display a more advanced stage of
art

On either side of the crypt,
and corresponding to the towers,
open two chapels with cross vault-
ing, apparently (so far as one can
judge from the one on the right,
which is still accessible, while the
door of the one to the left is
blocked up), of the same date as
the original crypt, as is shown by
the masonry. So that we may
infer that the church of Vere-

mundus was planned with its bell-towers set at the ends of the aisles and rising over
the choir ambulatory.

These imposing towers (Fig. 255) may claim to be older than the interesting
campanile, of rather more elegant appearance, belonging to the destroyed abbey of
Santo Stefano at Ivrea (Fig. 256), which was erected by Bishop Henry II (1029-1044)
before the year IO42. 1 And they are in the same relation to all the other Lombardic
towers in the adjoining Val d'Aosta, and also, perhaps, in the whole of Piedmont,
where the existing tower of the destroyed abbey church of Fruttuaria (,Figs. 257, 358,
2 S9) begun in 1003 and consecrated in 1006,* and that belonging to the ancient

Fig. 256, Ivrea.

Campanile of (he .il>!vy church of Santo
Stefano (1029-1042).

Savio, o/. tit.

i86

LOMBARDIC ARCHITECTURE

Fig. 259. San Benigno.
Campanile of the abbey
church of Fruttuaria.
Pulvin (1003-1006).

abbey of San Giusto at Susa, founded in 1028 or
1029 * (Figs. 260, 261), afford striking examples of
such erections. An exception, perhaps, must be
made in the case of the cam-
panile of the church at Sant'
Ambrogio, at the foot of the
Sanctuary of San Michele
alia Chiusa, to be ascribed to
the monk John of Pavia,
who became archbishop of
Ravenna (983-998), and after-
wards retired to the Monte
Pircheriano. His epitaph has
actually come to light in the
church at Sant' Ambrogio. 2

These towers, further, contain the oldest dated
specimens of those characteristic corbel pulvins, made
rather elongated to correspond to the depth of the
wall, and often cut
flat at the sides,
which, from the
second half of the
Xth century on-
wards, were se-
lected by the
Lombard gilds
for their bell-
towers instead of
the Ravennate

corbel pulvins with ordinary ovolo profile, found as
far back as the VHIth century in the guard house

of Theodoric's palace at

Ravenna.

The Lombard gilds

must be regarded as having

been the first to use these

crutch-shaped pulvins, as

they were not introduced

north of the Alps before

the Xlth century, and did

not appear in the Greek

world till after the epoch

of 1000, when they were

employed, for the first

time as I believe, in the

compound windows of the

1 Savio, op. cif.

2 Esposizione italiana, 1898

Arte sacra. Taramelli, La Sagra Fig. 260. Susa. Campanile of the abbey

church of San Giusto (1028 or 1092).

Fig. 257. San Benigno. Campanile of
the abbey church of Fruttuaria
(1003-1006).

Fig. 258. San Benigno.
Campanile of the abbey
church of Fruttuaria.
Pulvin (1003-1006).

di San Michele alia Chiusa,

THE RISE OF THE LOMBARDIC STYLE

187

Fig. 261. Susa. Campanile of
the abbey church of San Giusto.
Pulvin (1028 or 1029).

ancient baptistery, now church, of the Holy Apostles at Athens (Fig. 262). That

baptistery appears to me to have been built in the early years of the Xlth century, on

account of the rudeness of the supports in the three-light
and two-light windows, and the external wall decoration
consisting of fragments of brick arranged in different
ways so as to form various patterns, interspersed with
tufa blocks in the horizontal courses of the masonry, as
compared with the more advanced art of the window
supports and brick bands with lions, scroll work, leaves,
and other ornaments in compartments to be seen in the
churches of St. Nicodemus (about 1044) and St. Theodore
(1049), also at Athens. Later, I believe, it was altered so
as to form a church, by the addition of a nave to the west
and the present cupola. The pulvins used by the Greeks

before the epoch of 1000 in multiple

windows, were of the Ravennate type,

as may be inferred from St. Mary

Panachrantos at Constantinople, sup-
posing that in the northern small

basilica the apse windows go back to

the work of Lips (886-912).

Lastly, these towers are no less

interesting than the ambulatory of

the apse from the fact that, though

altered by various restorations, and

partly concealed by modern buildings,
they still present
the oldest known
example of bell-

. ... Fie. 262 Athens. Church of the Holy Apostles

towers set at the (XIth Century),

head of the aisles.

The only exception will be if, as we shall see when we come to
discuss the cathedral of Cologne, it can be shown that the ex-
periment had been previously made in Germany. The arrange-
ment must have been derived from
the spiral staircases taken out of the
angles on either side of the principal
apse in various Roman structures
(Figs. 263, 264). The architect of
Sant' Abondio at Como (1013-1095)
borrowed it soon after, without having
to go to the church of Saint Germain
des Pre"s at Paris (Fig. 265) for the
idea, as has been imagined. We may
be the more certain of this because the
of Saint Germain, as rebuilt by Abbot

Fig. 263. 1'lan of a
Roman building.
(From Braman-
tintfs sketches in the
Ambrosiana.)

abbey church

Morardus (f 1004), possessed only one bell-tower, probably

the one on the north side, near the monks' dormitory l ;

1 Bouillart, Histoire de tabbaye royale de Saint Germain des Prey.

c
c

' -*^"> *^_ ,' '._

Fig. 264. Plan of a Roman
building, (from Braman-
tino's sketches in the Am-
brosiana. )

i88

LOMBARDIC ARCHITECTURE

and this will be the tower mentioned by Hugh of Fleury : " turrem quoque cum
signo . . . construxit." 1 In any case, the abbey church would have to yield
precedence to the cathedral of Ivrea, being some years its junior.

Fig. 265. Paris. Abbey church of Saint Germain des Pres. (From Bouillai-l, " Histoire
iff Fabbaye royale de Saint Germain des Prez.")

THE CATHEDRAL OF AOSTA. Aosta, which in the Xth century, in consequence
of the Saracen raids, contained neither houses nor inhabitants, towards the end of the
century began to rise again from its ruins and recover its population, under the pro-
tection of various powerful families, the most important of which was that of
Challand. To this period, or more probably to the first years of the epoch about
1000, and after the cathedral of Ivrea was finished, or at least was on the way to
completion, we may assign the founding, or rather rebuilding, of the cathedral of

Aosta. Perhaps it took place in the years which followed
the settlement of the Benedictines of Fruttuaria in the Val
d'Aosta, which is believed to have happened in loig, 2 and
in San Giusto at Susa (1029-1042).

The nave was rebuilt when the church was vaulted.
The choir, though it has signs of a respectable antiquity, with
cubical capitals like those in Sant' Abondio at Como, has
still suffered from alterations ; nor can these capitals be used
as a safe guide to fix the date of the building of the choir,
inasmuch as similar ones may be seen, in Upper Italy, in
buildings up to the XVth century, e.g. in the Palazzo

Madama at Turin ' where thg y occur in the work carried

Century). out by order of Ludovico d'Acaja in 1416.

The crypt has obviously been altered. The furthest

1 Delisle, Recueil des historiens des Gauies et de la France Ex fragmento chron. fratris Hugonis
Floricuemis monachi. * Tibaldi, Storia della valle d'Aosta.

THE RISE OF THE LOMBARDIC STYLE

189

part, however, viz. that immediately under the apse, appears to be original. It
has groined vaults with visible arches which spring from miscellaneous Roman
capitals, with the exception of one which has been made for its place. This is a
cube, bevelled at the angles, and supported by an octagonal shaft. From the work-
manship one might say that it was by the same hand that carved the oldest
examples in the cathedral of Ivrea. The other part, that under the spacious
chancel, has capitals made expressly, as
well as others of miscellaneous origin.
Of the former, one recalls those in the
crypt (Fig. 266) of the cathedral at
Aquileia. It exhibits a range of minia-
ture arches supported by small, fluted
pillars, above which rise stiff leaves and
cauliculi. The design, execution, and
material of these capitals show them to
be of the same date as some in the
collegiate church of Sant' Orso at Aosta
(1133), and this comparison enables us
to fix the period of the alterations re-
ferred to above. The crypt of Sant'
Orso is far from being of the Carolingian
date which some have given it. 1 Bold
cross vaulting of this kind was never
seen in that period, and belongs to the
Xllth century.

The towers (Fig. 267) were built at
the same time as the church. The
northern one was rebuilt or finished in
the XVI Ith century, while the one on
the south side still retains the original
structure in its lower part, where we
find repeated the decorative motive of
arched corbel courses grouped in twos by
lesenas, which occurs in the lowest stage
of the towers of the cathedral at Ivrea.

Their position is interesting, as they
afford the oldest known example of bell-towers flanking the aisles at the east end.
And lastly, the date of their erection gives them a claim to precedence over the Lom-
bardic campaniles in which the Val d'Aosta is so rich. The oldest of these seems to
be that at Prd Saint Didier, which is mentioned as early as the XI Ith century.

Fig. 267 Aosla. Duomo. Campaniles (XIlli,
XlVth, and XVIIth Centuries).

IOO7

GALLIANO. PARISH CHURCH OF SAN VINCENZO. It was dedicated in
by Aribertof Intimiano, afterwards archbishop of Milan (ioi8-iO4S). 23

It consists of a nave and aisles, and ends in a round apse considerably elevated
owing to the crypt beneath. It has wooden roofs, and is lighted by wide round-
headed windows without splays, and loops splayed on both sides. The apse is
decorated externally with blank arcading.

1 Tibaldi, op. cit. 2 ruricelli, of. cit.

3 Annoni, Monumenti efalti politici e religiasi del bjrgo di Canlurio e sua fieve.

190

LOMBARDIC ARCHITECTURE

The crypt (Fig. 268) has groined vaulting with visible arches springing from four
marble columns supporting Corinthianesque capitals which, though not carved with
any elegance, are nevertheless designed with sure outlines, and show an undoubted
vigour of execution. They recall, though belonging to a different "order," the
Composite capitals in the crypt of San Miniato al Monte near Florence (1013), but
are superior in boldness of execution.

Both church and crypt were originally embellished with paintings, of which only
scanty but valuable remains are left.

The parish church of Galliano is interesting to us on account of the external

Fig. 268. Galliano. Crypt of San Vincenzo (1007).

decoration of the apse. There is no earlier example of a Pre-Lombardic church
with the outer face of the apse ornamented by blank arcading extending to the whole
height of the wall. Another point is that it provides one of the earliest dated
instances of a considerably elevated crypt.

THE BAPTISTERY OF GALLIANO. The first erection, or at least complete
rebuilding, goes back to the reconstruction of the adjoining parish church of San
Vincenzo by Aribert of Intimiano, that is to say to the years immediately following
the epoch of 1000, and before 1007. This is indicated by the painting formerly
existing in the apse of the church, and now transferred to the Ambrosiana at Milan,
in which Aribert is represented in the act of offering the church with its tower which
he had rebuilt and decorated, together with the adjoining baptistery of which the porch
is shown.

The plan of the interior is a square with four hemicycles projecting from it. At
the angles of the square are four isolated octagonal pillars from which spring the
arches which carry the gallery above and the cupola. On the outside (Fig. 269), the

THE RISE OF THE LOMBARD1C STYLE

191

recess containing the altar, and the one opposite in the front of the building, are
decorated with arched corbel courses broken by lesenas.

In the thickness of the wall of the western hemicycle is constructed a narthex,
beyond which is a square porch. From the latter two staircases ascend to the gallery,
which extends from the square of the central tower to the outer wall of the building.

No wood is used in the structure, which is lighted by wide round-headed
unsplayed windows, or else by loops with splays on one or both faces. The stone roof

Fig. 269. Galliano. Baptistery (1007).

rests directly, in Roman fashion, on the vaulting, which consists of half-domes and
rudely formed barrel and intersecting vaults.

The gallery receives its light from the inside through four pairs of openings.
Just above the point where it stops, the crossing passes into the octagon from which
the cupola starts, by means of four small conical or hood-shaped pendent! ves developed
at the angles of the square where the walls are strongest (Fig. 270). Their function
is to carry four of the sides of the drum (octagonal both externally and internally),
thus generating the Lombardic cupola.

With reference to these pendentives, we may remark that the earliest traces of

192

LOMBARDIC ARCHITECTURE

conical angle raccords date from the age of Hadrian. Thus, in the Great Baths of
Hadrian's Villa near Tivoli, built between 125 and 135,! the circular structure

described 2 as the
Calidarium is but-
tressed off from the
Tepidarium by a
small, undeveloped,
hood-shaped raccord
(Fig. 271). There may
also be noticed on the
outside of the " Piazza.
d'Oro," a few steps
from its octagonal
vestibule, the remains
of an undeveloped
compound pendentive
(Fig. 272), which I
can remember in
better condition than
at present ; and this
may have provided
formed their conical

came,

to support a cupola is to be

Fig. 270. Galliano. Baptistery (1007).

the model after which the Lombards, when the time
compound pendentive.

The oldest specimen of a hood-shaped niche used
found in San Giovanni in Fonte
or the Great Font in the cathe-
dral at Naples (Fig. 273). It was
built by Soter, 3 who, according to
Gams, 4 was bishop for twenty-one
years, after 465, in order to serve
the ancient primatial church of
Constantine, Santa Restituta, as its
position shows. 5 This baptistery
was followed by the one built by
Bishop Vincentius (5 54-577) 6 for
the new cathedral (known as
Stephania, from Stephanus who
was bishop for fifteen years, after
500), which was called the Lesser
Font.

The plan of Soter's baptistery
is a square, each side of which
measures about 25 ft. This square,
crowned by a heavy cornice, is con-

1 Lanciani, Reina, e Barbieri, La villa Fig. 271. Tivoli. Villa of Hadrian. Great Baths. Hood-
Adriana. shaped raccord (125-135).

2 Gusman, La villa im finale de Tibur.

3 Man. Germ. Hist. Gesta episcoporum neapolitanornm Appendix Cala/ogits episcoporum neapolitanorum.

4 Of. cit. 5 Bertaux, Vart dans I'ltalie nu'ridionale.
6 Man. Germ. Hist. Ges!a episcoporum neapolitanornin.

THE RISE OF THE LOMBARDIC STYLE

193

verted into the octagon from which the
dome springs by means of four hood-
shaped pendentives which span the angle
spaces and carry four of the vertical walls
of the low drum. I remark here that the
angle niches supporting pieces of wall are
a very old idea in Italy. For instance, in
Hadrian's Villa, in the basement walls of
the Imperial residence on the side of the
valley of Tempe, the angle of the outer
face is blunted, and the cross-piece which
replaces it is carried on a vaulted recess
formed between the two wall faces below
(Fig. 274).

