the dome: a study in the history of ideas

7/22/2019 The dome: a study in the history of ideas

1/220

A STUDY IN THE HISTORY OF IDEASBYE. BALDWIN SMITH


2/220


3/220

DATE DUE

Q726.5 S646dSmith, E. Baldwin (EarlBaldwin), 1888-1956.The dome, a study in thehistory of ideas.19iO


4/220

PRINCETON MONOGRAPHSIN ART AND ARCHAEOLOGY

XXV

BARR FERREE FOUNDATION

PUBLISHED FOR THEDEPARTMENT OF ART AND ARCHAEOLOGY

PRINCETON UNIVERSITY


5/220

THE DOMEA STUDY IN THE HISTORY OF IDEASBY E. BALDWIN SMITH

PRINCETON, NEW JERSEYPRINCETON UNIVERSITY PRESS


6/220

PUBLISHED BY PRINCETON UNIVERSITY PRESS,41 WILLIAM STREET, PRINCETON, NEW JERSEY 08540

IN THE UNITED KINGDOM: PRINCETON UNIVERSITY PRESS, GUILDFORD, SURREYCopyright, 1950 by Princeton University Press,

copyright renewed 1978 by Princeton University PressAll rights reserved

First Princeton Paperback printing, 1971LCC 75-160543

ISBN 0-691-00304-1ISBN 0-691-03875-9 (pbk.)Clothbound editions of Princeton University Press books are printed onacid-free paper, and binding materials are chosen for strength and durability.Paperbacks, while satisfactory for personal collections, are not usually

suitable for library rebinding.

PRINTED IN THE UNITED STATES OF AMERICABY PRINCETON UNIVERSITY PRESS, PRINCETON, NEW JERSEY


7/220

TO THE MEMORY OFBARR FERREE

Merely because this is the first monograph to be published with fundsof the Barr Ferree Foundation is not the reason that the author takesso much pleasure in dedicating this book to the memory of Barr Ferree.The real incentive comes from having known him and his great interestin the history of architecture.

This cultivated and learned gentleman devoted much of his life toan appreciation of the arts and a systematic study of the Gothic cathe-drals which he particularly admired. Born in 1862, he graduated in1884 from the University of Pennsylvania where for some years heserved as a special lecturer on architectural subjects in its new Schoolof Architecture. Before entering business he became known as alecturer and, as a result of his addresses delivered at the BrooklynInstitute of Arts and Sciences} he was made President of its depart-ment of Architecture and the Fine Arts. In the course of years hisarticles and interest in cataloguing the buildings of architecturalmerit everywhere in the world resulted in his being the first Amer-ican writer to be elected to honorary membership in the Royal Insti-tute of British Architects.

All his life, even after he became a successful businessman, he con-tinued to study the arts and compiled an extensive catalogue of theFrench cathedrals. At the same time he gathered a fine and rare col-lection of books on mediaeval churches and towns in France, whichnow forms the nucleus and chief ornament of the Barr Ferree Libraryat Princeton. During the First World War he gave eloquent expres-sion to his wrath when his beloved cathedrals were attacked andseriously injured. In New York he organized the Pennsylvania Society,now the largest of the state societies in the United States, and fromits foundation in 1899 until his death in 1924 he was its Secretary andDirector. Because of the patriotic work of the Society and his personalefforts during the war, he was decorated in 1922 with the Grand Crossof the Legion of Honor.His death, a sudden one, 'occurred October 14, 1924. In accordancewith his wish, however, his estate was converted into a Foundation forthe publication of books on architecture and related topics in the FineArts. Thus, the stimulation of intelligent cultivation of the Fine Arts,,toward which the sustained effort of his life was directed, now liveson, working through the avenues whereby a University influences itsstudents and the public. The dedication of his books and property tothe cause of the Fine Arts in America has become what he wanted

a lasting memorial of the ultimate indestructibility of hisintellectual purpose and spiritual conviction.


8/220


9/220

PREFACETHE admission that this book falls short of what it was intended to be is merely astatement of fact and not an apology. There was a time when the author opti-mistically believed that he could present the major aspects of domical ideologyand evolution in one study. That, however, was before the rapidly expanding com-plexities of the subject and the difficulties of organizing the material in a writtenform where the ideas would not be reburied under a mass of accumulated evidencehad become inescapable factors. Once it had become evident that the dome was notjust a utilitarian form of vaulting, which had originated for structural and environ-mental reasons in some one country, but was primarily a house concept, which hadacquired in numerous cultures its shape and imaginative values upon an ancestralshelter long before it was translated for ideological reasons into more permanentand monumental form by means of wood carpentry and masonry, the whole problemof the dome opened up into a comprehensible but infinitely complex chapter In thehistory of ideas.

After the broad outlines of this evolution from the primitive house had beentraced in the various ancient and retarded cultures of Europe, Asia, Africa, and theAmericas there arose the disquieting question of to what extent one scholar had thetime and equipment to reconstruct the whole development of domical beliefs. Thematter of time was settled conclusively when Karl Lehmann's The Dome of Heavenshowed that no one could expect to enjoy indefinitely a monopoly of domical ideas.The other question remains to be tested now that the scaffolding has been removedand The Dome in skeleton form has to stand alone. Since so many of the conclusionsare contrary to prevailing opinions, a partial study of the dome, which at least pre-cipitates the major issues, has the advantage of testing out the basic method of approachbefore it is applied to such controversial aspects of domical evolution as the originof the Iranian dome and the still more delicate question of whether even ancientGreece did not have its own tradition of a symbolic, wooden dome.

It is difficult to imagine how certain portions of this study could have been writtenif it had not been for the assistance and cooperation of Glanville Downey, whose wideknowledge of Byzantine literature and Greek architectural usage has made it possibleto base much of the essential evidence on the texts. Another contributing factor ofgreat importance, since it necessitated rewriting much of the manuscript, was thepublication of Andre Grabar's Martynum. Had the author's indebtedness been lim-ited to the Martyrium the references in the text might

have been an adequate ac-knowledgement. It was when Grabar read his last chapter and encouraged him topublish it, even though it advanced an explanation for the Syrian bema which wasquite different from the one Grabar had published, that the author became indebtedto the man himself and came to appreciate the generosity of his fine scholarship. Thefact that A. M. Friend has listened patiently to the mutterings of a dome-obsessedmind, has read the manuscript and endeavored to protect the author from the dangers

vn


10/220

PREFACEof Byzantine liturgies, does not, of course, make him responsible for the unorthodoxapproach to some of the problems. The author is indebted to Mrs. Estelle Brown forher help in preparing the manuscript and to Miss Rosalie Green for her scholarlycare in checking the references.

E. BALDWIN SMITHPrinceton UniversityApril 1949


11/220

CONTENTSPREFACE ViiI. DOMICAL ORIGINS 3II. THE USE OF THE WOODEN DOME IN THE NEAR EAST 10

1. Marneion, Gaza 142. The Holy Sepulchre, Jerusalem 163. Domus aurea, Antioch 294. Martyrium, Nazianzus 315. Martyrium, Nyssa 316. Martyria, Constantinople 327. S. Simeon Stylites, Kal'at Sim'an 348. Sion Church, Jerusalem 369. Cathedral, Etschmiadzin 37

10. S. Stephen, Gaza 3811. S. Sergius, Gaza 3912. Church, Mahoymac 4013. Church, Ba'albek 4114. The Islamic Wooden Dome 41

III. THE MASONRY DOME AND THE MORTUARY TRADITIONIN SYRIA AND PALESTINE 45

A. THE BRICK DOME 46B. THE STONE AND CONCRETE DOME 47C. THE DOMICAL MORTUARY TRADITION IN SYRIA 50

IV. DOMICAL FORMS AND THEIR IDEOLOGY 61A. THE ANCESTRAL SHELTER: QUBAB HUT AND KALUB 61B. THE SACRED KALUB 67C. THE MONUMENTAL KALUBfi OF MASONRY 70D. THE CONOID BAETYLS AND THE ANCESTRAL HOUSECONCEPT AS A MANIFESTATION OF DIVINITY 71E. OTHER SACRED AND CELESTIAL ASPECTS OF THEDOMICAL SHAPE 741. The Omphalos 752. The Cosmic Egg 773. The Celestial Helmet 77

F. THE COSMIC HOUSE 791. Indian Tradition 802. Asiatic Tradition and the Imperial Baldachin 81

ix


12/220

CONTENTS3. Pre-Islamic and Hebrew Tradition 834. Early Christian and Byzantine Tradition 85

V. DOMICAL CHURCHES: MARTYRIA 95A. CIRCULAR 98B. POLYGONAL 100C. SQUARE 105D. CRUCIFORM 108E. INSCRIBED CRUCIFORM mF. FOUR-LOBED, CRUCIFORM 115G. TRI-LOBED 120H. RECTANGULAR 124I. SUMMARY 131

VI. THE PLACE OF COMMEMORATIONA. THE PROBLEM 132B. THE MONUMENTS 1351. Domed Martyria 1352. Basilicas 139

C. THEORIES 1411 . An Altar? 1 4 12. An Ambon? 1413. Ambon, Choir and Mass of Catechumens? 144D. THE EVIDENCE OF THE TESTAMENTUM 145E. THE PLACE OF COMMEMORATION 147F. DOMICAL CHAPELS 151

APPENDIX: Description of the Church of S. Stephen at Gaza byChoricius, Sections 37-46 Translation and Notes by G. Downey 155

INDEX


13/220

THE DOME


14/220


15/220

I DOMICAL ORIGINS

How the dome took shape, where it originated, and why it became the outstand-ing feature of Byzantine and Islamic architecture are questions which havenot been satisfactorily answered either by the Orient oder Rom controversyor by the misconceptions implicit in the prevailing theories regarding the origin andpurpose of the domical shape.

