Byzantine Churches in Constantinople: Their History and Architecture Walter S. George CHAPTER IV THE CHURCH OF S. IRENE Cover. BYZANTINE CHURCHES IN CONSTANTINOPLE Printers mark. MACMILLAN AND CO., Limited LONDON · BOMBAY · CALCUTTA MELBOURNE THE MACMILLAN COMPANY NEW YORK · BOSTON · CHICAGO DALLAS · SAN FRANCISCO THE MACMILLAN CO. OF CANADA, Ltd. TORONTO PLATE I. Mediaeval Map of Constantinople by Bondelmontius. Mediaeval Map of Constantinople by Bondelmontius. Frontispiece. NOTE ON THE MAP OF CONSTANTINOPLE For the map forming the frontispiece and the following note I am greatly indebted to Mr. F. W. Hasluck, of the British School at Athens. The map is taken from the unpublished Insularium Henrici Martelli Germani (B.M. Add. MSS. 15,760) f. 40. A short note on the MS., which may be dated approximately 1490, is given in the Annual of the British School at Athens, xii. 199. The map of Constantinople is a derivative of the Buondelmontius series, which dates from 1420, and forms the base of all known maps prior to the Conquest. Buondelmontius' map of Constantinople has been published from several MSS., varying considerably in legend and other details:1 the best account of these publications is to be found in E. Oberhummer's Konstantinopel unter Suleiman dem Grossen, pp. 18 ff. The map in B.M. Arundel, 93, has since been published in Annual B.S.A. xii. pl. i. In the present map the legends are as follows. Those marked with a dagger do not occur on hitherto published maps. Reference is made below to the Paris MS. (best published by Oberhummer, loc. cit.), the Venetian (Mordtmann, Esquisse, p. 45, Sathas, Μνημεῖα, iii., frontispiece), and the Vatican (Mordtmann, loc. cit. p. 73). Tracie pars—Galatha olim nvnc Pera—Pera—S. Dominicus—Arcena—Introitus Euxini Maris. Asie minorus pars nvnc tvurchia.—Tvrchia. Tracie pars—Porta Vlacherne—♰Ab hec (sic) porta Vlacherne usque ad portam Sancti Demetri 6 M.P. et centum et decem turres—♰Porta S. Iohannis 1—Porta Chamici 2—Porta Crescu—Porta Crescea—♰Ab hec (sic) porta que dicitur Crescea usque ad portam Sancti Demetri septem M. passuum et turres centum nonaginta octo. Et ad portam Vlacherne 5 M. passuum et turres nonaginta sex—Receptaculum Conticasii 3-Porta olim palacii Imperatoris—Porta S. Dimitri—Iudee 4—Pistarie p. 5—Messi p.—Cheone p.6—S. Andreas—S. Iohannes de Petra—Hic Constantinus genuflexus—♰Ad S. Salvatorem—♰Columna Co(n)s?—Hic Iustinianus in equo7—Sancta Sophia—Hippodromus—S. Demetrius—S. Georgius-S. Lazarus—Domus Pape—Domus S. Constantini—Sanctorum Apostolorum—Porta antiquissima mire (sic) arte constructa8—S. Marta9—S. Andreas—S. Iohannes de Studio—Perleftos. F. W. H. 1 S. Romani? 2 Porta Camidi, Vat. 3 Receptaculum fustarum dein Condoscalli, Par. 4 Porta Judea, Par. 5 Porta Piscarii, Par. 6 Porta Lacherne, Par., delle Corne, Vat., del Chinigo (i.e. Κυνηγίου in the xvi. cent. Venetian maps. 7 Theodosius in aequo eneo, Ven. In hoc visus imp. Teod. equo sedens, Vat. 8 Porta antiquissima pulcra, Par. 9 St. Mam (as?) Ven. Sts. Marcus, Vat. BYZANTINE CHURCHES IN CONSTANTINOPLE THEIR HISTORY AND ARCHITECTURE BY ALEXANDER VAN MILLINGEN, M.A., D.D. PROFESSOR OF HISTORY, ROBERT COLLEGE, CONSTANTINOPLE AUTHOR OF 'BYZANTINE CONSTANTINOPLE,' 'CONSTANTINOPLE' ASSISTED BY RAMSAY TRAQUAIR, A.R.I.B.A. LECTURER ON ARCHITECTURE, COLLEGE OF ART, EDINBURGH W. S. GEORGE, A.R.C.A., AND A. E. HENDERSON, F.S.A. WITH MAPS, PLANS, AND ILLUSTRATIONS MACMILLAN AND CO., LIMITED ST. MARTIN'S STREET, LONDON 1912 PREFACE This volume is a sequel to the work I published, several years ago, under the title, Byzantine Constantinople: the Walls of the City, and adjoining Historical Sites. In that work the city was viewed, mainly, as the citadel of the Roman Empire in the East, and the bulwark of civilization for more than a thousand years. But the city of Constantine was not only a mighty fortress. It was, moreover, the centre of a great religious community, which elaborated dogmas, fostered forms of piety, and controlled an ecclesiastical administration that have left a profound impression upon the thought and life of mankind. New Rome was a Holy City. It was crowded with churches, hallowed, it was believed, by the remains of the apostles, prophets, saints, and