The East cannot show any dated
example of a conical pendentive earlier
than the baptistery of Soter. The most
we can say is that they used a kind of
small arches or niches, each forming the
base of a section, and thus (multiplying
the sides of the polygon from which they
start. This device may be seen, for in-
stance, in the four-faced arch at Lattakia,
assigned by De Vogue 1 to the Illrd cen-
tury (Fig. 275).

It is true we are told that from very
early times the Persians were acquainted

Fig. 273. Naples. San Giovanni in Fonte. Pendentive
(Vth Century).

VOL. I

Fig. 272. Villa of Hadrian. " I'iazza d'Oro."
Undeveloped compound pendenlive (125-135).

with the use of squinches for forming
the transition between a square base
and a dome. Dieulafoy - asserts that the
specimens from the palaces or castles of
Firuz-Abad (Fig. 276) and Sarvistan
belong, the former to the reign of
Xerxes I (486-465 B.C.) or of Arta-
xerxes I (465-425 B.C.), and the latter
to the period of the last Achaemenidae,
or possibly that of the Seleucidae. And
he insists that the conical pendentive
was the origin of the triangular one.

Perrot and Chipiez 3 have shown
with plausible reasons the incorrectness
of such dates, and place these buildings
in the 1st century of the Christian era,

1 Syrie Centrale Architecture civile el rcligieuse
du 1" au VII* siecle.

2 L'arl antique dans la Perse.

3 Histoirc de Fart dans faatiquiti Perse.

194

LOMBARDIC ARCHITECTURE

or even in the first Sassanid
period. It will be enough to
remark :

(1) That the triangular pen-
dentive, the origin of which we
traced in our account of the
mausoleum of Galla Placidia at
Ravenna, is earlier than the inven-
tion of the pendentive formed by
a semicircular recess to be seen
in San Vitale, and also than the
hood-shaped pendentive of Soter's
baptistery at Naples.

(2) The provinces of the old
Persian kingdom contain no
palaces earlier than trie " Domus
Augustana " on the Palatine at
Rome, with a ground plan like
that of the palace of Firuz-Abad,
where the disposition of the three
domed chambers in the front re-.
calls, with the variations one
would expect, that of the three
rooms of the interior western
faade of the Roman palace, and
indicates an influence derived from
the Roman conquests in Asia.

(3) Vaults of wide span, con-
structed of broken stones or lumps of tufa set in mortar (a method of construction
which some think was originated by the Etruscans 1 ), of which the dates are
authenticated, were not attempted by any peoples

before the Romans, 2 and are not to be found
anywhere outside Italy before the days when
Rome imposed on her Empire, far and wide, the
architecture of the arch and the vault. Now the
larger domes at Firuz-Abad cover in each case
a square with sides measuring about 43 ft., while
the crown is about 72 ft. above the floor, and they
are constructed of broken stones roughly shaped,
set in mortar.

(4) The raccords at Sarvistan and Firuz-
Abad, intended to form a direct transition from
the polygon to the circle, have no connection with
any similar experiment made in the East before
the Vth and Vlth centuries. Such raccords are
an imperfect reminiscence or rude application of
pendentives which would come from a combina-

Fig. 274. Tivoli.

Villa of Hadrian.
(I25-I35)-

Palace. Angle raccord

1 Isabella, op. fit.

2 Choisy, L'art de batir c/iez Ies\Roinai>is.

Fig. 275. Lattakia. Angle raccord in
four-faced arch (Illrd Century).

THE RISE OF THE LOMBARDIC STYLE

'95

tion of those in Soter's baptistery at Naples with the others in San Vitale at
Ravenna.

(5) Finally, seeing that it was under the Sassanidae (226-651) that Persia was
most receptive of Western influence in matters of building, so much so that Perrot
and Chipiez l describe Sassanid art as recalling in many of its aspects the art of
Rome under the Antonines and the dynasty of Severus, the erection of the palaces
of Sarvistan and Firuz-Abad may be placed, at latest, in the last years of that period,
and after the completion of the baptistery of Soter and of San Vitale ; possibly
when the Western builders had already erected the palace of Chosroes I (531-579)
at Ctesiphon, which
presents the striking
analogies with the
one at Firuz-Abad
pointed out by
Dieulafoy. They
had previously been
assigned to the Sas-
sanid age by Flandin
and Coste, 2 while
Lenoir 3 takes Firuz-
Abad back to the
time of Firuz (458-
484), who must have
given the place its
name. Fergusson 4
gives the precise
date of Sarvistan as
about 350, and of
Firuz-Abad as about
450 ; but on what
evidence he does not
tell us.

That the conical
pendentive of San
Giovanni in Fonte
was, after some cen-
turies of disuse, given
a new lease of life,
was due to the

Lombard gilds. The first efforts of the new birth are to be seen in the baptistery at
Biella: we find its full accomplishment at Galliano.

There is good ground for believing that this kind of pendentive was first applied
on a large scale in the cathedral of Parma, reconstructed about io6o, 5 after the fires of
1038 and 1055, by Bishop Cadalus (1046-1071), who became the Anti-pope Honorius II
(1061-1064), and consecrated in 1106. Later it was restored in consequence of the

1 Op. fit. 3 Voyage en feist. 3 Op. fit.

4 History oj architecture in all countries, from the earliest times to the present day.

5 Allocli, Serfe cronologica dei vescovi di Parma.

6 Muratori, Kenim ital. script. Chronicon Parmense.

O 2

Fig. 276. Firuz-Abad.

Angle raccord in palace.
antique de la I'erse")

(From Dieulafov, " Larl

196

LOMBARDIC ARCHITECTURE

Fig. 277. Constantinople. St. Saviour Pantepoptes (Xlth or Xlllh Century).

earthquake of 1117
which brought
down a consider-
able portion of the
church. The sim-
plicity of the mas-
sive restored piers of
the present dome,
compared with
those of the nave,
and also the pre-
sence of single in-
stead of compound
Lombardic penden-
tives, lead me to
think that the dome
is original up to the
low drum.

The new type of cupola, which I and others call " Lombardic," from the part
of Italy where it originated, was brought to perfection at a later date by being carried
on hood-shaped compound peridentives. The earliest example on a large scale is
still to be seen in San
Michele Maggiore at
Pavia (Xllth cen-
tury) ; but there can
be no question of its
employment in San
Lorenzo at Milan after
the catastrophe of
1 1 24, as is proved by
the evidence of Bassi. 1
The Lombardic dome
was elaborately deco-
rated on its exterior,
and in this form it
crossed the Alps.

There is no cer-
tainly dated example
of a Lombardic cupola
earlier than the bap-
tistery of Galliano.
Nor can the Byzan-
tine world show one.
There, from the time
of Justinian I (527-
565)onwards,only the
following kinds of cen-
tral dome were in use.

Op. dt.

Fig. 278. Salonica. St. Panteleemon (Xlth or Xllth Century).

THE RISE OF THE LOMBARDIC STYLE

197

Fig. 279. Athens. St. Niccxlemus (about 1044).

(1) Polygonal externally, and internally composed of sections alternately flat
and concave, as in SS. Sergius and Bacchus (about 527), and in two of the cupolas of
St. Saviour Pantocrator

(in 8-1 143), at Constantin-
ople.

(2) Completely cir-
cular in form, or else
polygonal externally and
internally constructed with
visible ribs, as in Isidore
the Younger's dome (551-
563) for St. Sophia, St. Mary
Diaconissa (582-602), St.
Saviour Pantepoptes (Fig.
277) erected by Anna
Dalassena, mother of
Alexius I Comnenus
(1081-11 18), 1 and in one of
the cupolas of St. Saviour
in the Chora, at Constan-
tinople. The last church
was rebuilt by Mary

Ducaina, mother-in-law of Alexius I Comnenus, and was partly reconstructed by
Theodore Metochita under Andronicus II Palaeologus (1282-1328).

(3) Circular both externally and internally,
the interior being treated as a continuous
spherical surface. This device has been fol-
lowed in St. Irene (about 740) and St. Mary
Panachrantos (886-912, 1282-1328), also at Con-
stantinople, and in St. Sophia at Salonica (about

495)-

(4) Circular or polygonal on the exterior,
while the whole of the interior surface is divided
into concave sections, as in the convent church
of Myrelaion built by Romanus Lecapenus (919-
945) at Constantinople, or in the manner to be
seen at Salonica in the church of the Holy
Apostles (Xlth century).

(5) With the interior surface divided into con-
cave sections continued through the drum below,
which is thus, as it were, fluted. St. Mary
Pammacaristos (1081-1 1 18) and St. Saviour in the
Chora at Constantinople are instances.

(6) Polygonal externally, with the interior
treated as a continuous spherical surface, as in
St. Theodore at Athens (1049), and in the

churches of the Virgin (1028), St. Elias (Xlth century), and St. Panteleemon (Xlth

or Xllth century) (Fig. 278) at Salonica.

1 Du Cange, Hist. Byz. Conslanlinopolis Christiana.

Fig. 280. Athens. St. Nicodemus
(about 1044).

198

LOMBARDIC ARCHITECTURE

All the preceding cupolas are supported on triangular spherical pendentives, with
the exception of SS. Sergius and Bacchus at Constantinople.

(7) The last kind of Byzantine dome rests on an alternation of arches and angle
recesses spanned by hood-shaped vaults, with or without triangular pendentives
above them, as may be seen in St. Nicodemus at Athens (about 1044) (Figs. 279,
280), and in the convent church of Daphni near Eleusis (Figs. 281, 282), which is
believed l to have been built before the end of the Xlth century to replace the original

_ ^^^ church of the Vth or

Vlth.

At this point we
may make a further
observation. Some
have suggested that
it is on the eastern
side of the Adriatic
that we must look for
the earliest examples
of the " Campanian "
pendentive from
which the Lombardic
cupola originated, i.e.
the hood-shaped pen-
dentive carrying a
vertical piece of wall,
which the Lombard
gilds made compound
and applied to octag-
onal domes, thus pro-
ducing the" Campano-
Lombardic " penden-
tive. Dalmatia, as a
matter of fact, con-

tains buiidi "g s c m -

monly attributed to a
far earlier period than
the baptistery of
Galliano, in which use is made of conchiform squinches. I mention here the small
churches of San Pietro Vecchio and San Lorenzo at Zara, San Nicol6 and Santa
Croce at Nona, and Santa Barbara at Trau. But these buildings are not as ancient as
they are supposed to be, and are later in date than the baptistery of Galliano. Thus,
for instance, the church of San Lorenzo at Zara, thought to belong to the VHIth or
IXth or, at latest, Xth century, is really a work of the Xlth. I am supported in this
view by the consideration that of the four capitals belonging to the columns which
separate the nave from the aisles, the two made expressly for the church (the others
being ancient and brought from elsewhere), viz. those of Corinthian type with, in
one case, the figure of a saint carved on it, exhibit precisely the Lombardic manner of
the Xlth century. On the other hand, another church of the same type, Santa
Barbara at Trau, is to be ascribed to the following century, and after 1123, in

1 Millet, Le monastere de Daphni.

Fig. 281. Daphni. Convent church (Xlth Century).

THE RISE OF THE LOMBARDIC STYLE

199

consequence of the destruction of the city by the Saracens : " Urbs direpta est,
moenia funditus excisa, publica et privata aedificia eversa." l

To return to the baptistery of Galliano, everything in it is irregular, plan as well
as elevation. The square of the interior is almost childishly incorrect ; the four apses
forming a Greek cross are not exactly opposite to one another ; while instead of
external buttresses corresponding to the arches of the interior, the thickness of the
walls was increased at the points from which the apses start. There are the most
patent inequalities in the thickness of the walls and vaulting, the span of the arches,
the diameter and height of the piers, the distribution, shape, and lighting capacity of
the windows, &c. On similar grounds, De Dartein, 2 though he calls attention to its
very interesting artistic features, considers it as
a work of great imperfection, the production of
unskilled builders, mere country labourers.

My view, on the other hand, is that, in spite
of the numerous defects referred to, and although
there is to be noticed a complete absence of
carved ornament, the structure must be regarded
as of no small value for the history of architec-
ture. The whole has no parallel in any building
either in Italy or beyond the Alps, not even
Theodulfs church at Germigny des Prds (801-
806). Its very complex forms, and the difficulties
involved in the attempt to vault every space,
suggest that its builders were not regular " ma-
gistri," while the person who designed it was a
capable master mason who, apparently, wanted
to experiment on a small scale (though with un-
suitable materials and workmen of little skill)
with a plan which was later to be carried out on
a larger one; and to attain this object he sacrificed
all artistic considerations.

However this may be, it tells us that, about

the epoch of the year 1000, the study of scientific vaulting had made notable progress
in Lombardy, and that the organic conception of the galleries and of the Lombardic
cupola placed at the crossing was already formed, and only waited to be applied on
a grand scale in the Italian basilicas.

Fig. 282. Daphni. Convent church
(Xlth Century).

THE CHURCH OF SAN BABILA AT MILAN. There is
record of the date of the erection of the existing building.
However, we shall find that it is not difficult to fix it
approximately.

It consists of a nave and aisles separated by com- _
pound piers (Fig. 283), from which spring the longitudinal
arches of the nave and the transverse ones which span
both nave and aisles (Fig. 284). These visible transverse
arches serve, in the aisles, to support the groined vaults,
concave at the crown in order to lessen the thrust, an
idea which came from Ravenna, where we met with it

1 Farlati, op. cit. Episcopi Tragurienses. a Op. fit.

no authentic historical

Fig. 283. Milan. San Babila.
'Section of pier (Xlth Century).

2OO

LOMBARDIC ARCHITECTURE

in San Vitale. In the nave they give strength to the barrel vaulting. Substantial
buttresses project from the corresponding points of the exterior (Fig. 285). The
transverse arches over the aisles carry arched ramping walls which bind the two sets
of buttresses together.

The continuous capitals round the piers, of heavy cubical Lombardic forms,
present interlacing, foliage, scroll work, cauliculi, the lamb bearing a cross, the dove,
a pair of griffons drinking, and two animals biting one another's feet or tails. Where

Fig. 284. Milan. San Babila (Xlth Century).

not renewed, these carvings are in shallow relief and flat. The bases of some of the
half-columns are spurred at the angles.