Ever since the nineteenth century it has been generally believed that the dome,from its inception, was a functional means of vaulting which originated for environ-mental reasons either in the brick architecture of the Orient or in the masonry con-struction of the Romans. This effort to trace the dome back to a single place of originhas disregarded the Syro-Palestinian region as a country which should have playedan

importantrole in the

development and spread of a Christian domical tradition,because the early explorers, such as De Vogue and Howard Butler, found so few extantdomes among the Syrian ruins and because more modern excavators have uncoveredno traces of masonry domes on central-type churches before the sixth century,

Even if it were true that the dome had started only as a utilitarian form of roofingand had a neat unilateral development as it spread from Mesopotamia, Persia, or someHellenistic center in the Roman Empire, it would still be necessary to explain, first,why men had come to build such curvilinear shapes; second, why they came to asso-ciate the dome in pagan, Christian and Islamic periods with tombs, memoriae,aediculae, tabernacles, ciboria, baldachins, martyria, baptisteries, churches, fire tem-ples, mosques, and audience halls; third, why the East Christians during the fifth andsixth centuries began to manifest so much interest in the dome as a form of churcharchitecture; and fourth, why the Islamic builders elected to make the dome thedominant feature of their tombs and mosques. All of these questions can be answeredin whole or in part by a study of the domical ideology which prevailed in the lateantique and Early Christian world and by the existing evidence for the use of thewooden dome in Syria and Palestine.Between 1935 and 1939 the Princeton Excavations at Antioch-on-the-Orontes un-

covered the plans of two important Early Christian churches of the central typethe fourth century martyrium at Kaoussie (Fig. 170), a suburb of Antioch, and thefifth century martyrium at Seleucia Pieria (Fig. 182), the port of Antioch. At the sametime the excavators of Gerasa disclosed other churches of the central type. Since it hasbecome apparent that in the cities of Syria and Palestine the churches were far frombeing all of the basilican type, there arise the questions of how these central-typemartyria and churches were roofed and what part they played in the growth of adomical, Christian architecture in the Near East. As long as the dome is thought ofonly as one kind of functional roofing, it is of relatively little importance whetherthese Syrian and Palestinian cult houses had pyramidal, conical, or domical roofs ofwood or of light volcanic scoriae, and there is little justification for devoting a wholemonograph to a study of the dome and a hypothetical restoration of two Antiochenechurches,

3


16/220

DOMICAL ORIGINSActually, however, the dome was of great symbolical interest to the Christians.

It was a shape which, regardless of the materials used for its construction, had anantique sepulchral association with memorials to the dead and a long and highlycomplicated history in various parts of the antique world. Not only is it possible toshow that the dome was ideologically an essential feature of the central-type martyr-ium, but it is also possible to demonstrate why the two Antiochene churches, andother similar Syro-Palestinian churches, must have been roofed with great woodendomes of conoid shape, sheathed in gilded metal, which had a symbolical content inthe Christian thought of the period. It is also becoming increasingly evident that thesacred buildings of the Holy Land, with their mosaics, frescoes and illuminatedgospels, had a powerful and lasting influence upon all forms of Christian art. Conse-quently it is difficult to believe that the popularity of the dome on the religiousarchitecture of Byzantium, Armenia, and northern Mesopotamia was not influencedby the revered churches of Palestine.The necessity of reexamining the prevailing conclusions regarding the origins ofByzantine architecture, and the value of relating the development of domical archi-tecture in the Near East to the history of ideas, are indicated by the fact that theresults of a study of domical ideas so closely parallel the conclusions of Andre Grabarin his Martyrium. Although Grabar has not attempted to deal with the problems ofthe dome in tracing the growing popularity of the martyrium-type church, the domewas peculiarly associated with the martyrium because of its traditional mortuarysymbolism. Therefore, because of this relationship and the bearing which some ofthe domical evidence has upon the pattern of development outlined by Grabar, itis desirable to integrate the two approaches.

It is Grabar's conclusion that the mortuary implications of the Cult of Relics, theprestige of the martyrium churches of Syria and Palestine, and the spread of Syriansymbolism regarding the meaning of the House of God exerted a widespread influ-ence throughout

the Near East, including Mesopotamia and Armenia, both in popu-larizing the central-type martyria and in transforming them into churches of theregular cult.1 This spread of the martyrium concept, he says, coincided with the dis-semination of the Areopagitica throughout the Greek world. By the sixth century,then, when the Syrian churchmen were insisting upon the idea of the church as a mys-tic temple, a replica of the comprehensible universe, it becomes more apparent why thedome, which the Christians had taken over from pagan mausolea and commemorativemonuments, should have become popular, because for centuries the dome had beena symbolic form of varied but related meanings. In explaining the growing popularityof the martyrium-type church, Grabar derives all the basic forms of martyria fromthe pagan tombs, memorials, and heroa of Rome and does not, because of the magni-tude of his investigation, attempt to deal with the other sources of domical ideologywhich were involved in the development of domical architecture. It is to be hoped,

1 A. Grabar, Martyrium, recherches sur le culte des reliques et I'art chretien antique, Paris, 1946.


17/220

DOMICAL ORIGINStherefore, that a partial history of domical concepts and the evidence for the earlyuse of the wooden dome will show that Syria and Palestine had a native domicaltradition which not only readily combined with the Roman and Hellenistic traditionsof a mortuary dome, but also account for certain specific types of free-standing domesand help to explain why the dome, as the manifestation of an idea, became so im-portant in Byzantine and Islamic architecture.

Behind the concepts involved in domical development was the natural and per-sistent primitive instinct to think in terms of customary memory images and to at-tribute actual being and inner power to inanimate objects, such as the roof and otherparts of the house. To the naive eye of men uninterested in construction, the dome,it must be realized, was first of all a shape and then an idea. As a shape (which ante-dated the beginnings of masonry construction), It was the memorable feature of anancient, ancestral house. It is still a shape visualized and described by such terms ashemisphere, beehive, onion, melon, and bulbous. In ancient times it was thoughtof as a tholos, pine cone, omphalos, helmet, tegurium, kubba, kalube, maphalia,vihdra, parasol, amalaka tree, cosmic egg, and heavenly bowl While the modernterms are purely descriptive, the ancient imagery both preserved some memory ofthe origin of the domical shape and conveyed something of the ancestral beliefs andsupernatural meanings associated with its form.A key to the origin of the domical shape as a house concept is furnished by thederivation of our modern word dome from the Greek and Latin domus. In Middleand Late Latin doma meant house/* roof, and only at times cupola/' while duringthe Middle Ages and the Renaissance it was used all over Europe to designate a re-vered house, a Domus Dei. This persistent association with the idea of an importanthouse, which will be seen going back to the first beginnings of domical architecture,survived in the Italian duomo, the German, Icelandic, and Danish Dom 9 meaningcathedral, and as late as 1656 in the English dome meaning Town-House, Guild-Hall, State-House, and Meeting-house in a city.

2 For centuries, apparently, dome wasapplied to any outstanding and important house, sacred or otherwise, which mightor might not have had a cupola roof. During the seventeenth century, however, theoriginal meaning began to fade into poetic usage and by 1660 dosme in France hadacquired the specific meaning of a cupola vault, which in the course of the eighteenthcentury became standard usage for dome in English. This gradual limitation of mean-ing was partly the result of the growing scientific need for technical terms, but largelybecause the eighteenth century, in its admiration of such churches as S. Peter's inRome, S. Paul's in London, Les Invalides and the Pantheon in Paris, as well as scoresof other domical structures, still considered the monumental dome as the designatingfeature of all truly impressive houses of God, and hence synonymous with domus.What is revealing in this derivation is that even in English the idea of a domebegan as a house concept, just as in ancient Italy, Syria, India and Islam words for

2 New English Dictionary, Oxford, 1897; T. Bloimt, Glossographia . . . , London, 1656, dome.5


18/220

DOMICAL ORIGINShouse, tent, or primitive shelter, such as tegurium, kalube, vikdra, and kubba, cameto designate a dome or domical structure.

It is impossible within the limits of a study of the domical tradition in Syria andPalestine to trace all the beginnings of domical shapes

and domical ideas in the dif-ferent countries of antiquity and to note their parallels in the retarded primitivecultures of Africa, Asia and the Americas, Instead, a series of already carefully investi-gated postulates, which can at least be checked against the evidence for the origin ofthe domical ideas of Syria, is advanced.

A. The domical shape must be distinguished from domical vaulting because thedome, both as idea and as method of roofing, originated in pliable materials upon aprimitive shelter and was later preserved, venerated, and translated into more perma-nent materials, largely for symbolic and traditional reasons.

1. At the primitive level the most prevalent and usually the earliest type of con-structed shelter, whether a tent, pit house, earth lodge, or thatched cabin, was moreor less circular in plan and covered by necessity with a curved roof. Therefore, inmany parts of the ancient world the domical shape became habitually associatedin men's memories with a central type of structure which was venerated as a tribaland ancestral shelter, a cosmic symbol, a house of appearances and a ritualistic abode.

2. Hence many widely separate cultures, whose architecture evolved from primi-tive methods of construction, had some tradition of an ancient and revered shelterwhich was distinguished by a curved roof, usually more or less domical in appear-ance, but sometimes hoop-shaped or conical.B. This domical shape, as an ancient and revered house form, was preserved in

many cultures and gradually translated into more permanent materials as a familyor royal tomb, a cult house and abode of the Great One, or as a utilitarian granary,sweat house and kiln.

1. Because of the animistic habits of thought which continued to attach innermeaning and magical power to the memorable shape of the ancestral round shelter,most early civilizations had deeply rooted domical ideologies which resulted inthe veneration of the domical shape as a mortuary, sacred, royal and celestial abodeof the Great One long after the ordinary domestic architecture had become rec-tangular with flat or gabled roofs.