martyrs of the Catholic Church; shrines at which men gathered to worship, from near and far, as before the gates of heaven. These sanctuaries were, furthermore, constructed and beautified after a fashion which marks a distinct and important period in the history of art, and have much to interest the artist and the architect. We have, consequently, reasons enough to justify our study of the churches of Byzantine Constantinople. Of the immense number of the churches which once filled the city but a small remnant survives. Earthquakes, fires, pillage, neglect, not to speak of the facility with which a Byzantine structure could be shorn of its glory, have swept the vast majority off the face of the earth, leaving not a rack behind. In most cases even the sites on which they stood cannot be identified. The places which knew them know them no more. Scarcely a score of the old churches of the city are left to us, all with one exception converted into mosques and sadly altered. The visitor must, therefore, be prepared for disappointment. Age is not always a crown of glory; nor does change of ownership and adaptation to different ideas and tastes necessarily conduce to improvement. We are not looking at flowers in their native clime or in full bloom, but at flowers in a herbarium so to speak, or left to wither and decay. As we look upon them we have need of imagination to see in faded colours the graceful forms and brilliant hues which charmed and delighted the eyes of men in other days. In the preparation of this work I have availed myself of the aid afforded by previous students in the same field of research, and I have gratefully acknowledged my debt to them whenever there has been occasion to do so. At the same time this is a fresh study of the subject, and has been made with the hope of confirming what is true, correcting mistakes, and gathering additional information. Attention has been given to both the history and the architecture of these buildings. The materials for the former are, unfortunately, all too scanty. No continuous records of any of these churches exist. A few incidents scattered over wide tracts of time constitute all that can be known. Still, disconnected incidents though they be, they give us glimpses of the characteristic thoughts and feelings of a large mass of our humanity during a long period of history. The student of the architecture of these churches likewise labours under serious disadvantages. Turkish colour-wash frequently conceals what is necessary for a complete survey; while access to the higher parts of a building by means of scaffolding or ladders is often impossible under present circumstances. Hence the architect cannot always speak positively, and must leave many an interesting point in suspense. Care has been taken to distinguish the original parts of a building from alterations made in Byzantine days or since the Turkish conquest; while, by the prominence given to the variety of type which the churches present, the life and movement observable in Byzantine ecclesiastical art has been made clear, and the common idea that it was a stereotyped art has been proved to be without foundation. Numerous references to the church of S. Sophia occur in the course of this volume, but the reader will not find that great monument of Byzantine architectural genius dealt with in the studies here offered. The obstacles in the way of a proper treatment of that subject proved insuperable, while the writings of Salzenberg, Lethaby, and Swainson, and especially the splendid and exhaustive monograph of my friend Mr. E. M. Antoniadi, seemed to make any attempt of mine in the same direction superfluous if not presumptuous. The omission will, however, secure one advantage: the churches actually studied will not be overshadowed by the grandeur of the 'Great Church,' but will stand clear before the view in all the light that beats upon them. I recall gratefully my obligations to the Sultan's Government and to the late Sir Nicholas O'Conor, British Ambassador at Constantinople, for permission to make a scientific examination of the churches of the city. To the present British Ambassador, Sir Gerard Lowther, best thanks are due for the facilities