As the nave walls are not carried up, the nave has no windows. The aisles, on
the other hand, are lighted, on the north by loops splayed on the inside, and by large
windows on the south.

The front was rebuilt when the church was lengthened by a bay and the existing
Lombardic cupola constructed, probably replacing a simpler one, just as happened in
Sant' Ambrogio at Milan. It must have been of the same type as that of the
bapistery of Galliano (1007), and, we may add, the original one in Sant'
Ambrogio ; and did not require piers specially built to support it, seeing that

THE RISE OF THE LOMBARDIC STYLE

201

later it was possible to raise the present cupola on the same piers. The apse,
approached through a barrel-vaulted presbytery, is flanked by two cross-vaulted
chapels also with apses. On the outside it is strengthened by buttresses, and
decorated with a range of elongated niches, to which an arched corbel course ^vvas
added as a cornice in the last restoration (Fig. 286).

In San Babila there appears for the first time the application to a Lombardic
church of the ingenious plan of raising ramping walls above the transverse arches
of the aisles, pierced by a passage opening, which connect together the buttresses of

Fig. 286. Milan. San Babila (Xlth Century).

the nave and those of the aisles. The architect may have derived the suggestion
from the Tepidarium of the Baths of Diocletian and Maximian (306), and also,
perhaps, from the Thermae Herculeae at Milan, which must have been built by
Maximian before his abdication (305).

I note at this point that, in my opinion, the architect intended to cover the
nave (which is about 27 ft. wide between the piers) with cross vaulting as well as the
aisles, and with this object he carried up a vaulting pier consisting of a pilaster and
two engaged shafts to the spring of the vaulting, and also prepared massive
buttresses on the outside to receive the thrust. Afterwards, when he was face to face
with the daring project, his courage failed him, and he was content to roof his nave

2O2

LOMBARDIC ARCHITECTURE

with cylindrical vaulting divided
into sections by transverse arches.
Hence, the pilaster was the only
member of which any use was
made, the two shafts being left
without any function to perform,
and the nave was deprived of
windows.

In this church we may also
observe the first timid appearance
of animal forms in the decoration
of the capitals. We shall find them
fully developed in San Flaviano at
Montefiascone (1032).

The date of San Babila is
shown above all by the apse, which,
with its bold buttresses and the
character of its niches, exhibits a
close relationship to that of San
Celso at Milan (996). They must
be almost contemporary. In the
second place, it is revealed by the
occurrence in it of not a few of the
chief elements of the Lombardic

system, which cannot be said of SS. Felice e Fortunate near Vicenza (985), so that

San Babila must be later than that church. Thirdly, it is demonstrated by the fact

that its aisles have groined vaults, whereas in 1032 ribbed

vaulting had already crossed the Apennines to make its

appearance in San Flaviano at Montefiascone. And no

less so by the carvings (where original) of its piers, which

show a less advanced stage of art than those in San

Flaviano. In fact, there must have elapsed between them

an interval which may be safely estimated at about a

quarter of a century.

For all these reasons the date of San Babila may

be fixed, without much chance of error, within the first

decade of the Xlth century, and perhaps about the time

when the Lombardic cupola made its first appearance in

the baptistery of Galliano.

Fig. 285. Milan. San Babila. South side (Xlth Century).

THE BASILICA OF SAN MINIATO AL MONTE NEAR
FLORENCE came into existence, apparently, in the early
years of the Vth century. Bishop Hildebrand (1008-
1024) began to rebuild it in 1013: " Ildebrandus . . .
basilicam restauravit ac magnifice exornavit." 1 And it
is known that by about the year 1062 it was finished. 2

It consists of a nave and aisles (the former being

1 Ughelli, Italia sacra Archiepiscopi Florentini.

2 Supino, Gli albori del? arte Jiorentina Architcttura.

Fig. 287. San Miniato al Monte
Plan of church (1013-1062).

THE RISE OF THE LOMBARDIC STYLE

203

double the width of the latter) separated by nine round arches springing from
cruciform 'piers ',and isolated columns (Fig. 287). The latter only carry the nave
arcades, while the compound piers furnish the supports, not only for the longitudinal
arches, but also for the transverse ones which span both nave and aisles. The nave
terminates in a semicircular apse. Nave and aisles alike have open timber roofs,
made diaphragmatic by the transverse arches (Fig. 288).

The outer walls though tampered with in the upper part, are original, with the
exception of the
facade which ap-
pears to belong
to three different
periods, viz. the
Xlth, Xllth, and
XHIth centuries.
Wall piers corre-
sponding to the
half-columns of the
nave piers support
the arches which
span the aisles.
The windows are
narrow and double
splayed. The heads
of those in the apse
have voussoirs
alternately of tufa
and brick.

Beneath the
elevated presby-
tery opens the
seven-aisled crypt,
covered with
groined vaults with
visible vaulting
arches (Fig. 289).
The capitals, made
for their places,
which surmount
the isolated
columns, are in
some cases Com-
posite ; and one of these, while presenting on the one side the mere outline of the ovolo,
with the volutes only blocked out and plain stiff leaves, on the other has the leaves
fully carved and the volutes (between which runs an egg-and-dart and bead-and-reel
moulding) filled by roses executed with some care like the other ornaments (Fig. 290).
This peculiarity may be due as much to the carver's wish to leave some evidence of
his skill and of the different kinds of work that his chisel was capable of as to his
arbitrary fancy. It is possible, too, that it may come from the instinct of imitating
the antique, for similar irregularities are to be found among the buildings of the

Fig. 288. San Miniato al Monte. Church (1013-1062).

204

LOMBARDIC ARCHITECTURE

Rome of the Empire, where we sometimes see capitals with the details left uncarved
in parts which were hardly visible or quite hidden. For instance, in the Mausoleum
of Santa Costanza outside the walls of Rome (IVth century) there are capitals
(among the twelve of the outer circle, which are earlier than Constantine and
have been brought from elsewhere) with the two rows of free acanthus leaves
elaborately carved, while in others the acanthus leaves are only outlined and,
as it were, enclosed in shells, waiting for the carver's chisel to do its work and

set them free. This
peculiarity, which was
also in some cases
intended by its au-
thors to go down to
future ages as a speci-
men of the methods
of execution in vogue
in their days, is illus-
trated by examples
in various Lombardic
buildings of the
mediaeval period,
among them Sant'
Ambrogio at Milan,
where the south gal-
lery contains a con-
tinuous capital show-
ing interlacing of
which some is merely
outlined, some par-
tially executed, and
some completely
finished (Fig. 291).

O n t he other
hand, the wall piers
in the crypt have Pre-
Lombardic cubical
capitals, hollowed out
at the angles (Fig.
292), and others in
which the circum-
ference of the part
corresponding to the
section of the column passes into the square at the top, which is crowned by a high
abacus.

The basilica of San Miniato suggests some considerations from which we may
derive useful information.

First and foremost, its organic structure, in spite of the retention of the form of
the Roman basilica, shows, with its compound piers alternating with columns and the
transverse arches which bind the whole edifice together in a rational and stable
manner, an advance towards the Lombardic style, and also affords the first example

Fig. 289. San Miniato al Monte. Crypt of the church (1013-1062).

THE RISE OF THE LOMBARDIC STYLE

205

Fig. 290. San Miniato al Monte. Church.
Capital in the crypt (1013-1062).

on the western side of the Apennines of an organic conception of this kind. The

fact indicates that this conception, within a few years of its origin and under the

powerful influence of the Lombard School, was

steadily making its way in Italy, and that the

Lombardic style was still in process of forma-
tion ; for otherwise, in the case of so important

a building, it surely would have appeared in

its complete form instead of one of simple

transition from Roman to Lombardic.

In the next place, the capitals, wrought ex-
pressly to fit the isolated columns in the crypt,

no longer show the Corinthianesque Pre-Lom-

bardic forms of the VHIth and IXth centuries

which we know so well, but are imitations of the

simplest type of Roman Composite. This

feature is worth notice for several reasons.

The first is that, as it is not met with in any

other building of which the date is certain

earlier than San Miniato, it follows that it is an

undoubted result of the diffusion of that revival

of art which not long before had made its ap-
pearance in Upper Italy.

The second reason is that, as these capitals exhibit a form more closely

approximating to the classical type than that of the imitation of Ionic and

Corinthian capitals wrought for the nave of SS. Felice e Fortunate at Vicenza (985)

and the crypt of
San Vincenzo at
Galliano (1007), it
may be inferred
that, in spite of the
influence exercised
in Tuscany by the
Pre-Lombardic
style which was
dying, and the
Lombardic which
was coming to the
birth, in that dis-
trict the traditions
of Roman art had
considerably more
vitality than in
Lombardy and the
adjacent regions,
where imitation of

Roman types is almost exclusively limited to the Corinthian, though with variations

in proportion and design.

A third reason why these capitals are important is that, if it was only in the

first years of the Xlth century that the artistic revival crossed the Apennines,

Fig. 291. Milan. Sant' Ambrogio. Capital in the gallery (Xlth Century).

206

LOMBARDIC ARCHITECTURE

its diffusion through the Lombard part of Tuscany, the Duchy of Rome, and in
Rome itself, must naturally have taken place at a still later period. It is on grounds

such as these that I believe we ought to assign to a

date not earlier than the first quarter of the century

following the epoch

of 1000 the capitals

imitating ancient

types, or Lombard-

esque in character,

which the City of

Rome has to show.

It is just in that

period that I place,

e.g., the capitals of

Santa Maria in Cos-

medin, imitating the

Fig. 292. San Miniato al Monte. simplest type of Com-
Church. Capital in the crypt
(1013-1062).

- 2 93- Rome. Santa Prassede.
Chapel of St. Zeno. Capital
(Xlth Century).

posite, and the Ionic Lombardesque capitals of the
columns flanking the entrance to the chapel of St.
Zeno (Fig. 293) in Santa Prassede. Both chapel and church are the work of Pope
Paschal I (8I7-824). 1 ^^^^^^^^^^^^^^^

THE BASILICA OF SANT'
ABONDIO OUTSIDE COMO was
erected above the ruins and partly
upon the foundations of the primi-
tive church of SS. Peter and Paul
(Vth century). The existing church
(Fig. 294) is the work of Alberic,
bishop of Como (1010-1028), a fact
which is proved by a diploma of
1013 issued from Paviaby Henry 1 1
of Germany (1002-1024). In 1027
the building was, apparently, still
unfinished, for legacies in honour
of St. Abundius were left in that
year by some citizens of Milan. In
1063, again, there is record of a
handsome donation made by Bishop
Raynaldus (1061-1092) to Arderic
the Benedictine abbot of Sant'
Abondio, and this may be con-
nected with the completion of that
part of the church which was
assigned to the general public.
In 1095 the church was
solemnly consecrated by Pope

1 Duchesne, Le liher pontificalis-

Fig 294. Como. Sant' Abomlio (1013-1095).

THE RISE OF THE LOMBARDIC STYLE

207

Fig. 295. Como. Sant' Abondio.
Capital in the nave (1013-1095).

Urban II. 123 It is possible that the ceremony did not correspond exactly with

the completion of the building, which I believe was reached several years before,

and that it merely coincided with the presence at

Milan of Urban, who had been a Benedictine monk

and a disciple of Abbot Hugh of Cluny (1049-1109),

on his way to the General Council at Clermont.

The date 1013 when the rebuilding of the church

began, gives Sant' Abondio a claim to precedence

over all other churches in Italy and outside it with

respect to the use of cubical capitals resulting from

the penetration of half a sphere and a cube (Fig.

295). It has generally been described by English

writers as the Cushion Capital.

No one has as yet pointed out that the cubico-

spherical capital is derived from the purely decorative

form of it which appeared in the IVth century, also

in Lombardy, as may be seen in an important

sarcophagus of that date which was unearthed at

Lambrate in 1905, and is now preserved in the Castello Sforzesco at Milan

(Fig, 296).

The capitals of this kind, though depressed in form, which exist in the crypt of

San Marco at Venice (Fig. 297) are to be referred, not to the first church decreed by

the Doge Giustiniano Paitecipazio ( 827-829) in consequence of the bringing from

Alexandria of the body of St. Mark (828), and finished by his successor Giovanni

Partecipazio (829-837) in 836 when the consecration took place, nor even to the work

of Pietro Orseolo I (976-978), but rather
to the rebuilding of the basilica taken
in hand by Domenico Contarini who
was Doge, according to Sansovino, 4 from
1043 to 1070 or 1071 (in 1071 "aedes
D. Marci coepta est reparari in earn
formam qua nunc visitur "), and dedi-
cated in 1094 after the finding of the
body of St. Mark. 5 6 7

At Venice the artistic revival about
the year 1000 showed itself in imitations
of the antique. Evidence is to be found
in Santa Eufemia alia Giudecca, which,
as may be gathered from Gallicciolli, 8
was founded in 865 or 870 or 890, and
rebuilt or restored in 983. In it may
be noticed six Composite capitals rudely

executed after Roman models, one bell-shaped Composite with wild acanthus leaves,

Fig. 296. Lambrate. Sarcophagus now in the Castello
Sforzesco, Milan (IVth Century).

1 Tatti, De gli aiinali sacri delta citlit di Como. 2 Giulini, op. cit.

'' Boito, Architcltura del media evo in Italia La chiesa di Sant' Abondio e la basilica di solto.

4 Venetia, citto, nobilissima &f.

5 Man. Germ. Hist.JoannisDiaconichronicon Venetian et Gradcnsc.
" .Von. Germ. Hist. Annates Venetici breves.

Muratori, Kemm ital. script. Andreac Dandiili chronicum Venetian.

* Delle memoric venctc antiche profane ed ecdesiastiche.