2. This tendency was strengthened by the primitive habit of visualizing boththe cosmos and divinities in the shape of the ancestral house.G. Therefore, there is no historical justification for the assumption that the domical

vault originated for purely structural and environmental reasons in either brick orstone. Instead, all the evidence shows that early vault forms, like the dome and thetunnel vault, were traditional roof shapes originating in pliable materials and laterimitated in masonry for ideological reasons.

i . Hence, in tracing the evolution of the dome in any particular region, a dis-tinction must be made between the cultural level when the domical idea took shapeand acquired symbolic values and the historical period when there was a social


19/220

DOMICAL ORIGINSorganization with the incentive, technical equipment, and craftsmen to translatean ancestral dwelling into a tholos tomb, to turn a royal tent into a domical audiencehall of brick and to erect monumental hemispheres, or conoid domes, upon temples,martyria, palaces, churches, baptisteries and mosques.

2. Moreover, the dome, like any other curvilinear form such as the horseshoearch, could not have originated in cut stone, because rock is shapeless and the imagehas to exist in the mind of the stonecutter. Stone architecture the world over, fromIndia to Stonehenge, began as an imitative and sculptural effort on the part of or-ganized society to reproduce venerated forms which had formerly been constructedin more pliable materials.D. There were various domical traditions in both the West and the East.

1. It was the mortuary, divine, royal and celestial meanings of these domicaltraditions with their symbolic ideologies which led to the popularity and monu-mental use of the domical shape in India and the late Roman Empire, then in theChristian and Sassanian East, and later in the Islamic Empire.

2. Because the conception and meanings of the domical shape were primarilyderived from primitive habitations, many cultures had domical ideologies beforethey had domical vaults of masonry. Even after some cultures developed or acquireda monumental architecture with temples, palaces and churches of stone and brick,they religiously preserved the shape, and often the ancient construction, of ancestraland ritualistic shelters for their inner sanctuaries, tabernacles, aediculae, ciboria andbaldachins.E. At the same time that the dome was taking on utilitarian values as a vault upon

granaries, baths (sweat houses) and kilns which in the beginning were special adapta-tions and survivals of the primitive round house the domical shape, regardless of itsconstruction, acquired persistent symbolic values.

1. It was the mortuary, divine, royal and celestial meanings of these domicalideologies which, in different civilizations and at different periods, furnished theincentive to translate the idea of an ancient tentorium, kalube, maphalia, tegurium,vihdra, and kubba into a monumental structure of wood, brick, stone, or concrete.

2. The process, however, was never wholly independent and indigenous in anycivilization. The formation of domical architecture in the Roman Empire, India,the Christian and Sassanian East, and the Islamic Empire was the result of an in-tricate fusion of various domical traditions and a multilateral dispersion of struc-tural methods of building domical shapes.

3. By the late Roman Empire, when the dome was acquiring so much distinctionas a symbol of celestial greatness and imperial immortality, its ideology was furtherenriched by the popular ideas already associated with similar shapes, such as thetholos, mundus, heroon, sacred baetyl, omphalos, divine helmet, umbrella, cosmicegg and pine cone, also by the interests of the Orphic cults in a celestial cosmogonyand a heavenly salvation, and by the introduction of ancient Indian beliefs regard-ing the cosmic significance of the dome.

7


20/220


21/220


22/220

II * THE USE OF THE WOODEN DOMEIN THE NEAR EASTBEFORE

discussing the historical evidence for the early use of the wooden domein the Near East, some consideration should be given to the theoretical advan-tages of the wooden dome and to the question of why a domical style of archi-

ture should have developed apparently so rapidly in Asia Minor, Byzantium andSyria regions which suffered such severe and recurrent earthquakes. If the domewas only a practical form of masonry vaulting, why was it adopted in regions wherethe heavy masonry dome was more difficult and dangerous to construct than gabledand flat roofs?

Because of its disruptive thrusts the masonry dome was in constant danger of col-lapsing if supporting walls, piers or buttressing were disturbed. When built largeand imposing, it required massive supports and buttressing up to its haunch, whichmade it impossible or at least very dangerous to make its domical shape fully visibleupon the exterior. The wooden dome, on the contrary, was light and could safely beraised up large, free standing and of wide span on relatively thin walls and highclerestories. Once constructed, its rigid framework exerted relatively little thrust.Furthermore, in wood carpentry the builders could easily and safely reproduce ona large scale the curved profiles of those conoid and bulbous shapes which were im-practical in masonry but which had become customary and symbolically significantupon the traditional ritualistic shelters. Once completed, the wooden dome could beprotected and made resplendent, like the celestial symbol that it was, by a gilded metalsheathing.And yet, in spite of the advantages of the wooden dome, we find in some countrieswhich started with the tradition of wood carpentry and where the masonry domeswere in danger of being destroyed by the recurrent earthquakes an evident desireor necessity to translate the domical shape into masonry. In those regions where timberwas becoming scarce through deforestation and limited transportation facilities itis easily understood why the wooden dome was abandoned. At the same time therewas in the antique world the conviction, regardless of fact, that masonry constructionwas in itself enduring and a mark of superior greatness. Roman interest in themechanics of vaulting, and imperial ideas of a state architecture of solid and enduringmasonry, had perfected domical construction and introduced it into various partsof the empire. The masonry dome had come to be a mark of royal and divine power.Since all large churches were built with the approval and assistance of the State, theChurch was strongly influenced by imperial building methods, and the masonrydome of stone, brick and concrete became more and more common. At the same timethe wooden dome continued, in Syria at least, to be so characteristic of the churcharchitecture that it was taken over by the Arabs as the distinctive feature of theirmosques.

Individual scholars have recognized and discussed the use of the wooden dome10


23/220

WOODEN DOME IN THE NEAR EASTin the ancient and Early Christian periods and its later importance in the domicalarchitectures of India, Russia and the Islamic world. 1 Nevertheless, the significanceof the wooden dome in the early evolution of domical styles has been largely disre-garded, partly, it would appear, because it has left so little archaeological evidence,but largely because historians of architecture have been schooled to the belief thatthe wooden dome was a derivative form of construction, imitating masonry vaults. Aslong as the dome was considered to have been only a kind of vaulting essentiallypeculiar to the exotic art of the Orient, as it was known in the nineteenth century,and the wooden roofs upon the circular and polygonal buildings of Greece and theHellenistic world were thought to have been always conical rather than domical,scholars were unable to reverse their reasoning and assume that domical traditionsof wood were not only earlier than those of masonry but were also developed inde-pendently in different parts of both Europe and Asia. The Etruscans certainly knewthe wooden dome, as did the later Roman builders, and in the Saar basin of theGermanic North it can be shown that the domical shape was used upon houses, tombs,temples and city towers and was not translated into masonry form until after theRomans came to dominate the country. It is also possible that domical houses andtemples of wood carpentry were common from an early period along the bordersof the Black and Caspian Seas from the Danube to Armenia and the Caucasus beforecontacts with the Byzantine and Sassanian Empires introduced the organized laborand technical means of reproducing the traditional domical shapes in more perma-nent masonry. It is only in India, however, that the importance of wood carpentryin the development of domical architecture has been fully recognized: 2 in a studyentitled Origin and Mutation in Indian and Eastern Architecture/' William Simp-son outlined the stages by which arcuated and domical forms developed from primitivehabitations of bamboo and thatch and were later translated into cut stone.a In fact,Simpson laid down the premise that wood carpentry was one of the necessary stepsbetween the first origin and its full development in stone,

In 1913, when Birnbaum endeavored to demonstrate from literary sources that1 H. Thiersch ( Antike Bauten fur Musik, 3 A. K. Coomaraswamy (History o/ Indian

Zeitschrtft -fiir Geschichte der Architektur, n, and Indonesian Art, 1927, 49) wrote, Practi-1908/9, 33ff.) restored the tholos of Epidauros cally, it can hardly be doubted that, as in otherwith a wooden dome whose construction re- countries, the form of the god's house iscalled the carpentry domes en parasol repro- derived from that of human dwellings andduced in a number of roek-cut Etruscan tombs tombs, the main source leading back to the(J. Martha, L'Art etrusque, 1889, 156, fig. 124; domed thatched hut and the barrel vaultedG. Dennis, The Cities and Cemeteries of types of the Todas ; and in tracing the originEtruria, 1878, r, 239, 274, 448; K. Lehmann, of the domical vihara and fire temple, A.The Dome of Heaven, Art Bulletin, xxvn, Foucher (L'Art greco-bouddhique du Ghand-1945, 20 n. 176). Other references to the use hara,i, 1905, 128) pointed out that the domicalof the wooden dome in Greece: P. Cawadias, forms went back to the primitive round hutLa Tholos d'pidaure et le peinture Pausias, and were constructed in wood long before theyMelanges Nicole, 1905, 61 1; H. Pomtow, Die were translated into stone.Grosse Tholos zu Delphi, Klio, xir, 1912, 3 W. Simpson, R.LB.A. Transactions, vn,2l6ff. 22gff,

11


24/220

WOODEN DOME IN THE NEAR EASTthe octagonal churches of Antioch, Nazlanzus and Nyssa had wooden roofs, he dis-regarded the literal meanings of the texts and insisted that their wooden roofs wereeither conical or pyramidal in shape because, as he wrote, a wooden dome is a para-dox. 4 Strzygowski, although he repeatedly pointed out the possible influences ofwooden prototypes upon the stone architecture of Armenia and suggested certainwooden derivations for the Asiatic adjustment of the dome to a square plan, alwayscame back to his undemonstrated conviction that the dome itself originated in thebrick architecture of Mesopotamia and Iran. 5 As early as 1921, however, Herzfeldinsisted that the Early Christian churches of Syria had wooden domes which weretaken over by the Arabs for their mosques.6 Later, K. A. C. Creswell in his EarlyMuslim Architecture undertook to trace the Islamic wooden dome back to its probableprototypes in Syria. 7 Having shown that the early Arab domes of wood must havebeen built by Syrian workmen and presumably represented the continuation of aSyrian tradition, Creswell apparently disregarded the logic of his own arguments,considering the evidence as somewhat ambiguous and unsatisfactory/* and invariablycame back to the traditional conclusion that the Syrian and Palestinian churches ofthe central type must have had wooden roofs of conical rather than domical shape.By 1935, however, C. Watzinger, without discussing the evidence, fully accepted

the existence of the wooden dome as the prevailing type of roof upon the circular,polygonal and cruciform churches of Syria and Palestine. 8 Somewhat later, the ex-perienced archaeologist D. Krencker, after careful study, restored the great octagonof S. Simeon Stylites at KaFat Sim'an with a pointed and bulbous dome of woodenconstruction. 9 In 1943 W. Born, in an article on the history of the bulbous dome,made no attempt to trace the wooden dome back to its early origins and was reluctantto admit that its bulbous form could have antedated the Islamic period in the archi-tecture of Syria. 10 At the same time he advanced the conclusion, which seems to re-quire further explanation, that the bulbous dome must have evolved in the stratumof wood architecture which extended through India, the Near East and Russia andeven suggested that Syria must have led in the development. Even though he saysthat the wooden dome in Islamic architecture was at an early date translated intostone, he does not intimate that it may have had a long previous history in widelyseparate regions.