enjoyed in the study of the church of S. Irene. I have been exceedingly fortunate in the architects who have given me the benefit of their professional knowledge and skill in the execution of my task, and I beg that their share in this work should be recognized and appreciated as fully as it deserves. To the generosity of the British School at Athens I am indebted for being able to secure the services of Mr. Ramsay Traquair, Associate of the Royal Institute of British Architects and Lecturer on Architecture at the College of Art in Edinburgh. Mr. Traquair spent three months in Constantinople for the express purpose of collecting the materials for the plans, illustrations, and notes he has contributed to this work. The chapter on Byzantine Architecture is entirely from his pen. He has also described the architectural features of most of the churches; but I have occasionally introduced information from other sources, or given my own personal observations. I am likewise under deep obligation to Mr. A. E. Henderson, F.S.A., for the generous kindness with which he has allowed me to reproduce his masterly plans of the churches of SS. Sergius and Bacchus, S. Mary Panachrantos, and many of his photographs and drawings of other churches in the city. I am, moreover, indebted to the Byzantine Research and Publication Fund for courteous permission to present here some of the results of the splendid work done by Mr. W. S. George, F.S.A., under unique circumstances, in the study of the church of S. Irene, and I thank Mr. George personally for the cordial readiness with which he consented to allow me even to anticipate his own monograph on that very interesting fabric. It is impossible to thank Professor Baldwin Brown, of the University of Edinburgh, enough, for his unfailing kindness whenever I consulted him in connection with my work. Nor do I forget how much I owe to J. Meade Falkner, Esq., for kindly undertaking the irksome task of revising the proofs of the book while going through the press. I cannot close without calling attention to the brighter day which has dawned on the students of the antiquities of Constantinople since constitutional government has been introduced in the Ottoman Empire. Permission to carry on excavations in the city has been promised me. The archaeology of New Rome only waits for wealthy patrons to enable it to reach a position similar to that occupied by archaeological research in other centres of ancient and mediaeval civilizations. But the monuments of the olden time are perishable. Of the churches described by Paspates in his Byzantine Studies, published in 1877, nine have either entirely disappeared or lost more of their original features. It was no part of wisdom to let the books of the cunning Sibyl become rarer and knowledge poorer by neglecting to secure all that was obtainable when she made her first or even her second offer. ALEXANDER VAN MILLINGEN.      Robert College, Constantinople. Πόλις ἐκκλησιῶν γαλουχέ, πίστεως ἀρχηγέ, ὀρθοδοξίας ποδηγέ Nicetas Choniates. CONTENTS   Page CHAPTER I Byzantine Architecture 1   CHAPTER II Church of S. John the Baptist of the Studion 35   CHAPTER III Church of SS. Sergius and Bacchus 62   CHAPTER IV Church of S. Irene 84   CHAPTER V Church of S. Andrew in Krisei 106   CHAPTER VI Church of S. Mary Panachrantos 122   CHAPTER VII Church of S. Mary Panachrantos 138   CHAPTER VIII Church of S. Theodosia 164   CHAPTER IX Church of S. Mary Diaconissa 183   CHAPTER X Church of SS. Peter and Mark 191   CHAPTER XI Church of the Myrelaion 196   CHAPTER XII Church of S. John the Baptist in Trullo 201   CHAPTER XIII Church of S. Thekla 207   CHAPTER XIV Church of S. Saviour Pantepoptes 212   CHAPTER XV Church of S. Saviour Pantokrator 219   CHAPTER XVI Church of S. Theodore 243   CHAPTER XVII Monastery of Manuel 253   CHAPTER XVIII Monastery of Manuel 262   CHAPTER XIX Balaban Aga Mesjedi 265   CHAPTER XIX Balaban Aga Mesjedi 265   CHAPTER XX Church of the Gastria 268   CHAPTER XXI Church of S. Mary of the Mongols 272   CHAPTER XXII Bogdan Serai 280   CHAPTER XXIII Church of S. Saviour in the Chora 288   CHAPTER XXIV Mosaics and Frescoes in the Church of S. Saviour in the Chora 321   CHAPTER XXV Dating and Classification of the Churches 332   BIBLIOGRAPHY 337   LIST OF EMPERORS 341   INDEX 343 PLANS AND ILLUSTRATIONS FIG.   