208

LOMBARDIC ARCHITECTURE

Byzantine in character, and one of Corinthianesque type. The first six, by the way in
which they are worked, exhibit an obvious affinity to the Corinthianesque capitals

Fig. 297. Venice. San Marco. Crypt (Xlth Century).

carrying high pulvins like truncated pyramids, in the aisles of the cathedral of San
Giusto at Trieste (Fig. 298), so much so that we may believe they all came from the

Fig. 298. Trieste. Duomo (Xlh and XlVth Centuries).

same school of Venetian carvers, and belong to the same date. For the nave and
aisles of the present Duomo at Trieste, though remodelled in the XlVth century

THE RISE OF THE LOMBARDIC STYLE

209

(when the cubical capitals hollowed out at the angles, and with plain leaves
occupying the hollows, were introduced), must certainly be ascribed to this revival,
and not to the IVth or Vth and Vlth centuries. 1 I say the nave and aisles, for the
small side apse, or Bishop's Chapel, with its capitals recalling those in the gallery
of San Vitale at Ravenna (526-547), clearly belongs to the Vlth century, or, more
precisely, the time of the first bishop of Trieste, Frugiferus (about 524-568), who was
the founder of his cathedral; 2 while the cupola in front of it is obviously later than
the Xlth century.

From the same Venetian School came the eight Corinthianesque capitals with
coarse foliage sometimes treated in
Byzantine fashion, in San Giovanni
Decollate at Venice, founded in
icx>7. 3

THE CHURCH OF SANTA
MARIA AT SUSA was erected a
short way from the church of San
Giusto in the same town (1028 or
1029). It is stated 4 that it was
founded in 1027 and certainly before
1029. The date is confirmed by a
comparison of the masonry and the
decoration of the campanile with
the same features (which are original)
in the contemporary church of San
Giusto.

The primitive structure is, to a
great extent, almost unrecognisable
from the mutilation which it has
suffered, the modern constructions
built against its exterior, and the
remodelling which the interior has
undergone. All the same, the parts
which are visible make it worthy of
our attention.

The value of these remains is
concentrated in the front of the
church, which, with the surviving
tall and graceful campanile of five
stories marked off by saw-tooth
courses, and lighted by single, two-
light, and three-light openings, pro-
vides the earliest available example of a church front flanked by Lombardic bell-
towers of the same date, and forming a part of it. The source of this arrangement,
in the present case, must have been, not so much Sant' Apollinare in Classe, as St.
John Lateran at Rome, where two bell-towers rise from the northern fa5ade, i.e. the
important one which faces the city (Fig. 299). The exact date of these towers is not

Fig. 299. Rome. St. John Lateran. Campaniles
(Xllih Century).

1 Handler, // Jiwma di Trieste. * Gams, op. cit.

* Prou, Raoul Glaber, Les ciny livres de ses histoires.
VOL. I

3 Gallicciolli, op. cit.

210 LOMBARDIC ARCHITECTURE

known. But we learn that the rebuilding carried out by Sergius III (904-91 1) in the
basilica did not extend to the transept, which, like the apse, had escaped the terrible
effects of the earthquake which devastated Rome in 896, and so required nothing more
than restoration. Further, we know that Pope John XIII (975-972) had a bell of
exceptional size hung in one of them. 1 We hear, again, that in the reign of Paschal II
(1099-1118) one of them was struck by lightning, which brought down the bronze
cock on the apex of the roof, as well as the bells within, and seriously damaged the
whole angle of the wall of the church. 2 These injuries must have been fairly soon
made good, for the bronze door of the time of Celestine III (1191-1198) in the
adjoining baptistery, has engraved on it a representation of the northern fagade of
the basilica with its two bell-towers intact, showing two stages with two-light
windows and sharply pointed roofs.

A careful examination of the masonry where visible, and of the decorative features,
reveals at once two things to a practised eye. The first is, that in the course of the
restoration of the Xllth century the two highest stages of the tower were rebuilt
These stages, though frequently tampered with and more or less damaged, still
exhibit the evidence for their real age in the three-light openings with arched heads
supported by shafts of ancient origin surmounted by rude corbel pulvins flattened at
the sides and these pulvins, it must be remembered, were not seen at Rome till the
appearance of the earliest Lombardo-Roman bell-towers, i.e. till the second half
of the Xlth century. So that the date must be in the Xllth century.

The second is, that in each case the square staircase and its newel show in the
quality of the masonry, among other things, very old brickwork, which must belong to a
time when the stairs were only intended to give access to the ceiling and the roofs, and
also other work, not so old, consisting of small blocks of peperino, which (where original
and not the result of restoration) must, apparently, date from the period when the two
staircase towers were raised above the transept roof and converted into campaniles,
an occasion which maybe connected, with considerable probability, with John Xlllth's
gift of a bell to the basilica, that is to say between 965 and 972.

The scheme was afterwards reproduced by Majolus in Saint Pierre le Vieux at
Cluny, which he completed in 982.

THE CHURCH OF SAN FLAVIANO AT
MONTEFIASCONE. The date of its first
_ _____ .^^^ erection is uncertain. If we are willing to

trust the well-known Bull of Leo IVth

h-Rji (845-857) confirming to Virobono, bishop

of Toscanella, jurisdiction over all places
belonging to that diocese (the Bui, is still
preserved in the archives of Toscanella, and
is published by Campanari and Turriozzi 3 ),

da y s of that Pope, and was dedicated to the

virgin.

The present building, however, cannot
boast of so remote an origin. The oldest

1 Rasponi, De basilica et patriarchio Lateranensi.

Fig. 300. Montefiascone. San Flaviano. Inscrip- 2 Rohault de Fleury, Le Latran au moycn-Age.

tion in the faade (1032).. 3 Opp. citt.

THE RISE OF THE LOMBARDIC STYLE

211

Fig. 30 1 . Montefiascone. Plan
of San Flaviano (1032).

part, that is to say the east end and the three nearest arches in the Lombardic
style, only dates from the year 1032, as we are assured by the inscription (Fig. 300)
built into the present front of the church, which has epi-
graphical peculiarities exactly corresponding to about that
period. We give the first three lines, which state the year
when the church was rebuilt :

Annis millenis currentibus atque tricenis

binis adiunctis ostendit pagina cunctis

hoc templum factnm denuo virtutibus aptum *

The two westernmost bays and the front in the
Pointed style were not erected till the beginning of the
XlVth century, when, on the occasion of the works
referred to in the Vatican Regesta Nos. 50 and 51, the
church was restored and lengthened by about 20 ft.

Our attention must be devoted to the Xlth century

church, which has a special interest for our subject. Its plan is externally a rectangle,
while the interior takes the form of a polygon with almost every side unequal. Three
of them have apses radiating from the centre (Fig. 301), the smaller ones being taken
out of the thickness of the wall, while the larger one starts directly from the end wall
of the nave. The outer walls have been rebuilt at the top.

In the interior (Fig. 302), the central space, of irregular form from floor to roof, is

surrounded by an aisle with a
gallery above it, communicating
with one another by two stair-
cases taken out of the thickness
of the walls, only that on the left
being original. The aisle, on the
north side of which is an original
very narrow window filled with a
transenna of interlacing circles, is
formed by four massive piers with
a column midway between them
on either side, thus producing two
bays. Corresponding to these
piers are wall piers strengthened
externally by buttresses of trian-
gular shape. The bays, which
form irregular squares, have raised
cross vaulting with diagonal ribs
of rectangular section, about 10
inches across at the base (Fig.
303). These ribs of stone, like
all the others, starting from tri-
angular springers developed be-
tween the longitudinal and trans-
verse arches, are nearly semicircular, and are constructed quite independently of
the cells. They served as centring when the latter were made.

1 De Angelis, Comentario storico-critico su forigine e U viccnde delta cittA e chiesa cattedralc di Montefiascone.

P 2

Fig. 302. Montefiascone. San Flaviano (1032).

212

LOMBARDIC ARCHITECTURE

The compartments of the vault-
ing, which geometrically form parts of
a cylinder and were originally plas-
tered over, were constructed by first
placing a rough wooden centring on
the ribs ; next, by modelling up the
surface in earth or in clay and water
to receive them, and then arranging
upon this layer lumps of tufa of
various sizes, one next the other, set
in mortar, after the Roman fashion,
thus producing a kind of coating in-
tended to make the centring firmer.
Above this came the backing of rubble
concrete.

The aisle extends round the
sanctuary, forming a sort of ambula-
tory, with ribbed vaults tripartite in
the four bays at the sides, and quad-
ripartite in the middle one. In one
of the bays the ribs are not of rect-
angular section, but rudely moulded
like a torus.

The compound piers have con-
tinuous cubical capitals carved with
foliage, scroll work, cauliculi, inter-
lacing, knots, flowers, creatures of real
or fanciful character (Figs. 304, 305).
In these capitals, as well as in those
belonging to the columns (Fig. 306)
and half-columns (carved in high relief with, in places, deep shadows produced
by the use of the drill, or completely under-
cut), the foliage is in some cases treated

- 33-

-Montefiascone. San Flaviano.
aisle (1032).

Vaulting of

Fig. 306. Montefiascone. San Flaviano. Capital
(1032).

Fig. 305. Montefiascone. San Flaviano. Capital
(1032).

THE RISE OF THE LOMBARD1C STYLE

213

,1 '

Fig. 304. Montefiascone. San Haviano. Capital (1032).

l-'ig. 307. Aquileia. Pluieus in the Duomo (Xlth Century).

2i 4 LOMBARDIC ARCHITECTURE

with freedom, in others very coarsely. The animal figures, too, devoid as they are
of proportion arid anatomy, show unequal treatment, with the exception of the birds,
which are marked by correct design and careful execution.

Anyone who cares to follow my example
and to compare, not once but several times and
with the examples before the eye, the capitals of
San Flaviano with those that precede them in

Fig. 310. Montefiascone. San Flaviano.
Lion from the original portal (1032).

San Babila and those that follow them, of the
same kind, in the ambulatory of the cathedral

Fig. 308. Bari. San Nicola. Capital (1087- . _ ,, . ~. .

1098-1105). at Aversa, in Santa Maria e San Sigismondo at

Rivolta d' Adda, and in Sant' Ambrogio at Milan

(the date in every case being nowadays ascertained), will find in them, allowance
being made for differences of material, the history of the gradual and progressive
development of Lombardic carving in the Xlth century.

Fig. 309. Montefiascone. San Flaviano. Gallery (1032).

Next, if we put the carvings in San Flaviano side by side with the nearly
contemporary work on the capitals and plutei (Fig. 307) in the cathedral at Aquileia
belonging to the time of the patriarch Poppo (1017 or 1019-1042 or 1045), and also

THE RISE OF THE LOMBARDIC STYLE

215

Fig. 311. Corneto Tarquinia. Museo Bruschi. Part 01
Etruscan amphora.

with that on the original capitals (Fig. 308) in San Nicola at Ban (1087-1098-1 105),

we are struck by a general superiority of the Xlth century carvings of Lombardic

design and workmanship over the

contemporary productions of Vene-
tian or Apulian chisels in the Lom-

bardo- Venetian or Lombardo-Apu-

lian styles.

So much of the gallery (Fig.

309) as is original has on either side

an arcade with columns carrying the

lofty walls which support the modern

wooden roof with its two unbroken

slopes covering the whole building.

One of the bases has the plinth

protected at the corners by spurs

carved in the form of heads. The

capitals are hollowed out at the

angles, each of which is filled by a

coarsely carved leaf. There is one

exception, a rather singular one : a

capital cut into polygonal faces. Remains of one of the old windows, now blocked

up, may be traced on the exterior of the south side of the church.

From the original west front two valuable fragments survive, both of a

decorative character. One is a panel carved with a Siren holding up the two ends of

her tail. The other is a small
lion (Fig. 310) of very archaic
form, barbarous both in design
and execution, holding between
its paws the remains of another
animal. One side and the hind-
quarters are left rough, which
means that it must have been set
against one side of the portal and
have supported some sort of shaft
belonging to it.

There are four noticeable fea-
tures in San Flaviano to which
we will call attention. The first
holds a very important place in
the genesis and development of
the Lombardic vaulted basilica.

I. The raised cross vaulting
with visible ribs of dressed stone :
the earliest of certain date that
I can point to.

II. The Lombardic continu-
ous capitals : not a new form, for we have already noticed them in San Babila at
Milan. There, however, animal representations barely make an appearance, whereas
in San Flaviano they are fully developed, and no longer fettered by early Christian

Fig. 312. Corneto Tarquinia. Museum. Etruscan
sarcophagus.

2 I 6

LOMBARDIC ARCHITECTURE

symbolism in the manner still to be seen in San Babila, but transcend the limita-
tions which restricted representations of living nature in Western religious carving

Fig. 313. Cerveteri. Representation in an Etruscan tomb.

l''ig- 3'4- Constantinople. St. Saviour Pantocrator (1118-1143).

during the VHIth, IXth, and Xth centuries. Such representations were partly a
legacy from Roman and Etruscan art (Figs. 311,312, 313), particularly the latter,

THE RISE OF THE LOMBARDIC STYLE

217

which has not yet been studied as it deserves to be in its relations with the genesis
of mediaeval art. Partly, too, they were due to the imagination of the carvers. Their

intention, to my mind, was
rather decorative than symboli-
cal. It was not till about the
epoch of 1000 that they were
used to ornament capitals in
churches, either in Italy or out-
side it. Nor do they appear in
Byzantine churches ; for, in the
period preceding or imme-
diately following the invention
of the Lombardic capital, the
time-honoured fashions of the
East were maintained in those
countries, with rare exceptions
in which traces of Lombardic
influence are sometimes appar-
ent (Figs. 314, 315, 316,317).

And if we occasionally find
decorative motives inspired by
those of Byzantine cubical capi-
tals of the Vth and Vlth centuries, as for instance at San Flaviano in the two
capitals nearest the presbytery, still they are conceived and carried out in a way
that is quite new and original, so that they acquire a character of their own, which is
neither more nor less than the typical Lombardic character, specimens of which
cannot be found earlier than the last years of the Xth century.

III. The small lion forming part of the decoration of the original doorway, which
must have had a portal of the Lombardic type. In fact, the church of Sant' Andrea,

Fig- 3 1 5- Salonica. Church of the Holy Apostles. Capital
(Xlth Century).

Fig. 316. Athens. Church of Kapnikaraea.
Capital (914-922).

317. Salonica. Church of the Virgin.
Capital (1028).

also at Montefiascone (which I was the first to notice), and contemporary with San
Flaviano as the carving of its capitals shows, still preserves its doorway which, though
tampered with, retains its original features, being those of a Lombardic portal in its

218

LOMBARDIC ARCHITECTURE

Fig. 318. Rome. San Lorenzo in Agro Verano. Lion at
the entrance (1216-1227).

earliest stage of development, open-
ing between two orders of jamb
shafts from which springs a stout
roll decorated with carving.