In 1945 K. Lehmann, in his study of the celestial symbolism of the dome, cited4 A. Birnbaum, Die Oktogone von Anti-

ochia, Nazianz und Nyssa, Repertorium furKunstwissenschaft, xxxvi, 1913, 181-209.

5 J. Strzygowski, Altai-Iran, 1917; Die Bau-kunst der Armenier und Europa, 1918; Diealtslauische Kunst, 1929; Origin of ChristianChurch Art, 1923, 8.6 E. Herzfeld, Mschatta, Hira und Badlya,Jahrbuch der preussischen Kunstsammlungen,XLII, 1921, 120-121.

7 K. A. C. Creswell, Early Muslim Architec-

ture, 1932-40, i, 83-87.8 C. Watzinger, Denkmaler Palastinas, 1933-35, ii, 131.

9 D. Krencker, Die Wallfahrtskirche des Si-meon Stylites in Kal'at Sim'dn (Abhandlungender preussischen Akademie der Wissenschaf-ten, Phil. hist. Klasse, 1938, no. 4), 1939.10 W. Born, The Origin and Distributionof the Bulbous Dome, Journal of the Ameri-can Society of Architectural Historians, in,1943,

12


25/220

WOODEN DOME IN THE NEAR EASTmuch of the evidence for the Importance of the wooden dome In Etruscan and Romanarchitecture and suggested that it may have been used In ancient Greece, 11 The fol-lowing year Herzfeld wrote: 'In Syria wood was throughout antiquity the specificmaterial for ceilings and said that the great Syrian domes of Bosra, Jerusalem andDamascus, all of wood, were constructed as double cupolas with an elaborate systemof girders and ribs which was the result of experience acquired In shipbuilding.12Before the central-type churches of Antioch can be restored and properly relatedto the history of the domical tradition in Syria from pagan times down to Its adoptionby Islam, it is necessary to review the evidence for the use of the wooden dome. Atthe same time it is essential to note to what extent the dome was primarily the dis-tinguishing feature of the mortuary shelter and the martyrium. With the growthof the ritual of the dead and the architectural elaboration of the simple provisionsfor the mensa martyrium into a commemorative monument and church, the symbol-Ism of the primitive shelter and house of the dead was extended to the whole church:the dome gradually became the manifest symbol of the martyrium. In Syria thisassociation of the dome with the martyrium and its transformation from a symbolicshape into a monumental form of architecture was first brought about by wood car-pentry, which made it possible to construct In the architecture itself a sepulchralciborium, a royal baldachin, a divine form and celestial symbol over an altar, throne,tomb, pulpit and baptismal font.

Early archaeologists and many later students of Early Christian architecture mini-mized the importance of the dome in Syria and Palestine because they found so fewremains of masonry vaults on churches of the central type, and they disregarded thepossibilities of domical construction in wood because so much of Syria and Palestinewas thought to have been barren and tlmberless. Even the acknowledged use of greattimbered roofs on both the pagan and Christian temples and the knowledge that finetimber was plentiful in the Lebanese mountains, while forests are known to haveexisted north of Hebron and in the region of Lake Tiberius, did not outweigh theexisting testimony of stone roofs in the Hauran and the unquestioned convictionthat the dome was primarily a form of vaulting. The existence of forests and the useof timber as late as the sixth century are clearly verified by the account of Procoplus.In describing the new church of the Virgin at Jerusalem, which may have beendomical, he tells how the builders, In order to find large enough timbers for theroof, searched through all the woods and forests and every place where they hadheard that very tall trees grew and found a certain dense forest produced cedars ofextraordinary height/'

13

Ever since recent excavations have made it clear that the central churches oiAntioch, Bosra and Gerasa could not have had masonry domes, many excavators ancscholars have somewhat casually assumed, because they did not realize how mud

11 K. Lehmann, The Dome of Heaven/' tecture. Ill, Ars Islamica, xi/xii, 67.Art Bulletin, xxvn, 1945, 1-27. 13 Procopius, Buildings, v, vL12 Herzfeld, Damascus: Studies in Archl-13


26/220

WOODEN DOME IN THE NEAR EASTcontent the Christians attached to specific architectural forms, that their woodenroofs must have been either pyramidal or conical. Nevertheless, the evidence forthe use of the wooden dome in Syria and the Near East is very specific and surprisinglylarge considering how few records attempt to describe the appearance and construc-tion of churches and the fact that excavations can never reveal more than very in-direct indications as to the use of carpentry domes upon the ruined churches. Oncethe use of the wooden dome is recognized and the mystic importance of the domicalshape is understood, it becomes necessary to reexamine not only the prevailing theoriesregarding the origin and dissemination of domical beliefs but also the plans of manySyrian churches which may have had domical, rather than gabled, wooden roofs.1. MARNEION, GAZA (c. 130 A.D.)The Marneion was one of the most renowned pagan temples and the earliest knownbuilding in the Syro-Palestinian region which presumably had a wooden dome.14Since Karl Lehmann has already shown that the dome in Roman, Persian and Chris-tian architecture had a clearly recognized celestial symbolism, 15 it is significant thatthis temple was dedicated to Marnas, a sky god, who was probably a Palestinianadaptation of the Cretan Zeus, the ruler of the universe.16 The temple and its finaldestruction by fire in 402 A.D. are reliably described in the Life of Porphyry byMark the Deacon 3 who knew the Bishop of Gaza between 382 and 392 A.D. In hisdescription of the temple, Mark is very specific in saying that it was round, beingsupported by two colonnades, one within the other, and in the center was a dome(Kifi&piQv), puffed-up and rising on high, 18 According to this account, then, its domewas free standing ( rising on high ), bulbous ( swollen ), and perhaps pointed likea pine cone. That it was built of wood can only be inferred from the account whichtells of a burning beam falling from the roof upon the tribune who was supervisingthe efforts to save this most renowned center of a dying heathen world.Although most historians of the period have accepted the fact that the Marneion

had a wooden dome, puffed-up, which resembled a pine cone, some scholars have14 F. Cabrol and H. Leclercq, Dictionnaire taische Kuste, 1852, 599-600; Strzygowski,

d'archeologie chretienne et de liturgie, vi, Kleinasien, 1903, 101; Watzinger, Denk. Palas.,cols. 6955.; Pauly-Wissowa, Realencyclopadie, n, 87.xiv, cols. iSggff. Creswell (op.cit., i, 84) points out that15 Lehmann, op.cit., 1-27. there was no opening in the top of the dome16 A. B. Cook, Zeus: a Study in Ancient as Dehio, Stark, Sepp and Strzygowski inferredReligion, 1914-40, 111,549 . from the Latin translation. Although Mark

17 G. F. Hill, The Life of Porphyry, Bishop is the first writer to use the term kiborion forof Gaza, by Mark the Deacon, 1913, 75-87, 140; the dome of a building, the domical meaningCabrol, Dictionnaire, xiv, cols. 1464* .; Cres- of the word is clearly indicated by Athenaeuswell, Early Muslim Architecture, i, 83 ; G. (Deipnosoph, I, iii, 72) who describes it asDehio and G. von Bezold, Die kirchliche Bau- cuplike, by its use for a domical covering overkunst des Abendlandes, i, 1887, 36; H. Gre- either a tomb or altar (J. Braun, Der christ-goire and M.-A. Kugener, Marc le Diacre, liche Altar, 1924, n, 192), and by its later1950; J. N. and B. Sepp, Die Felsenkuppel, Byzantine use for a domical vault.1882, 46; K. B. Stark, Gaza und die philis-