Page 1. Kasr Ibn Wardan 4 2. DerÉ Aghsy 6 3. DerÉ Aghsy (Section) 6 4. S. Nicholas, Myra 7 5. Church of the Koimesis, Nicaea 8 6. Church of the Koimesis, Nicaea (Section) 9 7. Map of Byzantine Constantinople facing  15 8. The Saucer Dome or Dome-Vault 16 9. The Dome on Pendentives 16 10. The Drum Dome 17 11. Diagram of Vaulting in Outer Narthex of S. Saviour in the Chora. 22 Church of S. John the Baptist of the Studion 12. Plan of the Church 56 13. Long Section 57 14. Cross Section, looking east 58 15. Cross Section, looking west 58 16. Elevation of the Narthex 59 17. Longitudinal Section of western portion of the Nave—Half-cross Section of the Nave 59 18. Details of the Narthex, Colonnade, Doors, Windows 60 19. Details of Doors; Details from Church of S. Theodore; Details from S. Saviour in the Chora 61 Church of SS. Sergius and Bacchus 20. Inscription on the Frieze in the Church 74 21. Exterior View of the Dome 77 22. Brick Stamps in the Church 79 23. Ground Plan (looking up) 80 24. Gynaeceum Plan (looking up) 80 25. Plan at base of Dome (Cross Section) 81 26. Transverse Section 81 27. Section through South Aisle 81 28. Constructive Section of the Interior Arrangement, showing Gynaeceum Floor, Vaulting, Roof, and Springing of Dome 82 29. Constructive Section of the Rear, with Gynaeceum, Floor, and Roof removed 82 30. Sections of Mouldings 83 Church of S. Irene 31. Ground Plan of the Atrium and Church facing  104 32. Gallery Plan  "     104 33. Longitudinal Section  "     104 34. South Elevation  "     104 35. West Elevation 105 Church of S. Andrew in Krisei 36. Plan of the Church (restored) 118 37. Plan of the Church 119 38, 39. Longitudinal Sections 120 40. Isometrical Section (restored) 121 Church of S. Mary Panachrantos 41. Details of the Shafts in East Windows of South Church 124 42. Inscription on Apse of North Church 131 43. Plan of the Church (conjectural) 135 44. Plan of the Church 136 45. Section through the North Church 137 46. Section through the South Church 137 Church of S. Mary Pammakaristos 47. Plan of the Church (conjectural) 152 48. Brick Details from the Parecclesion 154 49. Inscribed String-course on Apse of the Parecclesion 157 50. Plan of the Church—Plan of the Parecclesion—Plan of the Gynaeceum in the Parecclesion facing   160 51. Cross Section of the Church, looking east 161 52. The Parecclesion, east end of south side 162 53. Sections in the Parecclesion—Plan of Dome in the Gynaeceum 163 Church of S. Theodosia 54. Interior of the Church, looking west 171 55. Details from the Church—Details from Church of S. Theodore—Capital and Shaft found near Unkapan Gate 174 56. Ground Plan 179 57. Plan of the Gynaeceum 180 58. Section in the Gynaeceum 181 59. Longitudinal Section of the Church 181 60. Isometrical Section, showing scheme 182 Church of S. Mary Diaconissa 61. Plan of the Church 189 62. Longitudinal Section 190 Church of Ss. Peter and Mark 63. Font in the street to the west of the Church—A Window in S. Saviour in the Chora 194 64. Plan of the Church 195 65. Longitudinal Section 195 Church of the Myrelaion 66. Plan of the Church 200 67. Longitudinal Section 200   Church of S. John the Baptist in Trullo 68. Details from the Church—Details from the Pammakaristos—Details from the Pantepoptes 203 69. Details from S. Andrew in Krisei—Details from the Chora 204 70. Plan of S. John in Trullo—Longitudinal Section—Plan of the Dome 206 Plan of the Church—Cross Section 206 Church of S. Saviour Pantepoptes 72. Sketches from the Church 213 73. Plan of the Church—Longitudinal Section 217 74. Details from the Church 218 Church of S. Saviour Pantokrator 75. Details from the Church—Details from S. Saviour Pantepoptes 225 76. Inlaid Marble Pavement in the Pantokrator—Tile Pavement in the Gallery of S. Theodosia 234 77. Plan of the Pantokrator facing  240 78. Longitudinal Section through the North Church 241 79. Longitudinal Section through through the Central Church 241 80. Longitudinal Section through through the South Church 242 Church of S. Theodore 81. Details from the Church 245 82. Plan of the Church by Texier 249 83. Part of South