It is the forerunner of the
lions, griffons, and other creatures,
and also of the crouching telamons
which, at a later period, were used
to support the monsters which
flank the doorways, and the Lom-
bardic porches of churches. Notable
examples of such "stylophorous"
creatures are to be seen in Rome
at the doors of San Lorenzo in
Agro Verano (Fig. 318) of the time
of Honorius III (1216-1227), and
of SS. Giovanni e Paolo (Xllth or
XII Ith century) 1 (Fig. 319). North
of the Apennines, even more re-
markable specimens are provided,
to give only two instances, by the

principal entrances of the cathedrals of Ancona and Parma, where they belong to

the XHIth century.

The conception of animals either stylophorous or warders ot a door was of

Eastern origin (Fig. 320). The

architects of Syria and Chal-

daea employed such, of either

realistic or fanciful character,

in quite early times. But we

must not forget that, in Italy

too, the Etruscans used to set

lions and sphinxes to guard

the entrances of tombs, or to

form acroteria ; and that the

crouching lions of the Lom-

bardic portals often copy the

pose, archaic expression, and

hook-shaped locks of the manes

of their Etruscan prototypes,

e.g. the two belonging to the

German Institute at Rome,

which came from a tomb at

Vulci ascribed to the Vlth

century (Figs. 321, 322). And

this is not all, for these Lom-

bardic creatures again suggest

Etruscan inspiration when they hold between their paws, with either a protective or

destructive intention, the forms or the remains of human beings and animals. For

1 P. Germano di S. Stanislao, La casa Ce/imontana dei SS. martiri Giovanni e Paolo.

Fig. 319. Rome. SS. Giovanni e Paolo. Lion at the entrance
(Xllth or Xlllth Century).

THE RISE OF THE LOMBARDIC STYLE

219

Fig. 320. Constantinople. Imperial Museum. Stylophorous sphinxes from Sindjrli.

Fig. 321. Rome. German Institute.
Ktruscan lion from Vulci (Vlth
Century B.C.).

Fig. 322. Rome. German In-
stitute. Etruscan lion from
Vulci (Vlth Century B.C.).

Fig. 323. Florence. Archaeological Museum.
Etruscan lion from Vulci (Vlth Century B.C. ).

Fig. 324. Colchester. Sphinx in the Museum.
(From a photograph provided for me by the Curator.)

220

LOMBARDIC ARCHITECTURE

instance, a stone lion from the necropolis of Vulci (generally ascribed to the Vlth
century B.C.), now in the Archaeological Museum at Florence, rests its right fore paw
on a human head with closed eyes (Fig. 323).

With reference to this last sepulchral representation, I would remark that it
survived in Italy up to Imperial times. For instance, there is a sepulchral altar
carved with a sphinx holding a bull's head between its fore paws. 1 It also crossed
the Alps, and appeared in Britain in the Roman period. In the Museum at
Colchester is preserved a winged sphinx of stone, with a human head between its
fore legs (Fig. 324). It is supposed to have adorned the gate of the necropolis of
Roman Camulodunum ; but it is more likely that it may have guarded the entrance
to a tomb erected during the Roman conquest for some important personage from
Etruria. As a final illustration I may call attention to a coin of Teos showing a
griffon with its left fore paw on a human head. 2

Lastly, we may notice that the Lombardic porch with columns resting on the
backs of animals, realistic or imaginary, and on telamons, was modelled on the
Roman type of projecting porch, like the one belonging to the Constantinian St.
Peter's at Rome. This is a fact which has hitherto failed to attract attention.
Against the further side of the cloistered atrium of that famous church was built
a projecting porch, exactly opposite to the central door, the " silver door " or " porta
regia maior." 3 This porch is represented in the plan of Alfarano (1590), who tells us

that it belonged to the age
of Constantine, and had a
very ancient bronze roof
supported by two porphyry
columns, which protected a
marble statue of St. Peter.*
A drawing of the facade of
St. Peter's left by Grimaldi 5
shows the porch with the
statue referred to. It may
also be seen in other views,
e.g. the one reproduced here
(Fig. 325) from a print in
the Uffizi.

If Cattaneo 6 had taken
to heart the kindly advice
of De Rossi 7 " It is not
necessary to go on pilgrim-
age with De Vogue to the
Syrian desert in order to
make new discoveries in the

history of Christian architecture" and had looked for them at home, he would have
saved himself his travels to those distant lands (though they were only on paper) in

1 Altmaim, Die rumischen Grabaltdre der Kaiserzeit Grabaltar der Cornelia Cleopatra.

2 Milani, Studi e materiali d'arclieologia e numismalica, 1904. 3 Duchesne, Le liber ponlificalis.

4 Alme urbis dim Petri veteris novi templi descriplio Archivio capitolare di -Sa Pietro al Vaticano.

5 Volumen diagrammattim veteris Vaticanae basilicae Archivio capitolare di San Pietro al Vaticano.

6 Op. cit.

' Bull, d'arch. crisliana, ifylSpicilegio d'anheologia cristiana nelt Umbria Delia basilica di San
Salvatore presso Spoleto.

Fig. 325. Rome. Facade of the Old St. Peter's. (From a print in
the Uffizi.)

THE RISE OF THE LOMBARDIC STYLE

221

search of Vlth century prototypes of porches with stone or wooden roofs. And
he would have realised that when Leo III, bishop of Nola (about 700), wanted to
erect the porch of the chapel of the Martyrs at Cimitile near his cathedral city, there
was no need for him to go
to Eastern architects for
the model. Certainly no
Eastern artist had a hand
in the execution of its
carving, as is proved by
the Corinthianesque capi-
tals of a type not in vogue
at that period in the East,
and undoubtedly the work
of Campanian chisels.

The Lombardic porch
was sometimes surmounted
by a niche or canopy, or
else by a loggia, the latter
being an adjunct of Raven-
nate origin, as we noticed
in our description of the
guard-house of Theodoric's
palace at Ravenna (VHIth
century). In other cases it
was constructed with two
tiers of columns, one above
the other.

I have discovered the
prototype of the Lombardic
porch in the cathedral of
Modena (1099-1 106) ; and,
accordingly, it is to the
architect Lanfrancus, and
his collaborators the master

masons Wiligelmus or William and Nicholas, that the honour of its creation is due
(Fig. 326).

IV. The column base with spurs strengthening the lower torus of the plinth.
These adjuncts, which hitherto have appeared only under the form of claws (e.g. in
SS. Felice e Fortunate, Vicenza, and San Babila at Milan), here assume a new shape,
viz. that of living creatures.

San Flaviano, which, in spite ot all that can be said to the contrary, 1 is an
important landmark in the history of Lombardic architecture, teaches us that at the
beginning of the second quarter of the Xlth century, the Lombardic organism with its
raised cross vaults and longitudinal, transverse, and diagonal visible arches ; with its
compound piers prepared to support them, surmounted by characteristic heavy
Lombardic cubical capitals, and resting on not less characteristic spurred bases ; with
its external buttresses corresponding to the internal transverse arches ; and with its
galleries, no longer intended to reproduce an Oriental Christian usage, or rather one

1 Mothes, Die Bauknnst des Mittelalters in /fa/ten.

Fig. 326. Modena. Duomo. Porta de' 1'rincipi (Xlth Century).

222

LOMBARDIC ARCHITECTURE

derived from Pagan Rome, but designed to counterbalance the pressure of the
vaulting ; it teaches us, I say, that this organism had not only been created, but
had even crossed the Apennines. And this fact enables us to place its first
appearance, with some show of reason, at the close of the first quarter of the
Xlth century.

THE CATHEDRAL OF AVERSA was built after the Norman count Rainulf
(1030-1047) had founded the city (IO3O), 1 and when the episcopal see was instituted

for its first bishop, Azolino (about
1049- 1056). 2 Its founder was
Richard I, proclaimed Prince of
Capua in 1062. His son, Gior-
dano I (1078), completed the
building as stated by the inscrip-
tion over an original door of the
church at the northern end of the
transept. It was damaged when,
in 1134 or 1135, Roger I, King of
Sicily, set fire to the city, but was
restored. Traces of both fire and
restoration are to be seen in the

Fig. 327. Aversa. Duomo. Plan of choir (about . , , , , ,

1049-1078). central tower, which has been a

good deal altered. Further restora-
tions were made necessary by the earthquakes of the XlVth, XVth, and XVIIth
centuries, those of the year 1349 involving extensive rebuilding. In the XVIIth
century it was reduced to its present condition.

Of the original structure there remains the spacious choir with an arcaded
ambulatory from which project
three radiating chapels vaulted
with half-domes (Fig. 327). The
arcade, now blocked up, has com-
pound piers supporting round
arches. From these piers and
from the wall piers spring
massive transverse arches and
powerful ribs of rectangular sec-
tion, about i ft. 8 in. across at
the base, built to carry the heavy
vault cells of the uncouth but
impressive ambulatory (Fig.
328).

This vaulting was con-
structed in the same way as
that in San Flaviano at Monte-
fiascone ; and like that its sur-
face was plastered. That it is

markedly raised, comes from the varying diameter of the arches, and also from the
fact that the diagonal ribs are nearly semicircular. There may be noticed an

1 Parente, Aversa (chiesa vescovile) Enciclofedia delf ecclesiastico, anno 1845. z Gams, op. cit.

Fig. 328. Aversa. Duomo. Ambulatory (about 1049-1078).

THE RISE OF THE LOMBARDIC STYLE 223

improvement in construction over the vaulting in San Flaviano : the ribs rest un-
diminished on shafts made to receive them, instead of being compressed at the
end (after the Roman manner) in order to find their way between the arches to their
support on the wall or nave piers.

The capitals, whether cubical or Corinthianesque, of the piers have representa-
tions of living beings, which are rather rude, but display an art somewhat more
advanced than that to be seen on a capital (now
used for holy water) in the abbey church of the
Trinita at Venosa (Fig. 329), founded by Count
Drogo between 1046 and 1051, and consecrated in
1059 by Pope Nicholas II (1059-1061). Not so
the foliage, which, though somewhat clumsy in
design, is carved with a sure hand, and is superior
to that in San Flaviano at Montefiascone.

The choir of the cathedral of Aversa, of one
and the same date in all its parts, is of great im-
portance, both for the history of Lombardic archi-
tecture, and for the origin of the Pointed style,
inasmuch as it was the first to exhibit an ambula-
tory with ribbed vaulting which a minute examination shows to be original. It has,
indeed, been stated that the vaulting cannot have formed part of the original
structure, 1 and is the result of alterations and additions, 2 for the curious reason that
ribbed vaulting had not then been attempted in France, even in this rude form,
forgetting that cross vaulting with diagonal ribs had made its appearance in San
Flaviano at Montefiascone as early as 1032. And that date is beyond dispute, being
officially stated, quite apart from the evidence of the building itself, by the original
inscription of the Xlth century which I have reproduced above.

THE CHURCH OF SANTA MARIA E SAN SIGISMONDO AT RIVOLTA D'ADDA was
erected by the people of Rivolta, and was given with all its possessions to Pope
Urban II by the regular canons, in exchange for special privileges, afterwards con-
firmed by the successive pontiffs, Paschal II (1099-1118), Calixtus II (1119-1124),
Innocent II (1130-1143), and Celestine II (1143-1144), as is stated in a Bull of
LuciusII (1144-1145), dated II44. 34 It must, then, have been built in the years when
Urban II filled the chair of St. Peter (1088-1099), and probably before 1095, that is
to say before the erection of San Giacomo at Como, the church of Rivolta being
designed with an apse which has no external open gallery. It may even have been
begun in the pontificate of Victor III (1086-1088).

What is certain is that it must have been completed (" propriis sumptibus a
vestris civibus aedificata") before 1099 if it was possible to hand it over, already built
and consecrated, to Urban II. The recent removal of the later accretions which
veiled the structure has restored the primitive appearance of the church.

It consists of a nave and aisles (Fig. 330). The former, about 32 ft. wide
between the bases of the piers, is of three square bays, two of which have ribbed
vaults decidedly raised, while the third has a barrel vault sustained midway by an

1 Bertaux, op. cit.

1 Schultz, Denkmaler der Knnsl des Mittelalters in Unteritalien.

* Vignati, Vila di S. Alberto Quadrelli Documcnti storiti su S. Alberto Quadrelli vcscovo di Lodi.

* Biscaro, / documenti intorno alia chicsa di San Sigismondo di Rivolta d'Adda.

224

LOMBARDIC ARCHITECTURE

arch. Each aisle has six square bays with ordinary intersecting vaulting, but raised.
San Babila had clearly produced a school of imitators.

The two cross-vaulted bays of the nave
(Fig. 331), with massive rectangular stone ribs,
describing almost a semicircle, but slightly
pointed at the summit (where they measure
about i ft. 8 in. across), are concave-crowned.
They are constructed, so far as the cells are
concerned, in the same way as those of San
Flaviano at Montefiascone, only with a difference
in the material which here consists, in the case
of the stratum laid upon the wooden centring
kept up by the vaulting ribs, of broken bricks,
stones, and pebbles, sometimes arranged in
herring-bone fashion. In this vaulting, though
the ribs are pointed, they are not sufficiently so
to show an application of the principle of the
pointed arch.

The piers are alternately larger and smaller.
The former are in section like those of San
Babila at Milan (Xlth century), i.e. cruciform
with engaged shafts in the re-entrant angles
(Fig. 332). The latter have a section identical
with those of the church of Montalino at Stradella (Fig. 333), viz. quadrangular piers
with a half column on each face. The continuous cubical capitals belonging to these
piers and the corresponding wall piers (partly renewed or retouched, though keeping

Fig. '330. Rivolta d'Adda. Plan of church
of Santa Maria e San Sigismondo Xlth
Century).

Fig. 331. Rivolta d'Adda. Church of Santa Maria e San Sigismondo (Xlth Century).

THE RISE OF THE LOMBARDIC STYLE

225

Fig- 332. Rivolta d'Adda. Church
of Santa Maria e San Sigismondo.
Section of one of the large piers
(Xlth Century).

Fig- 333- Stradella.
Church of Monta-
lino. Section of
pier (Xth Century).

the original design and execution) are ornamented with scroll work, foliage, cauliculi,
and living creatures of real or fanciful origin, which, on the whole, show very little

difference in modelling and execution from those in
Sant' Ambrogio at Milan. In fact,
it might almost be said that the
best had come from the same hand
as the latter. The bases have at
the corners of the plinth strengthen-
ing spurs, in some cases carved as
human heads.