14


27/220


28/220

WOODEN DOME IN THE NEAR EASTchurch mentioned by Choricius is the domical church of S, Sergius and since the latermosaics at Ma'in show the city of Gaza as characterized by a cruciform church, we mustassume that both cruciform sanctuaries were the first church of Gaza.25 If this is areasonable deduction, then it is merely unfortunate that

the Ma'in mosaic is soschematic that it only shows the three gabled arms of a cruciform plan and, becauseof the restricted space, did not undertake to depict a domical crossing. While thisidentification of S. Sergius with the Eudoxiana is still problematic, it does prove thatthe Christian builders at Gaza carried on a domical tradition which went back topagan times,2. THE HOLY SEPULCHRE, JERUSALEM (326-335 A.D.)The memorial tomb of Christ erected by order of Constantine was the most revered

sepulchral monument of Christendom. It is, therefore, of supreme importance to thehistory of architecture to know whether or not this sacred martyrium had a gildedwooden dome. If domical, this omphalos of the Christian faith, built in the centerof the New Jerusalem, would have established the type for subsequent martyria andcarried over into Christian symbolism the antique ideas of the domical shape as anancestral abode which was given to man by God, as a celestial form, a divine heroon,a ritualistic sanctuary, a royal baldachin and a cosmic house. Unfortunately, verylittle is known about the appearance and construction of the Constantinian rotunda,which was burnt by the Persians in 614 A.D. and then underwent a long succession ofrestorations and three rebuildings, one by Modestus between 616 and 618 A.D., asecond by the Byzantine Emperor Constantine Monomachus in 1048 after the destruc-tion of the building by the Fatimite Caliph in 1009, and the third by the Latincrusaders in the twelfth century. There has been no agreement as to whether theoriginal martyrium had a domical or conical roof constructed of masonry or wood:Heisenberg and Leclercq believed that it had a massive, hemispherical dome ofmasonry; Grabar, while favoring a masonry dome, entertains the possibility that itmay have been wooden; 27 Crowfoot and Watzinger are of the opinion that its domewas wood; 28 Creswell arrived at the conclusion that it had a conical roof of wood, andVincent endeavored to prove that its original dome of masonry was changed in theseventh century to a truncated cone of carpentry construction, open at the top. A fewscholars, it should be noted, have found evidence in the accounts of Eusebius andAetheria to show that the circular rotunda could not have been erected until after thefourth century. 29 For the purpose of this study it is possible to disregard their theories,more, if we accept the evidence of the mosaics was the Constantinian dome ; Watzinger,at Gerasa (Figs. 30, 31) the martyrium of S, Denk. Palas., n, 131.John the Baptist at Alexandria was domical 29 H, G. Evers ( Zu den Konstantinsbauten28 R. de Vaux, Une Mosaique byzantine a am Heiligen Grabe in Jerusalem, ZeitschrijtMa'in, Revue biblique,-XLVii, 1938, pl.xiv/2. /. agypt. Sprache u. Altertumsk., LXXV, 1935,27 Grabar, Martyrium, i, 257. 53-60) without advancing any new evidence28 J. W. Crowfoot, Early Churches in Pales- suggests that the Holy Sepulchre was an opentine, 1941, 105, The later dome over the semicircular exedra; Ejner Dyggve ( Grav-Anastasis was built of timber and so probably kirchen i Jerusalem, Studier fra Sprog-og

16


29/220


30/220


31/220

WOODEN DOME IN THE NEAR EASTcated cone with its beams showing character-ized the rotunda at Jerusalem, architecturalhistorians and numismatists have assumedthat all the representations of the HolySepulchre on the lead seals of the Crusaderswere intended to represent a conical roof. Itis necessary to question this conclusion whenthe different types of seals depicting the HolySepulchre are arranged in chronological orderand the variations in the treatment of the roofare noted. The only series which unquestion-ably depicts the building with a truncatedcone of straight timbers converging on theopening at the top is that of the Canons ofthe Holy Sepulchre (Figs. 222, 223). What ismost

significant isthat the earliest examplein the series is dated 1172 (G. Schlumberger,

F. Chalandon and A. Blanchet, Sigillographiede I'orient latin, 1943, 134, nos. 163-167, pis.v/g, xx/s).

In contrast, the royal seals start out at thebeginning of the century by showing the HolySepulchre with a conoid dome (Fig. 218), likethe one on the Templum Domini. Followingthe capture of the city in 1099 and the estab-lishment of the Kingdom of Jerusalem, thenew rulers began to renovate and enlarge theHoly Sepulchre; at the same time they con-verted the Aksa mosque (Fig. 43) into a partof the palace and transformed the Dome ofthe Rock (Figs. 37, 38), which the Arabs hadbuilt in imitation of the Holy Sepulchre, intoa church known as the Templum Domini.When the royal seal was designed during thereign of Baldwin I (1110-1118) the two greathistoric and domical buildings of Jerusalemwere presented on either side of the Towerof David, which guarded the western andprincipal entrance of the city. Since the Sepul-chre lay immediately to the left of the PortaDavid, while the Templum was on the easternside of the city, it was natural for the designto have the Anastasis on the left, balancingthe Templum, which was crowned with across in place of the Arab crescent. Therefore,it is important to note that the seal of BaldwinI, the first of the series (Fig. 218), shows theHoly Sepulchre with a conoid dome similarto the one on the Templum (ibid., no. i, pi.XVI/l).The method of representation, while notentirely true to fact because at this time theAnastasis had an oculus in the top of its dome,proves that the building was still symbolicallythought of as a domical structure. One of thedifficulties in interpreting the architectural

evidence of these seals has been the disregardof a chronological development, for Vincent(Jerusalem, n, fig. 386) attributes a seal ofBaldwin III to Baldwin I The strongestproof that the designers were at first con-sciously presenting the Holy Sepulchre asdomical, and not conical, is the seal of Bald-win II (Fig. 221) which shows that he hadtaken over the royal design of a tower flankedby- two domical buildings for the seal of hiscity of Ramah (ibid., no. 126, pi. xix/j), withthe difference that one dome is surmounted bya crescent and the other by a cross. The domi-cal intent of the die-makers is again illustratedby another seal of Ramah, belonging to Bald-win III of Mirabel and dating between 1 168and 1174, on which the domical buildingsflanking the tower are surmounted by flyingbanners (ibid., no. 127, pi. xvm/4).The subsequent seals of Baldwin II (1118-1 131) and Baldwin III (1144-1162) modify thedesign and present the Holy Sepulchre moreaccurately (Fig, 219) with a hypaethral open-ing, so exaggerated in many cases that it lookslike a crescent (ibid., no. 5, pi. xvi/a). Al-though this change, intended to distinguishthe Anastasis from the Templum, gave riseto some stylized simplifications which in somecases make the roof look like a truncated conesurmounted by a crescent, the series as a wholeshows that the intent was to present the roofas a dome with an oculus (Fig. 220). Thisintent is evident in the way that the single,curved lines, intended to depict the gores ofthe dome, are clearly different from thestraight, double lines of the timbered cone onthe seals of the Canons (Figs. 222, 223). Fur-thermore, the curious variations and even mis-understandings of the roof on the later sealsall indicate the persistence of a domical tradi-tion: on the seal (Fig, 220} of Amaury I (i 162-1175) the Holy Sepulchre looks like a domicalstructure surmounted by a crescent (ibid., no.12, pi. xvi/3); on the later seal (Fig. 225) ofBaldwin V (1183-1185) the Holy Sepulchre,now moved to the right of the tower, appearsto have an onion-shaped dome surmountedwith a crescent (ibid., no. 17, pi. xvi/5); itsdomical shape (Fig. 226) on a seal of Guy deLusignan (i 186) is still very apparent (ibid.,no. 18, pi. 1/2); and on the seal of Jean deBrienne (1210-1237), executed after the lossof Jerusalem (Fig. 227), the traditional intentof presenting the dome with an hypaethralopening is more evident (ibid., no, 26, pi. 1/3).What is most significant in this whole series

19


32/220


33/220


34/220

WOODEN DOME IN THE NEAR EASTLater in the history of the monument the casualness of the Western interest in the

shape of the roof is illustrated at the end of the seventeenth century by De Bruyn, whocalls it a dome, but describes it as a truncated cone and so shows it in his drawingof the exterior; and yet at the same time his drawing of the interior (Fig. 4), which isdone in careful perspective, gives it the appearance of having been domical. Twenty-five years after Ladoire described and drew the interior as conical, Pococke, in 1745,shows the interior as domical in the same way that De Bruyn did and writes, Theroof was of cypress, and the King of Spain, giving a new one, what remained of theold roof was preserved as reliques. . . . There is a hole in the top of the dome to givelight, as in the Pantheon at Rome/'41 By 1810 the roof, which had been seriouslydamaged in the fire of 1808, was rebuilt with a dome that was replaced between 1863and 1869 by the present metal roof consisting of two domical shells pierced by anoculus. It is, therefore, impossible to see how the history of the Holy Sepulchre, whichthe Arabs consistently described as a domical building, furnishes any evidence insupport of the belief that there was a conical roof on either the seventh century or theByzantine structure.

Regarding the construction and appearance of the original and presumably Con-stantinian building, authorities for the most part have accepted Pere Vincent'srestoration (Fig. i) and assumed that it had a gigantic masonry dome supported upona circular colonnade in two stories. Quite apart from the historical evidence for theuse of a wooden dome, and without considering the question of either the gallery orthe shape of the actual tomb, this restoration, in its relation of solids to voids (Fig. 3),shows why a rotunda of this kind must have always had a wooden roof, as it did afterthe rebuilding by Modestus. Structurally a masonry dome of seventy-two feet in span,such as is shown in the restoration, could not have been supported on light columnarsupports and without massive and much heavier buttressing up to its haunch. Sinceall the evidence for the Holy Sepulchre combines with the other evidence for theSyrian use of the wooden dome to prove that the rotunda had a wooden roof, it isonly a question of establishing its shape.