Elevation showing the Side Chapel by Texier 249 84. Plan of the Church 251 85. Longitudinal Section 251 86. Front Elevation—Half-Plan of Central Dome in the Narthex 252 87. South Elevation and Section through Vaulted Bay of Narthex—Half-Plan of Central Dome 252 Monastery of Manuel 88. Plan of the Refectory 261 Monastir Mesjedi 89. Plan of the Church—Cross Section 261 Balaban Mesjedi 90. Plan of the Building 267 91. Section 267 Church of the Gastria 92. Plan of the Church 267 Church of S. Mary of the Mongols 93. Exterior View 273 94. Interior View 274 95. The Dome (Interior View) 276 96. Plan of Church 279 97. Plan of S. Nicholas Methana 279 Bogdan Serai 98. Plan of Upper Chapel—Half-Section of Apse—Half-Section of East End—Longitudinal Section 287 Church of S. Saviour in the Chora 99. Details from the Church 305 100. Details of a Window in the Gallery 309 101. Plan of S. Sophia, Salonica 313 102. Plan of the Chora (restored) 314 103. Bay in the Chora (restored) 315 104. Plan of Church of the Archangels, SygÉ 316 105. Plan of the Chora and the Parecclesion 317 106. Cross Section, looking west 318 107. Plan of Upper Gallery 318 108. Section through Church 319 109. Section through Chapel 319 110. Plan of Dome 320 111. Section through Inner Narthex 320 112. Plan of Gallery between the Church and the Parecclesion 320 113. Section of North Gallery 320 114. Plan of the Narthexes, indicating the positions of their Mosaics 321 115. Model of the Church in the Mosaic over Main Door in the Inner Narthex 326 116. Plan of the Parecclesion, indicating positions of its Frescoes 328 PLATES I. Mediaeval Map of Constantinople by Bondelmontius Frontispiece     facing page II. The Myrelaion (since it was burned). (1) From the north-west. (2) From the south-east 20 III. (1) The Myrelaion (since it was burned). The Interior, looking east. (2) Suleiman Aga Mesjedi 24 IV. (1) Bracket in S. Saviour in the Chora. (2) Sculptured Slab in S. Theodore. (3) S. Mary Diaconissa. Heads of Windows in south arm. (4) Sculptured Slab on the West Wall 28 Church of S. John of the Studion V. The Ruined Interior, seen from the minaret of the mosque. (2) The West Side of the Church 36 VI. (1) FaÇade of the Narthex. (2) The Ruined Interior, at the West End of the North Side 40 VII. Entablature and Anta Capital in the Narthex. (2) Cornice in the Narthex, looking up 44 VIII. The Church from the south-east 48 IX. (1) The East End of the Church. (2) East End of the North Side of the Church 52 X. (1) The Cistern beside the Church. (2) Another View of the same 54 The Church of Ss. Sergius and Bacchus XI. The Interior of the Church, looking north-west 62 XII. (1) A Capital in the Church. (2) A Capital in the Narthex of S. John of the Studion 66 XIII. (1) The Church from the south-east. (2) View in the Gallery over the Narthex 70 XIV. (1) The Interior of the Church, looking north-east. (2) Portion of the Entablature in the Church 74 XV. The Baptistery of the Church of S. Sophia. (1) The Exterior from the north-east. (2) The Interior of the Dome, showing Continuous Pendentives 76 The Church of S. Irene XVI. The Church from the south-east 84 XVII. (1) The South Side. (2) The North Side 86 XVIII. The Interior, looking east 90 XIX. (1) Vaulting at the north-western Corner of the Atrium. (2) The Northern Arch of the Main Dome, seen from the South Gallery 92 XX. (1) Mosaic in the Soffit of an Arch. (2) Portion of the Mosaic Inscription on the Outer Arch of the Apse 96 XXI. (1) The Interior, looking west. (2) The Door at the East End of the North Aisle 98 XXII. (1) Vaulting over the South Aisle. (2) A Compartment of Vaulting in South Aisle, looking up 100 XXIII. (1) A Capital in the South Arcade. (2) Base of a Column in the South Aisle 102 The Church of S. Andrew in Krisei XXIV. The East End of the Church 106 XXV. (1) The Church from the south-west. (2) The Interior, looking south 108 XXVI. (1) A Capital in the Inner Narthex. (2) A Capital in the Arcade under the West Dome Arch 110 XXVII. (1) A Capital in the Outer Narthex. (2) A Capital in the Outer Narthex 112 XXVIII. (1) View in the Outer Narthex. (2) View in the Inner Narthex, looking south 114 XXIX. (1) View in the Cloister. (2) View in the Courtyard 116 XXX. (1, 2, 