With one exception, the lunette
produced by the junction of the
lateral cell with the side wall is
pierced by two windows larger than
those in the aisles. A similar
arrangement may be seen in an

ancient Roman building the design of which has been preserved by Serlio. 1 We learn
from a Vatican MS. 2 that it formerly existed on the Via Appia (Fig. 334). In the
facing of the walls the use of " opus spicatum " is raised to the level of a real system of
construction.

Answering to the transverse arches in the aisles are substantial buttresses, as in
San Babila at Milan, connected by ramping ones with the similar series that meet
the thrust of the transverse arches
of the nave (Fig. 335). The open
arcaded gallery round the exterior
of the apse is the result of altera-
tions which also affected the front
of the church, apparently in the
Xllth century.

I f it possessed a matroneum or
triforium gallery, the church of
Rivolta d'Adda would exhibit the
complete Lombardic organism.
Everything points to its having
been built before Sant' Ambrogio
at Milan, but only just before.
The similarity of the salient fea-
tures in the two structures is ob-
vious. The carvings at Rivolta not
only reproduce the same decorative
elements among them the new
one of animals rampant but even
suggest the hands and the charac-
teristics Of the gild Which worked Fig. 3 34- -Rome. Tomb on the Via Appia. (From Serlio,
in Sant' Ambrogio. The artistic " De le aniiguiict.")

advance sometimes noticeable in

the capitals of the latter can easily be explained by the greater experience gained by
the carvers. In the secondary piers there is just the same arrangement of a half-

1 Of. cit. * Vatican Library. Cod. Lat 3439.

VOL. I Q

226

LOMBARDIC ARCHITECTURE

column attached to the face in order to resist the thrust of the transverse arch in the
aisle. In Sant' Ambrogio it also carries the shaft supporting the arched corbel course
below the gallery, whereas here it has no such function to perform as there is no
gallery to decorate.

Again, the ribbed vaulting of the nave in the two churches is closely related. In
either case it covers a square bay corresponding to two bays of the same form in each
aisle. Both transverse and longitudinal arches are semicircular. In Sant' Ambrogio
the latter are constructed, like the ribs, of brick with bands of stone at intervals. The
diagonal ribs are alike in section, and slightly pointed. This strong supporting frame-
work is, in both churches, quite independent of the masonry of the cells which rest
upon it. The latter, in the case of Sant' Ambrogio, are made of brick, with a maxi-
mum thickness of about i ft. 8 in. at the summit.

The somewhat rude construction of the vaulting at Rivolta d'Adda compared
with that in Sant' Ambrogio is to be explained by the old material which was used

Fig- 335- Rivolta d'Adda. Church of Santa Maria e San Sigismondo. South side (Xlth Century).

up in it, rather than by any great lack of skill in the builders. Between San Flaviano
at Montefiascone and the church of Rivolta there elapsed an interval long enough for
the evolution of ribbed cross vaulting from its elementary form to the complete
system of ribs sustaining cells of brickwork to be seen at Sant' Ambrogio, and for its
application to the church of Rivolta, though in a less advanced form so far as the
cells are concerned. The intersecting vaults of the aisles in both cases are raised.
The hood moulding of the main door is a roll springing from two attached shafts.
The lateral doors have plain jambs, and above them are very narrow windows splayed
on both faces.

The church of Rivolta d'Adda is a compound cf San Babila at Milan and San
Flaviano at Montefiascone, but nevertheless it marks a notable advance beyond them
in the principles of construction and statics. From San Flaviano came the idea of
ribbing the vaults, from San Babila that of giving them a raised and concave-crowned
form, of which the earliest example that I know is afforded by the presbytery of San
Vitale at Ravenna (526-547) (Fig. 336). From San Babila, too, it derived the system

THE RISE OF THE LOMBARDIC STYLE

227

of buttressing. But at Rivolta it was used to counterbalance the thrust of the wide
vaulting of the nave with its great span, and discharge it on to the powerful buttresses
of the aisle vaulting. And it was the adoption of cross vaulting that made it
possible to light the nave directly, by inserting windows in the lunette wall spaces.

In short, the church reveals an understanding of the principles of thrust and
abutment which will not be surpassed in later days by that of the Pointed style. But
the Lombardic organism still required for its completion the construction of galleries

Fig. 336. Ravenna. San Vitale. Upper part of presbytery (526-547).

above the aisles to counterbalance the thrust of the vaulting, and the direct lighting
of the nave. We will now pass to the building which marks the last stage but one
on the way towards the attainment of this goal.

THE BASILICA OF SANT' AMBROGIO AT MILAN. Most of those who have
written about this celebrated example of a vaulted basilica have indulged their fancy
in guessing at its date as though they were playing a card in a game of chance.
But architecture and the science of statics obey a law of progressive development.
And so, we cannot investigate the date of a given architectural organism until
we have first of all mastered the history of those which preceded, or were
contemporary with, or followed it, in order to make sure that what is regarded
as an original feature in such an organism may not turn out to be an original element
in an organism of quite another type and period. Nor is it possible to fix within
definite limits, even approximately, the age of a building, without a full knowledge

Q 2

228 LOMBARDIC ARCHITECTURE

of that history, and without a series of dated monuments to guide us, which,
like links in a chain, mark the fixed points in the gradual formation of the
organism. That formation is a slow process, for it is not in the nature of a new style
of architecture to come into the world, by an act of spontaneous generation, in all
its beauty and completeness.

It is because these ways have remained untrodden ways which are hard
to traverse, though sure and leading to results of certainty, that we find ourselves
in the presence of two schools of writers, at loggerheads with one another, and both
of them with logic and with facts. The first puts the date of the nave and aisles
of Sant' Ambrogio in the episcopate of Angilbert II (824-860), thus making
a perfectly new system of construction and equilibrium come into existence all
in a moment, at a date when the element from which it sprang, viz. the compound
Lombardic support, consisting of piers combined with columns, had not yet seen the
light. So that we are asked to accept a phenomenon belonging to the sphere
of the miraculous : an organism, that is to say, which has reached almost its full
development before it has passed through the embryonic stage, and then by some
mysterious process dies away to rise again, like some new Phoenix from its own ashes,
in the Xlth century.

The Lombardic pier, the plan of which contains all the elements of the
development in elevation of the building, appears in its elementary form, viz.
a quadrangular block with a half-column attached to each side, only in the Xth
century, as we saw in the case of the church of Montalino at Stradella. In its more
developed form of two piers set cross-wise with four columns in the angles, it is seen
at the opening of the next century in San Babila at Milan. These two types formed
the models for every compound support used in Lombardic architecture and
its derivatives.

The origin of this pier is not to be sought in the bundles of shafts used in
remote times in Syria with the object of producing a multiplied impost, but rather
from the Roman piers with engaged columns, and also from the cruciform examples
with vaulting shafts engaged in the re-entrant angle, like those used in the Basilica
Julia at Rome, and intended to provide the imposts for the longitudinal arches, the
secondary transverse arches, and the springers of the vaulting ; or, thirdly, from
compound piers of the kind adopted for a portico near the Theatre of Marcellus also
at Rome, illustrated in Fig. 1 1 5.

The second school, on the other hand, in spite of the recent revelation of the
original church of Santa Maria e San Sigismondo at Rivolta d'Adda, and in the
face of the ribbed vaulting in San Flaviano at Montefiascone and in the cathedral of
Aversa, and though it has to travel (generally only on paper) to the North of France
to find a type of vaulting in the Xllth century which was created in Italy a
century before this school, I say, would bring the Ambrosian Basilica down to a
point well on in the Xllth century. It never seems to have struck them as
inconceivable that the Lombard gilds could have built years before, in an unim-
portant place, a church such as that of Rivolta d'Adda, with an organic structure of
so advanced a kind, and then have started to erect another of only a slightly more
developed form in the most important centre of life in Lombardy. And that too,
when the organism had already attained its completion in San Michele Maggiore at
Pavia, marking the final stage in the way towards the perfection of the Lombardic
system ; so that they would make the architect of Sant' Ambrogio take a step
backwards on the ascent which the development of that system forms.

THE RISE OF THE LOMBARDIC STYLE

229

But besides all this, a recently published document 1 proves, however much
people may try to minimize its force, the circumstances and the date of the erection
of Sant' Ambrogio as we see it. Having said so much, I proceed to give a summary
of the information which has come down to us about the principal changes which the
church has undergone.

The first basilica was raised in honour of Saints Gervasius and Protasius
by St. Ambrose (374-397) : " ... in basilicam quam ipse
proprio aedificavit studio."

In 784 Archbishop Peter (784-805) handed it over to the
charge of Benedictine monks, who in 789, with their first
abbot Benedict (784-806), took up their residence in a convent
built to receive them.

In 835 Archbishop Angilbert II (824-860) gave the
monks the famous altar made by Wolvinius, after having
provided them with a new abbot in the person of Gaudentius
(f 842), and restored the monastic discipline.

Archbishop Anspert of Biassono (869-882) carried out
the works referred to in a line of his epitaph : " Atria vicinas
struxit et ante fores."

Archbishop Anselm V of Pusterla(i 126-1135) had anew
campanile erected on the canons' side of the church, corre-
sponding to the " Monks' Tower" on the other side, and gave
it (in an unfinished state) to the former in 1128. It was
built, at the expense of the city, by the same architect who
had rebuilt the church: "... cum eiusdem ecclesie architectus
ipsum (campanile) sicut aliam. ecclesie fabricam de communi
construxerit."

After being partially ruined in 1 196 it was restored, and
the cupola was rebuilt under Archbishops Hubert (i 195-1 196)
and Philip (i 196-1 2o6). 23 * 5

The story of these changes may be told more fully as
follows (Figs. 337, 338, 339).

Between 789, when the monks were definitely installed
in the basilica, and 824 when Angilbert II was consecrated
as head of the Milanese Church, the rebuilding of the apse
was taken in hand, and also the construction of the rectangular bay in front of
it, an arrangement evidently due to the need felt by the monks, after their
definite settlement in Sant' Ambrogio, of enlarging the space set apart for religious
functions. Further, they built the campanile, which was appropriately christened
the " Monks' Tower."

Next, in the episcopate of Angilbert, the two lateral apses were added
(manifestly of later date than the previous work), and there was also carried out the
rebuilding of the nave and aisles, and the reconstruction of the fagade.

Later, Archbishop Anspert added the atrium, as is stated in his epitaph. This
atrium, as designed, was open all round on both sides, at any rate in the front, but

1 Archiino storico lombarao, 1904, 1905. Biscaro, Note e documenti Santambrosiani.

2 Man. Germ. hist. Cregorii rpiscopi Turoncnsis liber in gloria martyrum.

3 Puricelli, op. fit. 4 Giulini, op. fit.
6 Archivio storico lombardo, 1904, 1905. Biscaro, Note e documenti Santambrosiani.

A Choir (789-824).
BB'Apsidal chapels

(824-860).
C Body of the church

(Xlth Century).
D Narthex (Xlth Century).
E Atrium (X I Ith Century).
F Canons' Tower

(1126-1128).
G Monks' Tower

(789-824).

F 'g- 337- Milan. Plan of
Sant' Ambrogio.

230

LOMBARDIC ARCHITECTURE

was closed soon afterwards, for the masonry of the arches and of the walls which
block them up evidently belongs to about the same date.

Finally, in the last quarter of the Xlth century, and before 1098, the year of
the institution of the festival in honour of SS. Gervasius and Protasius on the occa-
sion of the rediscovery of the bodies of the martyrs, 1 a discovery which must have
taken place in the course of the rebuilding works, the conversion of the nave and
aisles separated by columns into a nave and aisles covered with vaulting was carried
out. But this was done without altering the original arrangement of the church, or

Fig. 338. Milan. Sant' Ambrogio (Xlth and Xllth Centuries).

touching the three apses and the frontal wall connected with them, and utilising to a
considerable extent the old foundations. At the same time was built the narthex
with the characteristic and striking loggia above it, but quite independent of any
plan of an atrium in front of it.

This conversion was, in all probability, the result of the entrance of the popular
element into the government of Milan. We know, as a matter of fact, that the
people of Milan, which had obtained from Archbishop Aribert (1018-1045) the right

1 Archivio storico lombardo, 1904, 1905. Biscaro, Nets e documenti Santambrosiani.

THE RISE OF THE LOMBARDIC STYLE

231

to bear arms, and with it the consciousness of its own power, achieved its triumph at
the death of that proud prelate, when the government of the city took the form of a
Commune. What wonder then if the people, remembering the steps taken by
Angilbert and Anspert to affirm the archiepiscopal lordship, desired in its turn, now
that those prerogatives were at an end, to assert in the most impressive form their
own advent to power, their own supremacy ?

In this transformation the Monks' Tower was left untouched, but now it was
incorporated with the main building. The next step was to engraft, in the early years
of the Xlth century, the present atrium with its enclosing cloister on to the narthex.
Lastly, between 1126 and 1128, the erection of the Canons' Tower was taken in hand,

Fig. 339. Milan. Sant' Ambrogio. Atrium and fa9ade (Xlth and Xlllh Centuries).

involving the partial demolition of the left side of the basilica. The tower in modern
times has been raised in height, a fact which is patent to the observer.

The changes and enlargements here set forth find their confirmation in a number
of considerations which I will briefly state.

It is unnecessary to discuss the Monks' Tower, as there can be no question about
its date, or the three existing apses and the bays separated by walls in front of them
(an arrangement which recalls that in the large basilica of St. Symphorosa on the Via
Tiburtina near Rome), for we know that they date back to a period earlier than the
body of the existing basilica, as may be seen in plans which have been published. 1 '-

These plans make the remains of the columned nave and aisles go back to the
time of St. Ambrose, whereas they only date from the episcopate of Angilbert II.

1 Landriani, op. fit.

2 Beltrami, AmbrosianaLa basilica ambrosiana primitiva e la ricostruzione compittta nel secolo IX.

232

LOMBARDIC ARCHITECTURE

To begin with, where do we find churches erected before the end of the Xth
century which show scroll work and figure capitals of the type of those in Sant'

Ambrogio (Figs. 340, 341,
342) ? It has not been my
fortune to come across a
single specimen. And there-
fore the nave and aisles of
Sant' Ambrogio were not
constructed before that
period.