41 A drawing of 1586 (G. Zuallardo, // domes, the one over the tomb being made ofDevotissimo viaggio di Gerusalemme, 1595; R. cedars and covered with lead, and the otherKrautheimer, Santo Stefano Rotondo a Roma over the Crusaders' choir being made of stone,e la chiesa de San Sepolcro a Gerusalemme, and the crude drawing shows the roof of theRivista di archaeologia cristiana, xn, 1935,88, Anastasis as a truncated cone; in 1681 (Dr.fig. 7) depicts in a crude fashion a wooden roof O. Dapper, Asia, Beschreibung des ganzenwith an oculus which gives the impression of a Syrien und Palestins . . . , 1681, 345) the roofcone on the interior and a dome on the ex- is depicted as a wooden, truncated cone; C. deterior; the Callot drawings in the account of Bruyn (Voyage an Levant, 1700, 260, 288,Bernardino Amico (Trattato alle Piante


35/220

WOODEN DOME IN THE NEAR EASTOne reason several competent scholars have favored a conical rather than a domicalroof for the original Holy Sepulchre is that they have attached too much Importanceto the few representations which show the Holy Sepulchre with a conical roof (Fig. 9).The representations of the tomb of Christ, and especially those of Western origin,which are nearly always In connection with the scene of the Holy Women at the

Sepulchre, are unreliable as evidence of the actual appearance of the Anastasis becausethe intent of the Christian artist was not to show a specific building that few had seen,but to present a traditional and ideal tegurium whose Implications of salvation wouldbe readily comprehensible. This habit of the mediaeval imagination of dealingprimarily with the symbolic content of architectural forms resulted, as Krauthelmerhas shown in the Western imitations of the Holy Sepulchre,42 in never copying abuilding

in to to but, instead, presenting it figuratively as a divine type.Because of this subjective interest in the meaning of the forms, the memorial ofChrist was represented in a great variety of ways. Sometimes it was shown as a simpleround hut covered with either a domical (Fig. 7) or conical roof, for the Christian,like the antique man, was fully accustomed to think of a shepherd's hut, an ancestraldwelling like the Domus Romulus, as an ideal eternal house in the celestial gardenof an afterlife. Symbolic imagery of this kind, as will be seen, was taken over by theChristians with the result that various forms of ritualistic shelters, like a mortuaryciborium over either a tomb or altar, were frequently called a tegurium, tigurium,tugurium and tiburium. As late as the seventh century Arculph described the tombof Christ as in medio spatio huius interiores rotundas domus rotundum inest in unaeademaque petra excisum tegurium. Because of this predominant interest in the mean-ing of the architectural forms, the representations of the sepulchral dwelling of Christvary according to local types and iconographic purpose. On the Trivulzio book cover(Fig. 9), for example, it is not a two-storied mausoleum which is depicted. Instead,the representation is made up of two quite separate iconographic scenes which theartist combined in superimposed registers. In the upper register the symbolic tombis a traditional tegurium with the tree of the garden growing beside it and the guardingsoldiers asleep on the ground, while in the lower register is the subsequent event ofthe angel receiving the Holy Women at the door of the empty Sepulchre. Later in theWest, partly as a result of this fortuitous combination and also because of the influenceof existing pagan heroa and mausolea, the Holy Sepulchre was actually visualizedin more monumental terms as a two-storied structure (Fig. 8). In the end it Is likelythat this new conception of the Sepulchre actually influenced the rebuilding of theinner tomb.48

42 R. Krautheimer, Introduction to an ion that the western ivories preserve the'Iconography of Mediaeval Architecture/ appearance of the tomb. The representationsJournal of the Warburg and Courtauld Insti- have been compiled by N. C. Brooks, Thetutes, v, 1942, 1-33. Sepulchre of Christ la Art and Liturgy, with

43 E. Baldwin Smith, A Source of Mediaeval Special Reference to the Liturgic Drama/'Style in France, Aft Studies, it, 1924, 90; University of Illinois Studies in Language andGrabar (Martyrium, I, 277-282) is o the opin- Literature, vn, 1921, 6-110.

23


36/220

WOODEN DOME IN THE NEAR EASTThe purely symbolic intent of these representations is best shown in the mosaics

of S. Apollinare Nuovo (Fig. 1 2) where the tomb is a traditional classical and mortuarytholos, a memorial, or heroon. Even in the East, where one might expect more fidelityto fact, the forms of the tomb are more ideological than naturalistic: on the Madabamosaic (Fig. 1 1) it is only a domelike omphalos rising out of the center of the HolyCity,44 and on the Rabula Gospels (Fig. 10) it is a Syrian rustic shelter, a kind of nativeand ritualistic kalube. At about the same period, in the scene of the Entry into Jeru-salem on the Rossano Gospels (Fig. 16), the city is dominated by a structure whosemelon dome, regardless of the sequence of events, was intended to represent the HolySepulchre.

Quite apart from the various ways in which the tomb of Christ was presented in theideational art of this period, the Holy Sepulchre had to be domical because at the timewhen Constantine undertook to honor the King of Heaven the domical mausoleumwas an established type of imperial tomb or memorial, which had a profound symbolicsignificance in relation to the Cult of the Caesars. On the eternal memorial of animmortal kosmokrator the dome, surmounted by the celestial eagle (Figs. 17-21), hada heavenly implication which appealed to all those who by the fourth century hadcorne to visualize the Saviour in all the conventional terms of a divine imperator.45Therefore, the domical tomb, as a glorified tegurium and cosmic dwelling whichoccurs so frequently during the Constantinian period on the coinage as an aeternamemoria and from the time of Diocletian became the standard mausoleum of Westernand Early Byzantine emperors, was the only form under which Constantine couldrecognize a heavenly master.46

44 Grabar, Martyrium, i, 235.45 A. Alfoldi, Insignien und Tracht derRomischen Kaiser, Romische Mitteilungen,

L, 1935, 1-158.46 The importance of the domical mauso-leum during the Constantinian period isshown by its appearance on the imperial coins.While there are several types of memorial, ortomb, they are all characterized by a domesurmounted by a celestial eagle. The simplesttype (Fig. 17) has a plain, masonry cylinderpierced by a rectangular doorway (J. Maurice,Numismatique constantinienne, 1908-12, I,pis. xvii/i*, xix/io); it usually occurs on coinsdedicated to Romulus, but a variant with anarcuated doorway occurs on coins of ValeriusRomulus (ibid., pi. vn/i4), of Divus Maxi-mianus (H. Cohen, Description historiquedes monnaies frappees sous I'empire romain,1880-92, vi, 533), and on others (Fig. 18) struckin honor of Constantius Chlorus (Maurice,op.dL} ii, pi. 711/5; F. Gnecci, / Medaglioniromani, 1912, n, pi. 128/2). The second typehas columns, six on a coin of Maxentius (Fig.20) and four on others (Fig. 19) (Maurice,

op.cit., i, pis. XVH/IO, 11 and xix/i, 2, 9;Cohen, op.cit., vn, 102, 183; Collection R.Jameson, n, 1913, pi, xv/jgg). The third type(Fig. 21), on coins struck by Maxentius, hasits drum decorated with statues in projectinggabled niches and suggests the appearance ofTheodoric's mausoleum at Ravenna (J. Lie-gle, Architekturbilder auf antiken Munzen/'Die Antike, xn, 1936, 221, Abb. 25/d; Voetter,Munzen . . . Diodetianus bis Romulus, figs.44, 48). It is to be noted that the mausoleumwith one door open undoubtedly indicatesthat no body was actually buried in the me-morial.Although the domical tomb had great an-

tiquity in various parts of the Mediterraneanworld, it had become by the late antiqueperiod essentially a Roman form of mauso-leum. There is no evidence, however, that itbecame an imperial type of monument untilafter Diocletian erected his domical tombtemple next to his royal dwelling and madeof it a kind of omphalos in the center of hiscity of Spalato. This idea of constructing aheavenly and eternal memorial in the center


37/220


38/220

WOODEN DOME IN THE NEAR EASTSepulchre at Jerusalem, or Constantius's erection of a mausoleum adjacent to thechurch of the Holy Apostles, had a lasting influence upon imperial sepulchral archi-tecture. At Rome the round and domical tomb of Honorius was built adjacent to thechurch of S. Peter; at Ravenna the cruciform and domical tomb of Galla Placidia(Fig. 73) was near to the church of Santa Croce; the tomb of Theodoric, based on animperial type, was close to his palace, and, in Carolingian times, the domical tombof Charlemagne was next to his palace at Aachen.The only essential difference between this imperial mortuary tradition and theHoly Sepulchre was in the shape and construction of the dome. Instead of havingbeen built in the Roman manner as a geometrical hemisphere of masonry, whichwould have required concealing much of its domical shape on the exterior by but-tressing, the dome of the Anastasis was a free-standing and gilded form of wood car-pentry which rose on high to make manifest the heavenly character of the abode ofChrist. It was constructed of wood in the Syrian manner and by Syrian workmen, asthe name of its architect, Zenobius, would suggest. That it was slightly puffed-up andpointed, having the form of a truncated pine cone like the domes on the pagan Marne-ion, the church of S. Stephen at Gaza, and the Arab mosques, is indicated by somegraphic evidence and by the fact that the conoid shape had long been a manifestationof divine presence in Syria. Later, in Chapter iv, will be considered the other beliefswhich were combined with these traditional conceptions of a domical structure as akingly tomb and ideal dwelling in the afterlife to give added content to the shape ofthe Holy Sepulchre.The appearance of its light, soaring dome o wood is probably preserved in thescene of the Resurrection painted on the reliquary from Jerusalem in the Museo Sacroof the Vatican (Fig. 14), which cannot be dated later than the seventh century.49 Thebox is the only certain piece of Palestinian painting of the Early Christian periodwhich we possess and, according to C. R. Morey, affords us the best indication wehave of the -appearance of the Holy Sepulchre in the sixth century/'50 The shape of thedome suspended above the ciborium is slightly puffed-up and pointed, while itsevident symbolism is shown by the stars painted on its under surface51 and by the treesin the background.52

49 C. R. Morey, 'The Painted Panel from who arrived at the unanimous conclusionthe Sancta Sanctorum/' Festschrift zum seek- that the script cannot be of so late a date, iszigsten Geburtstag von Paul Clemen, 1926, earlier than the eighth century and so far as151-156, fig. 13; P. Lauer, Le Tresor du existing criteria indicate might date as earlySancta Sanctorum, Man. Piot, xv, 1906, 978., as the sixth (Morey, op.cit., 151). Grisar at-pL xiv/2; H. Grisar, Die wmische Rapelle tributed the box to the ninth or tenth cen-Sancta Sanctorum und ihr Schatz, 1908, 113, turies; Diehl favored a late date; Wulff andfig- 59 > O. Wulff, AUchristliche und byzan- Dalton were both inclined to the sixth cen-tinische^ Kunst, i, 1918, 312, fig. 290; Grabar, tury; Vincent (Jerusalem, n, 177, fig. 108)Martyrium, i, 259. published the scene as a representation of the60 Lauer dated the box by the inscription as Constantinian Holy Sepulchre on a tenth cen-late as the tenth century. C. R. Morey obtained tury miniature from the Lateran.the opinions o Monsignor Mercati, Pio 51 For celestial symbolism, see p. 91.Franchi de* Cavalieri and Professor Mercati, ln the Resurrection scenes on the Sancta