3) Three Views of the Decorated Doorway in the Cloister 118 The Church of S. Mary Panachrantos XXXI. (1) Vault of the Passage on the West of the Dome in the South Church. (2) The Interior of the North Church, looking north 122 XXXII. (1) The North Church, looking east. (2) The North Church, looking west 126 XXXIII. (1) The Diaconicon, looking east. (2) The Western Dome Arch in the South Church 128 XXXIV. (1) The East Window of the South Church. (2) The Outer Narthex, looking south 130 XXXV. (1) The East End of the South Church. (2) The East End of the North Church 132 Church of S. Mary Pammakaristos XXXVI. (1) The Church from the south-east. (2) The Church from the west 138 XXXVII. (1) The Inner Narthex, looking south. (2) The Dome, looking west 142 XXXVIII. The Parecclesion from the south-east 144 XXXIX. (1) The East End of the Parecclesion. (2) The West Column in the Parecclesion 148 XL. (1) The East Column in the Parecclesion. (2) The Column flanking the East Window in the Apse of the Parecclesion. (3) The West Column in the Parecclesion 150 XLI. (1) Interior View of the Dome in the Parecclesion. (2) Mosaic in the Dome of the Parecclesion 154 XLII. South Side of the Parecclesion 156 Church of S. Theodosia XLIII. (1) The East End of the Church. (2) The Church from the south-east 164 XLIV. (1) The Interior, looking north-east. (2) The Eastern Dome Arch 168 XLV. (1) The Dome over the Stairway to the Galleries. (2) The Narthex, looking north, and the Stairway to the Galleries 172 Church of S. Mary Diaconissa XLVI. (1) The North-west Side, seen from the Aqueduct of Valens. (2) The North Arm, looking east 182 XLVII. (1) The Interior, looking north-east. (2) The Interior, looking south-east 184 XLVIII. (1) The Lower Part of the North Side of the East End. (2) The Upper Part of the North Side of the East End 186 XLIX. (1) South Eikon Frame. (2) Detail in the South Eikon Frame 188 L. (1) The Interior, looking west. (2) A Capital on the Column at the Entrance to the Church 190 Church of Ss. Peter and Mark LI. (1) The Church from the south-east. (2) Font outside the Church 192 LII. (1) The Dome, looking north. (2) Looking across the Dome, from the south-west 194 Church of the Myrelaion LIII. (1) The South Side of the Church. (2) The Narthex, looking north 196 LIV. (1) The Interior, looking east. (2) The south-west Angle of the Cross 198 LV. (1) S. John in Trullo from the south-west. (2) The Interior of Balaban Mesjedi 202 LVI. The Church of S. Thekla. (1) From the north-west. (2) The East End 208 The Church of S. Saviour Pantepoptes LVII. (1) Door leading from the Outer to the Inner Narthex. (2) The Dome, looking west 212 LVIII. (1) Decoration in Brick on the Exterior of the South Wall. (2) Bracket in the Parecclesion of S. Mary Pammakaristost 214 Church of S. Saviour Pantokrator LIX. The Church from the west 220 LX. (1) The Church from the north-west. (2) Fragments of Sculptured Marbles found in the Church 222 LXI. (1) Interior of the South Church, looking east. (2) The Southern Arm of the Church 224 LXII. (1) Entrance from the Narthex to the South Church. (2) Interior, looking from the South Church through into the North Church 226 LXIII. (1) The Gallery in the North Church, looking south. (2) Interior of the North Church, looking east 228 LXIV. (1) Arch in the North Wall of the South Church, seen from the South Church, looking north. (2) Arch in the North Wall of the South Church, seen from the Central Church, looking south 230 LXV. (1) Narthex of the North Church, looking north. (2) Outer Narthex of the South Church, looking north 232 LXVI. (1) South Bay in the Gallery of the South Church. (2) View in the North Church, looking south 234 LXVII. The Pulpit in the South Church. (2) West Side of the Central Bay in the Gallery of the South Church 236 LXVIII. (1) Interior of the East Dome in the Central Church. (2) Interior of the Dome in the South Church, looking north 238 LXIX. (1) The East End from the south. (2) East Window of the Central Church. (3) The East End from the north 242 Church of S. Theodore LXX. (1) The North End of the Western FaÇade. (2) The Church from the north-west 244 LXXI. (1) The Central Dome from the south. (2) The Western FaÇade from the south 