Where, too, before the
second half of the Xlth cen-
tury, are we to look for
basilicas in the Lombardic
or derived styles, of ascer-

Fig. 340. Milan. Sant' Ambrogio. Capital (Xlth Century).

tained date, with a complete
system of raised cross vault-
ing for nave and aisles, wholly or partially ribbed ? I am not in a position to adduce
a single example either in Italy or in Normandy, in spite of the remarkable progress
made by the Lombardic
style in the latter country
after the epoch of 1000.

Moreover, we must re-
member that, while the Lom-
bardic basilicas of the early
Xllth century, with their
complete system of vaulting,
e.g. San Michele Maggiore at
Pavia, admitted windows not
only in the side walls of the
aisles and galleries above
them, but also in the walls
of the nave, and with this
object had the imposts of the nave vaulting raised ; the architect of the Ambrosian
basilica, on the contrary, perhaps from a fear of imperilling the stability of his nave
vaulting if the walls were raised sufficiently to allow of windows being made in

them, was content to light
the aisles and galleries with
windows in their side walls,
while he relied on those in
the west front for lighting
the nave.

The conclusion is that
the conversion of the arcaded
Sant' Ambrogio into a
vaulted church was carried
out before the Xllth century.
Added to all this is
Fig. 342. Milan. Sant' Ambrogio. Capital (Xlth Century). the evidence provided by the

Fig. 341. Milan. Sant' Ambrogio. Capital (Xlth Century).

THE RISE OF THE LOMBARDIC STYLE

233

principal door of the basilica (Fig. 343), which, although it has been tampered with,
is still, as a whole, the work of Master Adam, who may with some reason be regarded
as the architect of the existing basilica and of its new campanile. There are no
grounds for making an exception even in the case of the jambs, which have been

Fig- 343. Milan. Sant' Ambrogio. Portal (Xlth Century).

thought to belong to the IXth century ; whereas any trained eye will see at once the
close relationship between the carving on the bottom piece of the right jamb and that
on the two shafts on either side, which are unquestionably the work of Master Adam.
Anyone, too, who is acquainted with the facts knows that the subject of Hercules

234

LOMBARDIC ARCHITECTURE

preparing to attack the Nemean lion, carved on the lowest piece of the left jamb
(Fig. 344), was not treated by the Lombard mosaic workers and carvers before the
Xlth century. So that we shall not be far wrong if we say that, of the six pieces
of marble which form the jambs, one is from the hand of Master Adam himself, and
the others are the work of contemporary artists.

The date of the portal cannot be earlier than the second quarter of the Xlth
century. Anything else is contradicted by the bases with figure spurs. The oldest
dated examples of these spurs belong only to the second quarter of that century,
and we noticed them in San Flaviano at Montefiascone. On the other hand, it is

Fig. 344. Milan. Sant' Ambrogio. Part of left jamb of the portal (Xlth Century).

later than the first half of the Xlth century, by reason of the carving on the shafts, the
lintel, and the archivolts, which marks an obvious advance over that in San Flaviano
at Montefiascone, especially in the treatment of animal life. At the same time it
must be rather earlier than the portal of the cathedral at Modena, on account of the
somewhat more advanced character of some of the carving on the latter. And it is
decidedly older than the portals at Pavia belonging to San Michele Maggiore (erected
just after the memorable earthquake of 1117) (Figs. 345, 346) and San Pietro in Ciel
d'Oro (built after 1117) (Fig. 347), in which we find a fairly obvious advance,
especially in the treatment of figures, over the carving in Sant' Ambrogio.

In short, the date which we arrive at agrees with that which may be gathered
from the documentary evidence. 1 And that evidence authorizes us with good reason

1 Archivio storico lombardo, 1904, 1905. Biscaro, Note e docuincnti Santambrosiani.

THE RISE OF THE LOMBARDIC STYLE

235

in placing the erection of Sant 1 Ambrogio and the Canons' Tower within the
possible limits of the forty years between 1088 and 1128. There fall into their proper
places in this period: (i) the institution in 1098 of the festival in honour of SS.
Gervasius and Protasius, when the basilica must have been finished ; (2) the building
of the atrium against the narthex, some years after 1098, as shown by the advanced
art of the original carvings on the three sides which enclose it ; and (3) the incorpora-

Fig. 345. I'avia. Facade of San Michele Maggiore (Xllth Century).

tion of the Canons' Tower with the body of the basilica. And this period stands in
the relation which we should expect to the date of the church at Rivolta d'Adda.

Everything points to the maker of the portal, " Adam Magister," being the
architect of the basilica. The name Adam appears on the well-known inscription,
with the date 1098, still to be seen on the outer front of the atrium of Sant'
Ambrogio. It is also registered in a deed of purchase of a piece of land at
Comabbio, executed at Milan in 1087, with the description as son of Albert
" qui dicitur Melanense de loco Comabio." Thirdly, it occurs in the form of " Adam
magister de Sancto Sepulcro " in the parish of Brebbia, in a second deed of purchase,
dated 1094, of another piece of land, also in the territory of Comabbio. 1 Both
1 Archh'io storica lombardo, 1904, 1905. Biscaro, Note e documents Santambrosiani.

236

LOMBARDIC ARCHITECTURE

places, San Sepolcro and Comabbio, were the property of the monastery of Sant'
Ambrogio.

In the portal of Sant' Ambrogio the animals rampant standing up against the
shaft of a column should be noticed. At Rivolta d'Adda the motive had already
been applied to another member, the capital. This motive of animals rampant, and
sometimes " regardant," on the columns of portals, those at Sant' Ambrogio being the
prototypes, was derived from Etruria. The sepulchral cippus from Settimello, now

Fig. 346. Pavia. San Michele Maggiore. Portal at the side of the church (Xllth Century).

in the Archaeological Museum at Florence, and considered to belong to the second
half of the Vlth century B.C., 1 has four rampant lions with their heads and fore feet
turned outwards (Fig. 348).

This portal is the oldest surviving example of the kind which I can point to in
Italy. The earlier one in Sant' Andrea at Montefiascone is not in its original state.
North of the Alps we shall find the earliest specimen of a Lombardic portal, though
1 Milan!, Cippo di Settimello Atti del/a R. Accad. dei Lined, 1903.

THE RISE OF THE LOMBARDIC STYLE

237

not so characteristically Lombardic as that of Sant' Ambrogio, in St. Iitienne at Caen
(1066-1086), the work of Lanfranc of Pavia.

The creation of the doorway penetrating a wall with widely splayed sides and
two orders of jamb shafts, surmounted by a lunette sunk in the middle of concentric
archivolts and roll mouldings corresponding to the shafts, took place after the epoch
of looo, and was the
work of the Lombard
gilds. In the West
it first appeared in
buildings of which
they were the au-
thors, and in the East
no building earlier
than the Xllth cen-
tury contains it at all.

The basilica of
Sant' Ambrogio is a
combination of San
Flaviano, with its
galleries and un-
broken roof, and of
the church of
Rivolta, with its
raised cross vaulting
for both nave and
aisles, and system of
buttresses. And
though the architect
secured the stability
of the nave vaulting
by keeping its spring
rather low, and flank-
ing it by the galleries,
thereby sacrificing
the direct lighting of
the nave, still there
is the fact, unparal-
leled at the time, of
a nave at least 44 ft.
wide, covered by
cross vaulting in
brick, sustained by
arches of brick and

stone.

Fig. 347. Pavia. San 1'ictro in Cit-1 <T Oro. Portal (Xllth Century)-

Attention, too, must be called to the construction of the intersecting vaulting
with diagonal ribs in the nave (the first two bays being in the main untouched),
designed, like the masonry of the walls, to be visible, and not plastered over as at
Rivolta d'Adda, Aversa, and Montefiascone. Here the vault cells, besides being made
of straight courses of bricks, are no longer geometrically parts of one cylindrical

2 3 8

LOMBARDIC ARCHITECTURE

surface as in the past, but each cell shows internally
a concave surface curved in every sense. A hori-
zontal section taken at any point in the vaulting
would show a quadrilateral figure with curvilinear
sides. The curve of such vaulting is arbitrary,
that is to say it rested with the constructor to carry
it up just as high as he pleased from the spring.
And the raised outline was not merely intended
to resist pressure, but was partly chosen for its
appearance.

The vaulting was constructed by first putting
up temporary centres for diagonal and transverse
arches, and building these in brick bound with
stone, or entirely in stone. Next, on these per-
manent centres was formed a temporary rough
boarded centring, with the surface modelled up
in earth or in clay and water, to receive the cells.
The slight raising forming the concave crown was
obtained by an additional small centring. This
concavity was intended, among other objects, to
strengthen the vault in its weakest point. The
vaulting of this kind in Sant' Ambrogio, showing
as it does an advance beyond the other cases of

ribbed vaults

Fig. 348. Florence. Archaeological
Museum. Etruscan cippus from
Settimello (Vlth Century B.C.).

Fig- 35 - Como. San Giacomo. Apse
(Xlth and Xllth Centuries).

which we have

discussed, is the

earliest of its

species in any building either in the West or in

the East.

And now we will conclude our argument
with the examination of a peculiarity presented by
the oldest part of Sant' Ambrogio, the apse, on
the exterior of which we see high up a range of
deeply recessed arched niches, divided into groups
by lesenas. This treatment, which is another
creation of the Lombard gilds, and apparently
derived from the rectangular cavities grouped in
threes by vertical rolls, as seen in San Pietro at
Toscanella (739), is something quite new ; for
although long before the date of Sant' Ambrogio
the central apse of St. Sophia at Salonica (about
495) exhibited small arched niches (provided that
originally they were not openings intended to throw
light on the mosaics of the interior), the motive
has no direct connection with the arrangement
shown in the apse of Sant' Ambrogio. Originally
confined to apses, it was later applied to circular
buildings and baptisteries, as may be seen, for
instance, in the baptisteries of Agliate (824-860),

THE RISE OF THE LOMBARDIC STYLE

239

Fig. 349. Milan. San Lorenzo Maggiore. Chapel of Sant' Aquilino. External gallery

(Xlth Century).

Fig. 351. Modena. Uuomo (1099-1106).

240

LOMBARDIC ARCHITECTURE

Biella (Xth century), and Novara (Xth century), and in the "Rotonda" at Brescia
(Xlth or XI Ith century).

The next step was to convert them into external open galleries, the earliest
known example of which is afforded by the chapel of Sant' Aquilino attached to
San Lorenzo Maggiore at Milan (Fig. 349), where it does not belong, as some suppose,
to the original structure of the chapel, but is an addition made after the fire of 1071,.
with the object of strengthening the cupola. The fact can easily be verified by anyone
who compares the masonry of the original building with that of the more recent gallery.
This type of gallery was afterwards embellished in the manner shown by San
Giacomo at Como (Fig. 350). The date of this church comes between the last decade
of the Xlth century and the year 1117; in other words, before the ten years' war

Fig. 352. Como. San Fedele. Apse (Xllth Century).

between Como and Milan. 1 It must be later than the erection of Sant' Abondio in
the same town, for it is evidently some years younger, and derives some of its features
from that church ; but it cannot have been begun long before 1095, for if it had been
finished by that year it would have been consecrated by Urban II like Sant'
Abondio, which was not the case.

In their ennobled and elaborated form the galleries were employed with a new
intention by the architect Lanfrancus, in the form of the wholly or partially communi-
cating galleries encircling the cathedral of Modena (Fig. 351) which he designed and
carried out (1099- ii o6). 2 The master builders of Pavia, again, applied them to the

1 Rivista arch, della proiniicia di Como, fascicolo 30. Barelli, La chicsa di San Giacomo in Como,
' 2 Muratori, Rerum Ital. script. Translatio tarpon's s. Gcminiani.

THE RISE OF THE LOM BARDIC STYLE

241

fronts of churches, stepped so as to follow the slope of the gable. San Michele
Maggiore at Pavia was the first to exhibit this treatment.

Being a Lombard creation, they spread rapidly in Italy through the agency of the
Lombard gilds. Thus, for instance, they were used to encircle the exterior of the apses
of San Frediano at Lucca (i 1 12-1 147), 1 Santa Maria Maggiore at Bergamo (begun in
1137, as is proved by the inscription in the south porch), 2 and San Fedele at Como
(Fig. 352). The apse in the last case retains nothing of the structure of 914 (as others
besides me have pointed out 3 ), and after a careful examination I am inclined to place
it in the Xllth century. External galleries were also carried round the cathedrals
of Parma (Xllth century) (Fig. 353) and Piacenza, rebuilt in 1 122 4 (as is confirmed by

Fig. 353. Parma. Duomo (Xllth Century).

the inscription on the front) after the destruction of the old church by the earthquake
of 1 1 17. At Rome it made its first appearance in the apse of SS. Giovanni e Paolo,
as part of the works of restoration and embellishment (iO99-i2i6). 5 In the Venetian
region a notable specimen is afforded by the church of SS. Maria e Donato at
Murano (Fig. 354), which must have been restored after the earthquake of 1117,* and
finished by 1 140, as that date is inscribed in the beautiful mosaic pavement.

The introduction of open external galleries in the church at Murano, whereas
the apse of St. Mark's at Venice (1071-1094) is merely decorated with plain blank

1 Ridolfi, Guida di Lucca.
3 Dehio und von Bezold, op. cit.
8 P. Germano di S. Stanislao, of. cit.
VOL. I

3 Ronchetti, Memorie storichc della cittti e chiesa di Bergamo.

4 Muratori, Kentm Ital. script. Chronicon Placentinum.
8 Man. Germ. hist. Annales Venetici brrves.

242

LOMBARDIC ARCHITECTURE

arcading, enables us to fix approximately the date of the apse of Santa Fosca
at Torcello, where the arcades are still blank, but have been to some extent
elaborated. The date, then, will come between the rebuilding of St. Mark's and
that of San Donate, that is to say in the last years of the Xlth century, or at latest
in the first years of the Xllth.