26


39/220

WOODEN DOME IN THE NEAR EASTSanctorum panel (Fig. 14), the Trivulzio ivory(Fig. 9), the Munich ivory (Fig. 8) and theRabula Gospels (Fig. 10), the combination ofa tree with a circular and domical sanctuary,or mortuary tholos, was a traditional form ofpagan symbolism appropriated by the Chris-tians. The association of a sacred tree with adomical tholos, rustic shrine and ancestraltegurium was a common theme in both Hel-lenistic and Roman art, sometimes used todepict the ancient abode of a god in his sacredwoods, but more often to show the antiqueveneration for the mythical golden age whenmen, gods and animals lived together in asylvan and earthly paradise. The subject wasalso a popular motif in sacred, divine andfunerary gardens (P. Grimal, Les Jardins ro-mains, 1943, 61) of Hellenistic origin. A repre-sentation of an early Hellenistic funerarygarden is preserved on an inlaid cover of apyxis (Fig. 25) found in a tomb at Tresilicoin the Italian province of Reggio Calabria,which antedates the first century B.C. and was,as were other objects in this and similar graves,presumably of Alexandrian origin (E. Galli,Riflessi di pittura allesandrina in Calabria,Rivista del R. Istituto d'archeologia e storiadell' arte, vi, 1937, 32-46, tav, i). Here on thissepulchral pyxis are all the elements of anideal paradise the round, domical dwellingfor the soul, the sacred trees of the garden, theromantically ruined columns, the statuary andthe narrow bridge over which the soul had topass. In using this scene I am indebted toMiss Berta Segall, who tells me that the scene,although completely Hellenized, reflects anunderlying Egyptian tradition of the Isle ofthe Blessed which was a Nilotic conception ofthe future life, connected with the Isis cultthat influenced Roman representations of fu-nerary gardens and late antique ideas of theafterlife. Since this study went to press, MissSegall has heard rumors that the cover maybe a forgery.While it is difficult to disentangle the vari-ous antique beliefs which combined to givethe domical tholos a mortuary significanceand to make its association with a sacred treethe symbol of a future paradise, it is easy tosee how the Christians came to take over thisprevalent, classical combination of a domicalshrine, an ancestral and divine tegurium, witha sacred tree, for their own ideological repre-sentation of a heavenly paradise. In Romanart the subject began to appear in Pompeiianpainting and on the stucco ceilings of Roman

tombs: in the Domus Vesonius Primus (Fig.145) a great tree has its branches mingled withthe supports of a domical baldachin whichcovers an open-air altar in an idyllic andsacred garden of love and happiness (M.Rostowzew, Die hellenistisch-rornische Arch-itekturlandschaft/' Rom. Mitt., xxvi, 1911, 44,fig. 24); the domical shrine, or heroon, ofMelicertes at Corinth appears on the coins(Fig. 22) with a tree on either side of therotunda dedicated to this particular hero cult(T. L. Donaldson, Architettura numismatica,1859, 61; F. Robert, Thymele, 1939, 1565.);and a Renaissance sketch (Fig. 24) by FraGiocondo (H. de Geymuller, Trois albumsde dessins de Fra Giocondo, Melanges d'arch.et d'hist., xi, 1891, 136, pi. i; R. Jaeger, DieBronzetiiren von Bethlehem/ 1 J&hrb. d. deutarch. Inst., XLV, 1930, no, Abb. 22), shows adomical shrine as a monumental version of asylvan tegurium in combination with a deadtree, figures of Pan, a maenad and two femaledivinities, all indicating the romantic interestin an idyllic past and a kind of lost paradise.

In Early Christian art the representation ofthe Holy Sepulchre as a traditional domicaltomb, tholos, kalube, or tegurium, in combina-tion with a tree, shows not only that thedomical shelter had a symbolic significancewhich the Christians had appropriated fromRoman and Hellenistic art, but also that therewere mystic implications in the domical shapeas the sacred dwelling of a divinity who oncelived on earth among men, and as the idealabode of the soul in a paradisus, or heavenlygarden. This conception of the tomb of Christas a funerary symbol in a celestial garden isalso indicated by the references of early West-ern pilgrims to the atrium in front of theAnastasis as a hortus and paradisus. In adapt-ing the antique symbolism of the domicalshelter and its rustic setting to their own use,the Christians did not limit the symbolism tothe sepulchral abode of Christ, for the tombof Lazarus is depicted on an early piece ogold-glass (Fig. 23) as a domical tegurium witha tree behind it (G. Ferretto, Nate storicobiblwgrafiche di archeologia cristiana, 1942,236, fig. 40), As early as the end of the thirdor the beginning of the fourth century theparadise to which the soul of Jonah is trans-ported is depicted in a tomb at Cagliari inSardinia (Fig. 70) by means of a domicaltegurium and two trees, one the olive tree ofpeace and happiness, and the other the palmas a symbol of Jerusalem (see p. 54; G. B. de

27


40/220

WOODEN DOME IN THE NEAR EASTAs far as the actual domical character of the Holy Sepulchre is concerned, this

scene is equally significant whether the reliquary was executed as Professor Moreyhas argued, in the sixth century before the Anastasis was burnt by the Persians in614 A.D. or was painted shortly after Modestus restored the monument in 616-618.Grisar's argument for a later date in the ninth or tenth century was based upon thefact that the panel shows a rectangular altar stone in front of the ciborium. ThisGrisar considers to be the rectangular altar that Modestus had cut out of one of thepieces of the round millstone at the door of the Sepulchre after it had been brokenin two by the Persians in order to remove the band of gold that decorated it, Sincethe inscriptions inside the box cannot be dated later than the seventh century,53 andthere is no reason to believe that Modestus made any change in the shape and construc-tion of the dome built for Constantine, it does not alter the domical evidence to assumethat the panel may have been executed shortly after 618 A.D., although Professor Moreyhas presented a strong case for a sixth century date. The emphasis given to the dome,decorated with yellow dots as stars, suspended like the canopy of heaven above thetomb of Christ suggests that the reliquary was made to contain one of the pieces ofTrue Cross which had been recovered from the ruins of the holy site and that the novelpresentation of the Resurrection was to commemorate the restoration of the HolySepulchre,The Palestinian origin of the box and the reliability of its representation of the

rotunda are further strengthened by the similarity of its scene of the Holy Womenat the Tomb to the same scene on two ampullae, one in the Dumbarton Oaks Collec-tion (Fig. 158) and another in the Detroit Institute of Arts.54 On these two phialsfrom the Holy Land the rotunda is also depicted as a circular building with interiorcolumns and round-headed, clerestory windows; the only difference being that thedesigner of the ampullae, because of the small, circular space at his disposal, left offthe celestial dome which is so prominent on the box.

In either event the building on the reliquary box, whether it preserves the appear-ance of the Constantinian or the early seventh century memorial of Christ, stronglysupports the other evidence of its having had a dome of wood. The radial and seg-mental lines on the dome, which appear in a good photograph of the box, suggest a

Rossi, Cubicoli sepolcrali cristiani adorni dipitture, Bolletino di archeologia cristiana,series 5, in, 1892, 130-144, pi. vi). The signifi-cance of sacred trees in relation to the martyriaof the Holy Land has been discussed by Gra-bar (Martyrium, r, 71-75) and will again beconsidered on p. 66.

53 In front of the tomb of Christ was the mill-stone which had been adorned with a band ofgold and'jewels as is shown on the Rabula Gos-pels (Fig. i o). In 614 A.D. the gold excited thecupidity of the Persians, who broke the stoneinto two pieces. When the monument was re-

stored the fragments were squared and used asaltars, one in the vestibule and the other in thepresbyterium, according to Arculph. Grisar'sarguments for the late date of the reliquarybox lay great emphasis upon the white, rec-tangular altar stone seen in front of the cibo-rium (Fig. 14). Morey dismisses this argumentas not a serious one, because the existenceof an altar at the entrance of the Sepulchreproper is attested by the Breviarius de 'Hiero-solyma and the Itinerarium Antonini, both ofthe sixth century (p, 152).See p. 99.

28


41/220

WOODEN DOME IN THE NEAR EASTdomical construction in wooden gores such as is described in Choricius' account of thedome at Gaza (p. 39) and are so frequently shown on representations of Palestiniandomes (Figs. 15, 16, 168). Also its slightly puffed-up shape, Its soaring character andsize are all indications that it was made of wood and not surely a vault of stone ashas been Insisted.55Two other bits of evidence, slight in themselves, support the contention that the

original dome of the Constantinian monument had the conoid shape which was char-acteristic of the domical tradition in Syria. In the section Be Sepulchris of theAgrimensorum Romanorum the drawing of a typical sepulchral monument (Fig. 13)has a dome puffed-up, free standing, and pointed.56 Although the AgrimensorumRomanorum is derived from early Roman sources, the drawings are a later addition,possibly of the sixth or seventh century, at a time when the Holy Sepulchre with Itsfree-standing and pointed dome of wood would have been pictured as the Ideal typeof memorial tomb rather than the traditional imperial Roman mausoleum with ahemispherical dome of masonry. Furthermore, a ninth century Greek manuscript(Ambr. 49-50) which Grabar says was strongly influenced by the archaic and Orientaltraditions, probably of Palestine, presents Jerusalem by a schematic and symbolicdrawing of the Holy Sepulchre (Fig. 15) as a tholos with a conoid dome within thewalls of a massive edifice.57 Therefore, when all this evidence is combined and takenin conjunction with the shape of the wooden dome on the church of S. Stephen atGaza and the tomb of Bizzos (Fig. 61), which may have been Influenced by theMartyrium of Christ, the form and construction of the dome on the Holy Sepulchrebecome consistent with the whole tradition of the dome In Syria and Palestine.3, DOMUS AUREA, ANTIOCH (327 and 526-588 A.D.)The other important early central church (Fig. 26), presumably begun by Con-

stantine but finished by Constantius, was the Great Church, or Domus aurea r atAntioch. Little is known of the Constantinian building before its rebuilding afterthe fire of 526 except that it was a large domical octagon with interior colonnadesof great magnificence and that its interior was raised to an enormous height/'58 Itwas dedicated to Christ-Concord and according to Eusebius was a unique buildingof particular beauty in order to be worthy of Antioch, which was the head of allthe peoples [of the Orient]. 59 That it was domical and situated on the island in theNew City as a part of the imperial palace of Antioch has been fairly well established