246 LXXII. (1) The South Cross Arm (exterior) from the south-east. (2) The East End from the south 248 LXXIII. (1) The Capital on the Southernmost Column in the FaÇade. (2) Capital in the FaÇade 250 LXXIV. (1) The Outer Narthex, looking north. (2) Capital on the North Side of the Door leading from the Outer to the Inner Narthex 254 LXXV. (1) The Interior, looking east. (2) The Interior (upper part), looking east 256 The Refectory of the Monastery of Manuel LXXVI. (1) From the west. (2) From the south-east 258 LXXVII. Cistern of Aetius 262 Church of the Gastria LXXVIII. (1) The Church from the east. (2) The Entrance 268 LXXIX. (1) The Church from the west. (2) The Interior 270 Bogdan Serai LXXX. (1) The Apse in the Upper Chapel. (2) A Pendentive of the Dome. (3) The Chapel, from the north-west 280 Church of S. Saviour in the Chora LXXXI. (1) The Church from the west. (2) The Church from the south-east 288 LXXXII. (1) The Church from the north-east. (2) The North Side of the Church 292 LXXXIII. (1) The Inner Narthex, looking south. (2) Another View of the same 296 LXXXIV. (1) A Capital in the Outer Narthex. (2) Another Capital in the Outer Narthex 300 LXXXV. (1) The Interior, looking north-west. (2) The Outer Narthex, looking south 304 LXXXVI. (1) The Eikon Frame on the South-eastern Pier. (2) The Interior, looking east 308 LXXXVII. (1) The Cornice above the Main Door (on the Interior). (2) The Archivolt on the North Wall of the Parecclesion. (3) Window Heads in the Central Apse 310 LXXXVIII. (1) The East End of the Parecclesion. (2) A Capital at the Entrance to the Parecclesion 314 LXXXIX. (1) The Parecclesion, looking south-east. (2) The Parecclesion, looking west 316 XC. (1) Mosaic representing the Miracle of Water turned into Wine. (2) Mosaic representing Mary caressed by her parents, and blessed by priests seated at a banquet 322 XCI. (1) Mosaic representing the Registration of Mary and Joseph at Bethlehem. (2) Mosaic representing Theodore Metochites offering the Church to Christ 326 XCII. The Archivolt on the South Wall of the Parecclesion, with the Epitaph in honour of Tornikes 330 CHAPTER I BYZANTINE ARCHITECTURE I. Planning At the beginning of the fifth century, which is a suitable point from which to date the rise of Byzantine architecture, three principal types of church plan prevailed in the Roman world:— I. The Basilica: an oblong hall divided into nave and aisles, and roofed in wood, as in the Italian and Salonican examples, or with stone barrel-vaults, as in Asia Minor and Central Syria. II. The Octagonal or Circular plan covered with a stone or brick dome, a type which may be subdivided according as (1) the dome rests upon the outer walls of the building, or (2) on columns or piers surrounded by an ambulatory. The Pantheon and the so-called Temple of Minerva Medica at Rome are early examples of the first variety, the first circular, the second a decagon in plan. S. George at Salonica is a later circular example. An early instance of the second variety is found in S. Constanza at Rome, and a considerable number of similar churches occur in Asia Minor, dating from the time of Constantine the Great or a little later. III. The Cross plan. Here we have a square central area covered by a dome, from which extend four vaulted arms constituting a cross. This type also assumes two distinct forms: (1) Buildings in which the ground plan is cruciform, so that the cross shows externally at the ground level. Churches of this class are usually small, and were probably sepulchral chapels rather than churches for public worship. A good example is the tomb of Galla Placidia at Ravenna. (2) In the second form of the Cross church the cross is enclosed within a square, and appears only above the roofs of the angle chambers. An example is seen in the late Roman tomb at Kusr en NÛeijÎs in Eastern Palestine. In this instance the central square area is covered with a dome on continuous pendentives; the four arms have barrel-vaults, and the angles of the cross are occupied by small chambers, which bring the ground-plan to the square. The building is assigned to the second century, and shows that true though continuous pendentives were known at an early date10 (Fig. 8). Another