The open galleries of the cathedral of Pisa (Fig. 355) might suggest that the
Lombardo-Tuscan school anticipated the Lombardic in its application of this
treatment, the church being identified with the one begun in 1063 and finished,
according to Rohault de Fleury, in iioo. 1 Another writer 2 thinks that the building
was for the most part complete in 1118, when it was consecrated by Pope Gelasius II

(1118-1119). It is
also stated 3 that in
1104 it was still
some way off com-
pletion, inasmuch as
between 1070 and
1185 the judges or
kings of Cagliari,
Gallura, and Ar-
borea, and, several
years after 1 100, the
Byzantine emperors,
made important
donations to the
still unfinished work
of Santa Maria.
Further, that the
structure itself is
obviously lacking in
unity, owing to the
way in which the
different parts meet
in the middle ; that
the south side of the

F "g- 3S4- Murano. Church of SS. Maria e Donate (Xllth Century). western limb shows

a deviation to the

extent of about 2 ft. 8 in. from the straight line, a circumstance which would
support Rohault de Fleury's idea that at this point came the angle of the
facade according to the original plan of a cross with equal arms ; and that
various similarities and differences betray some enlargement and alteration of
the edifice. Moreover, that the Rainaldus of the inscription on the west front
must be the master builder of the same name mentioned in a document of 1264 as at
work on the cathedral ; and that the platform round the building was made between
1298 and 1300. From all which it would appear that the imposing pile would
seem to be the result of a general remodelling and enlargement of the original
building of the Xlth and Xllth centuries, carried out in the XHIth.

The only comments I make are these. External open galleries made their first

1 Les monuments de Pise. * Supino, Italia artisticaPisa.

3 Rassegna settimanale ttniversale, 1898. Fontana, Alcune osservazioni intorno al duomo di Pisa.

THE RISE OF THE LOMBARDIC STYLE

243

appearance in Tuscany (I speak of cases where the date is known) in the apse
of San Frediano at Lucca (1112-1147). They were lavishly used on the front
of San Paolo a Ripa d'Arno at Pisa, recalling that of the Duomo, but not till after
1148, when Pope Eugenius III (1145-1150) consecrated the high altar. 1 To an
experienced eye, acquainted with the subject, the cathedral of Pisa does not present
an organic whole which came into existence at one time, designed to produce a
preconceived effect of lines and masses, as the exterior of San Paolo just referred to
does. On the contrary, it proclaims itself to be the result of alterations, and of
a change in the decorative scheme. Then, the two ranges of external galleries
round the apse are certainly not so old as the construction of 1063, for at that
date the scheme was not yet invented. And the four tiers of galleries on the facade

Fi g- 3SS- pi sa- Duomo (Xlth, Xllth, and XHIth Centuries).

must be later than the building of the cathedral of Modena (1099-1 106), for when the
latter was erected the arches of the galleries of the fagade were not independent but
enclosed by relieving arches. They are also later than the rebuilding of the
cathedrals of Piacenza and Parma, where the fronts are treated with one or two
detached ranges of galleries and another stepped so as to follow the pitch of the
gable.

And, lastly, of the three chief characteristics of the Pisan buildings the
banded facing, the blank arcades, and the open galleries the first was imported from
Syria ; the second was derived from Ravenna and Lombardy, but given an improved
form ; and the third was borrowed from the Lombards with a greater scope for effect
given to it by its use on facades.

North of the Alps the earliest specimens are to be seen in the cathedral at

1 Jafife, Kegesta pontificum Romanorum.

R 2

2 4 4

LOMBARDIC ARCHITECTURE

Speyer, where they are a result of the alterations carried out from 1 137 to 1 146, and
in that of Bonn belonging to the years from 1 130 to 1 169.

It now remains for us to glance at a building which in my belief marks
the last stage on the way towards the completion of the Lombardic vaulted
basilica.

THE CHURCH OF SAN MICHELE MAGGIORE AT PAVIA. We are entirely
without documentary evidence about its foundation and rebuilding. We are told 1
that the first church, erected in the days of Gothic or Byzantine rule, was in

existence in 642, as is confirmed by
Paulus Diaconus, and reached the Xlth
century without having suffered any
important injury in the course of the
devastations and conflagrations of
which Pavia was the scene. Further,
that it was rebuilt, either in conse-
quence of some catastrophe, the de-
tails of which have not come down
to us ; or because, in the new era of
prosperity inaugurated by the rise of
the Commune, the people of Pavia
wanted a new church more consistent
with their ideals in the Xlth century ;
or, thirdly, some time in the long
period of silence which envelops the
basilica after the year 1008 (when it
is mentioned in a donation by Otto
son of King Ardoin), and remains un-
broken till the year 1155, when it was
the scene of Barbarossa's coronation.

My view is that the existing
structure arose directly after the
terrible earthquake of 1117, recorded
by Muratori, 2 which must have brought
on Pavia a catastrophe similar to those
which ruined Verona, Vicenza, Parma,
Cremona, and other Italian towns.
South side This much is certain, that other ancient
churches in Pavia were rebuilt, such as
San Teodoro, San Giovanni in Borgo

(now destroyed, with a facade which combined the features of San Michele Maggiore
and those of San Pietro in Ciel d'Oro), and also the double cathedral, the date of which
is placed by De Dartein, 3 for Santo Stefano in the Xlth or Xllth century, and for
Santa Maria del Popolo in the Xllth, though what is left of the two churches shows
traces only of the Xllth century.

San Michele is really later than the church of Rivolta d'Adda and Sant'

1 Atli della R. Accademia dei Lined, anni 1895-1896. Merkel, Uepitaffio di Ennodio e la basilica di San
Michele Maggiore in Pavia.

' l Annali d 'Italia. * Of. cit.

Fig. 356. Pavia. San Michele Maggiore.
(Xllth Century).

THE RISE OF THE LOMBARDIC STYLE

2 45

Ambrogio at Milan, because we not only find it exhibiting a more perfect
organism, but also because it shows in its carvings, taking into account the material

F 'g- 3S7- Aosla. Remains of the Roman Theatre (about 25 B.C.).

used, a remarkable artistic progress compared with those of the two other churches,
especially in the figure subjects, which also display a fairly obvious improvement
over those of the cathedral at Modena (1099-1106). On the other hand, it is some
years earlier than San Pietro in Ciel d'Oro. If proof be needed one has only

Fig. 358. Plan of Roman building. (From San Fig. 359. Rome. Plan of tomb outside the Porta Salaria.
Gallons sketch book in the Vatican Library.) (From Montana, " Li cinque tibri cTarchilettura."')

to compare the archangel over the portal of the latter with the similar figures on the
doorways of San Michele ; or, again, the figure subjects on the capitals in the
two churches.

246

LOMBARDIC ARCHITECTURE

The old restorations and reconstructions in brickwork to be seen on the exterior

of San Michele (Fig. 356) date from 1489, when the structure threatened "maximam

ruynam in corpore et voltis

tarn de medio quam a lateri-
bus ipsius ecclesie"; 1 and
the danger was so immediate
that the urgent work of re-
newal and restoration had
been already taken in hand.
It was carried out by Master
" Augustinus de Candia filius
quondam M. Jacobi," as
stated in a document of
October 3rd, 1489. I note
here that this Master James,
with his brother, put up in
1487 the present vaulted roof
over the central part of San
Pietro in Ciel d'Oro (recon-
secrated in 1132), as is proved
by an inscription preserved
by Bossi in his manuscript
collection of Pavian inscrip-
tions now belonging to the
University of Pavia. This
roof replaced an older one. 2
From these documents and others preserved in the Museo Civico it appears that

the vaulting of the nave in San Michele was rebuilt, and that of the aisles restored

and put in order ; that

work was done on the

buttresses, the upper part

of the walls of the body

of the church, the cupola,

the presbytery, and the

apse ; while the barrel

vaults of the transept

were secured with iron

tie rods. A visit to the

space between the vault-
ing and the roof enables

one to estimate without

difficulty the amount of

reconstruction and altera-
tions which took place

at the close of the XVth

century (in which brick

was used), and what was

1 Museo Civico, Pavia. Petition (supplic.i) of Hieronimus Varixius de Roxate, provost of the church,
transcribed for me by Mgr. R. Majocchi, Rector of the Collegio Borromeo, Pavia.

2 Majocchi e Casacca, op. cit.

Fig. 360. Rome. Portion of enclosing wall of cemetery
near Sant' Agnese outside the walls (625-638).

Fig. 361. Rome. Villa called " Sette Bassi."
(Ilnd Century).

Ribbed vaulting

THE RISE OF THE LOMBARDIC STYLE

247

the original state of the building. Thus, one may still see the wall-arches of the

vaulting which contained the original windows of the nave, double-splayed like those

still remaining below in

the aisles and triforium.

There, again, are the

vaulting shafts, and one

can trace the alterations

made in the external

blank arcades, as well as

the rebuilding of the upper

part of the dome and of

part of the transept at its

highest point.

In its original plan
San Michele, which forms
a perfect cross with very
elongated arms, and is
divided into nave and
aisles (the former being
over 33 feet wide) by piers
alternately larger and

smaller, had its nave, aisles, and galleries covered with raised cross vaulting in square
bays ; barrel vaults for the arms of the cross and the spacious presbytery ; a half-
dome for the apse ; and an octagonal cupola resting on Lombardic compound
pendentives over the crossing. This cupola is the earliest specimen of a Lom-
bardic dome of considerable size, in its completed and elaborated form, existing
either in Italy or in the countries north of the Alps.

Fig. 362. Rome. Villa called " Sette Bassi." Ribbed vaulting
(Ilnd Century).

We are not to imagine that the credit
Lombardic organism, taken individually,
Masters. As a matter of fact :

(i) The Babylonians were acquainted

m--.-

1 S s- ^

Fig. 363. Rome. Villa called " Sette Bassi." Detail
ribbed vaulting (Ilnd Century).

of inventing the essential elements of the
belongs to the Lombard and Comacine

with buttresses, as is shown by the remains
of the well-known temple at Mugheir,
where the face of the walls is broken
from point to point by the buttresses
which give it support. The Romans
afterwards developed them on statical
principles (Figs. 357, 358, 359), placing
them in relation to the vaulting and
the arches in the interior. They also
sometimes disposed them simply be-
tween the openings of arches and
windows. The builders of Rome and
Ravenna and their immediate Italian
descendants shaped them in different
ways : as broad pilasters, continuous,
graduated or stepped (Fig. 360) ; and as

248

LOMBARDIC ARCHITECTURE

Fig. 364. Constantinople. Convent church of Myrelaion (919-945).

a rectangular or ramping wall
pierced by an arch.

(2) In the case of compound
piers formed to support several
arches, and also of cruciform
piers used as the starting point
of longitudinal and transverse
arches and groins of vaulting,
the Romans were the first to
employ them.

(3) Raised concave-crowned
cross vaulting had been used by
the Ravennate builders of the
Vlth century in San Vitale.

(4) Diagonal ribs were an
invention of the Roman builders,
who used them in their inter-
secting vaulting, not merely in-
corporated with the masonry, as
has been universally believed,
but also standing out and visible,

as I have discovered. Ribs of this kind, viz. visible and at the same time incorporated
with the vaulting, may still be
seen in the substructures of the
villa known as " Sette Bassi,"
where one room of about 25 feet
square, belonging to the reign
of Hadrian as shown by the
brick stamps, is covered by an
intersecting vault, the cells of
which rest on prominent massive
diagonal ribs of rectangular sec-
tion (Figs. 361, 362). These ribs
are made of compartments in
brickwork filled in with rubble
(Fig. 363) and die away at the
angles into a triangular point.
They gradually increase in breadth
till at the point of intersection
they measure about 2 ft. 4 in.
across. The cells are formed of
a layer of tufa lumps set by hand,
with the concrete backing above
it. Both cells and ribs were
originally plastered.

The Lombard gilds, how-
ever, deserve the credit of having
given to almost all these ele-

/ Fin. i6<;. Constantinople. Convent church of Myrelaion

ments new forms, new functions, (919-945).

THE RISE OF THE LOMBARDIC STYLE

249

new uses ; and of having combined them in a system providing both for the equi-
librium and the decorative effect of the building, different from the preceding systems,
and absolutely original. And in this way they not only initiated what Viollet-le-Duc *
calls one of the most complete and logical revolutions with which we are acquainted
in the domain of architecture, but they also created a rich and varied scheme of
decoration which might have been made on purpose to prove that the most effective
method of expression in architecture is to be found in a frank and intelligent revela-
tion of the structure.

The Lombardic organism had no predecessors in the Western world. And this
isequally true of the East. In the
four centuries before the epoch of
IOOO, the vaulted churches built
by the Greeks were more or less
modelled, at first on the old
Byzantine type of the age of
Justinian, and afterwards on that
type combined with the fashions
prevailing under Leo III the
Isaurian (717-740) and Constan-
tine V Copronymus (740-775).

The well-preserved church
of the Convent of Myrelaion
at Constantinople (919-945)
(Figs. 364, 365), which affords
a very rare example in the
East of external buttresses
placed in relation to the piers
of the nave and cupola, tells
us what was the constructive
scheme employed by the Greeks
at the time when the Lombardic
system was being evolved in
Italy. And when that system
came forth into the light of
day, the Byzantine builders still
held fast to the models we have Fig. 366. Salonica. Church of the Virgin (1028).

described, though they intro-
duced a new scheme of architectural decoration for the exterior of their buildings
and modified the external form of their domes, the drum of which, under the
potent influence of the Lombardic School, became polygonal, while the cupola, by
an original treatment, had its continuous spherical surface broken into convex
sections corresponding to the curved spaces below. The prototype of this last
Byzantine form is the church of the Virgin at Salonica (Fig. 366) built by Christopher,
" protospatharius " and " katepan " (or chieftain) of Langobardia, together with his
wife and their sons (1028), as is stated in the well-known inscription. This building,
I consider, is important for the history of ecclesiastical architecture in the East ; for,
with its characteristics as guide, it would not be difficult for one who was at home
in the subject to classify chronologically (approximately it may be) a number of

1 Diclionnairt raisonnt dc t architecture fran$aisc du XI' au XVI* sticle.

2 5 o LOMBARDIC ARCHITECTURE

similar churches in those countries, which are not dated, but have sometimes been
assigned to wildly imaginary periods.

# * *

Here ends the first part of this book. In the second and last part we shall cross
the Alps, and with the aid of historical proofs and of the buildings themselves we
shall see what is the truth about the origin of the styles of architecture derived from
the Lombardic, which flourished there in the Xlth and Xllth centuries.

RICHARD CLAY AND SONS, LIMITED,

BREAD ST. HILL, B.C., AND

BUNGAY, SUFFOLK.

NA Rivoira, Giovanni Teresio

1119 Lombardic architecture

L8R513

PLEASE DO NOT REMOVE
CARDS OR SLIPS FROM THIS POCKET

UNIVERSITY OF TORONTO LIBRARY