6* Grabar, Martyrium, i, 259. late as the fifteenth century Symeon of Thes-56 Wolfenbuttel, cod. 36.23 Aug. fol., Agri- salonica

in De sacro Templo (i3$-Migne,P.G.,mensorum Romanorum, fol. 77V ; C. Thulin CLV, col 341) says that /&/>* symbolizes the(Die Handschriften des Corpus Agrimensorum tomb of Christ.Romanorum, Preuss. AL, 1911) says that the s8 Cabrol, Diet, I, Antioche, cols. 23728.;manuscript shows a strong Byzantine influ- Eusebius, Vita Constantini, in, 50; Malalas,ence and that the actual archetype does not Chronographia, ed. Bonn, 1831, 318, 324, 325;antedate 450 A.D. Malalas records that it was founded by Con-

57 Grabar, op.cit, r, 237, fig. 19; Grabar, Les stantine but was completed by Constantius.Miniatures du Gregoire de Nazianze de lTAm- 59 Grabar, Martyrium, I, 220-227.brosienne, i, 1943, pi. XLVIH/J and LII/I. As

29


42/220

WOODEN DOME IN THE NEAR EASTby the fifth century topographical mosaic at Yakto, a village near Antioch. It is pre-sented as a polygonal building (Fig. 29) having a cupola and apsidal vault, with aman standing beside it looking up in prayer and veneration. 60By derivation, location and symbolism this domical church of Christ, according

to Grabar, was a martyrium like the Holy Sepulchre, a traditional funereal, triumphaland memorial heroon of the imperial cult, located like Diocletian's mausoleum atSpalato in the center of his city and next to his palace and intended to make anothersymbolical parallel between Constantine and Christ.61 Here in this temple, with thetitle of Concordia poenitentiae, was worshipped Christ, the heavenly ruler and con-queror, who accorded to his universe the gift of Concordia, just as the Emperor, afterhis triumph of 325, brought about a union of the Roman universe in the uniquereligion of Christ. As a traditional imperial memorial, which became known as theDomus aurea because of its gilded and celestial covering, it was a domical, centralstructure, but its religious purpose as a church of Christ necessitated an apse at thewest side. Beyond the fact that it was domical and presumably built of wood coveredwith gilded lead like the structure which replaced it, nothing more is known aboutthe roof of the original Great Church.When the Golden House was destroyed by fire in 526, Malalas says, The GreatChurch of Antioch, which had been built by the Emperor Constantine the Great,when the disaster occurred and everything else fell to the ground, stood for two daysafter the frightful visitation of God occurred, and it caught fire and was destroyed tothe ground, 62 Later, in 588, Evagrius, an eye-witness, described the effect of thesecond great earthquake which wrecked Antioch. Many buildings, he wrote, weredestroyed when the very foundations were thrown up, so that everything about themost holy church fell to the ground, only the hemisphere being saved, which had beenconstructed of wood from Daphne by Ephraemius after it had suffered in the earth-quake under Justinus. 63 It was tilted, he goes on to relate, toward the north bythe

earthquake which followed [under Justinian in 528] so that it received bracingtimbers. These fell through the violence of the shock [in 588] when the hemispheresettled back and was restored to its right place as though it was set there by a rule.Evagrius' account leaves no doubt that the sixth century dome was made of wood,for no masonry dome could have been tilted, braced with timbers and then settledback into place.

60 Doro Levi, Antioch Mosaic Pavements, 62 Malalas, op.cit., 419, 21, as translated by*947> 332> pi. LXXX/C; the theory of Eltester, G. Downey.that the inscription, PIAVA, signifies the Porta ' Evagrius, vi, 8, translated by G. Downey;Taurlana and thereby locates the church in- the Ephraim, whom Strzygowski (Klein-side the entrance gate of the palace area, is asien, 95) called the master-builder who erect-considered unsound by Levi but has been de- ed the wooden cupola, was Ephraemius, theyeloped by Grabar (Martyrium, i, 215-227) comes Orientalis who became Patriarch of An-into a convincing argument for the location tioch in 527 (Downey, Ephraemius, Patriarchand purpose of the church as part of the palace of Antioch, Church History, vn, 19558 364-on the island.

61 Grabar, op.cit., 223-225.30


43/220

WOODEN DOME IN THE NEAR EAST4. MARTYRIUM, NAZIANZUS (before 374 A.D.)

^

Gregory of Nazianzus, when describing the 'living memorial of his father, writes,It surrounds itself with eight regular equilaterals, and is raised aloft by the beautyof the two stories of pillars and porticoes, while the statues placed upon them aretrue to life; its vault [oupaz^> heaven] flashes down upon us from above, and it dazzlesour eyes with the abundant sources of light. - Although there is nothing in the ac-count to prove that this rnartyrium of eight sides (Fig. 28) had a wooden dome, asWatzinger says it had, 62 it is very evident from Gregory's use of words that Birnbaumwas wrong in thinking that the church had a polygonal roof with a hypaethral open-ing. 68 Regardless of whether the roof was constructed of wood or masonry, the useof the word heaven and the emphasis upon its being a dazzling source of lightshow, as Keil recognized,67 that it was a cupola, because the Christians with theircosmic symbolism always thought of the dome as a celestial shape.Further confirmation of the domical form of this memorial is furnished by ascholion of uncertain date, which compares the rnartyrium to an octagonal sanctuaryat Alexandria, presumably the rnartyrium of S. John the Baptist, and to the The-otokos naos at Tyre, 88 While nothing is known about the church of the Virgin atTyre, which must have been later than the fourth century, except that it was customaryin the Near East to erect domical martyria in honor of the Virgin, there is every reasonto believe that the famous rnartyrium of the Baptist at Alexandria was domical. Notonly did it replace the pagan Serapeion, when that renowned domical temple wasdestroyed by the Christians at the beginning of the fifth century, but it depicted asdomical in the sixth century mosaics at Gerasa (Figs. 30, 31).5. MARTYRIUM, NYSSA (c. 379-394 A.D.)The rnartyrium which Gregory had built around 380 at Nyssa in Cappadocia con-

sisted of a central apsidal octagon with four exedras (Fig. 27), making it, accordingto Grabar,69 a cruciform structure like the rnartyrium of Antioch-Kaoussie (Fig. 170)and the church of S. Simeon Stylites at Kal'at Sim'an (Fig. 32). The letter of Gregoryto Amphilochius sometime between 379 and 594 furnishes important and conclusiveevidence regarding the shape and construction of domical martyria in the fourthcentury. In describing the rnartyrium which he was undertaking to build at Nyssahe writes, Above the eight apses the octagon will rise. The pine cone which rises fromthis will be cone-shaped (/ccovoetSTfc), the vortex reducing the shape of the roof froma plane to a point. 70 This specific statement that the roof of the rnartyrium was like

64 Migne, P.G., xxxv; On the death of his 69 Grabar, Martyrium, i, 152*?., 157.Father, Nicene and Post-Nicene Fathers, vn, 70 Migne, P.G., XLVI, 1093$.; Birnbaum, op.chap. 39, 267. cit., 181-209; Keil (Kleinasien, 778., fig. 63)6'5 Watzinger, Denk. Palds., n, 133, interprets Gregory as saying that it was a66 A. Birnbaum, Die Oktagone von Anti- geometrically conical roof, and his plan (Fig.ochia, Nazianz und Nyssa/' Rep. f. Kunstwiss., 27), based upon Gregory's description, is struc-xxxvi, 1913, 207. turally inexplicable in its location of the67 Keil in Strzygowski, Kleinasien, 94. n. 4. columns.

68 See Chap, n n. 25 and Chap, iv n. 41.31


44/220

WOODEN DOME IN THE NEAR EASTa pine cone, when taken in combination with the description of the Marneion, theaccount of the wooden dome on the church of S. Stephen at Gaza, the conoid domeon the tomb of Bizzos, the persistence in the Christian East of the conoid shelter as aritualistic kalube and cosmic house, and the later Arab adoption of the wooden conoiddome, Is proof that the pine-cone shape and the idea of a domical martyrium werealready traditional by the last quarter of the fourth century.

It is also significant for the history of domical architecture that Gregory goes on tosay he would prefer to build this martyrium with a wooden dome if it were possible.After asking Amphilochius to send him the required workmen, he writes, It isespecially necessary to give attention to the point that some of them shall know howto build domes (eiXijo-ts) without centering, for I have learned that when this is doneit is steadier than if it rests on supports; the scarcity of wood, indeed, leads to this plan,namely that the whole structure shall be roofed with stones, because there is no woodsuitable for roofing in this region/' This is confirmation of the assumption that domesof swollen and pine-cone shape were commonly built of wood and were sufficientlycustomary so that the conoid shape, because of the scarcity of wood, was to be repro-duced in stone. It is difficult to visualize what Gregory meant by domes built withoutcentering being steadier than those resting on supports unless he is thinking of stonesquinches, or pendentives, rather than wooden supports at the corners where thedome had to be fitted onto the octagon.6. MARTYRIA OF CONSTANTINOPLE (fourth and fifth centuries)

It has been customary to think of the Byzantine dome as a masonry vault whichbegan to be common upon the churches largely because of Justinian's interest in amonumental state arch

the dome: a study in the history of ideas

Documents