example is the Praetorium at Musmiyeh, in Syria, 11 which probably dates from between 160 and 169 A.D. At some later time it was altered to a church, and by a curious foreshadowing of the late Byzantine plan the walls of the internal cross have entirely disappeared from the ground-plan. The dome rests on four columns placed at the inner angles of the cross, and the vaulted cross arms rest on lintels spanning the space between the columns and the outer walls. From these three types of building are derived the various schemes on which the churches of the Byzantine Empire were planned. Of the basilican form the only example in Constantinople that retains its original plan is S. John the Baptist of the Studion (p. 56), erected c. 463 A.D. The church of SS. Sergius and Bacchus (p. 70) and the baptistery of S. Sophia (p. 78) represent respectively the two varieties of the octagonal plan. In the former the dome rests on piers surrounded by an ambulatory; in the latter the dome rests upon the outer walls of the buildings. Both are foundations of Justinian the Great. Of the Cross church plan showing the cross externally at the ground level no example survives in the city. But at least one church of that form was seen at Constantinople in the case of the church of the Holy Apostles. This was essentially a mausoleum, built originally by Constantine the Great and reconstructed by Justinian to contain the sarcophagi of the sovereigns and the patriarchs of New Rome.12 The church of S. Mark at Venice was built on the plan of the Holy Apostles. It is a cruciform church with aisles, but the galleries which might have been expected above them are omitted. The central dome rests on four piers, and four smaller domes cover the arms. Professor Strzygowski gives examples of cross-planned cells in the catacombs of Palmyra,13 and in many Eastern rock tombs.14 Such cross plans are found also in the Roman catacombs. These subterranean chapels, of course, do not show the external treatment, yet there can be little doubt that the external cross plan was originally sepulchral, and owes its peculiar system of planning to that fact. On the other hand, it was adopted in such churches as S. Mark's at Venice and in the French examples of PÉrigord for aesthetic or traditional reasons. In passing now to a consideration of the distinct forms developed from these pre-Byzantine types of church building, the classification adopted by Professor Strzygowski may be followed. In his Kleinasien he has brought forward a series of buildings which show the manner in which a dome was fitted to the oblong basilica, producing the domed basilica (KÜppelbasilica), an evolution which he regards as Hellenistic and Eastern. In contrast to this, Strzygowski distinguishes the domed cross church (KreutzkÜppelkirche), of which S. Theodosia in Constantinople (p. 170) is the typical example and which is a Western development. A comparison of the two forms is of great importance for the study of certain Constantinople churches. Fig. 1.—Kasr Ibn Wardan (Strzygowski). The domed basilica, as the name indicates, is a basilica with nave and aisles, in which a square bay in the centre of the nave is covered by a dome on pendentives. To north and south, within the arches supporting the dome, appear the nave and gallery arcades of the basilica; and as the galleried basilica is a usual Eastern form galleries are usual in the domed basilica. As seen from the central area, therefore, the north and south dome arches are filled in with arcades in two stories, and the side aisles and galleries are covered with barrel vaults running parallel to the axis of the church. At the west end a gallery over the narthex may unite the two side galleries. At Kasr ibn Wardan, instanced by Strzygowski as a typical domed basilica,15 there is such a western gallery (Fig. 1). According to Strzygowski the domed basilica is older than the fifth century. The domed basilica remains always an oblong building, and whilst the two sides to north and south are symmetrical, the western end retains the basilican characteristics—it has no gallery or arcade communicating with the central area. The narthex communicates with the nave by doors, and if a gallery is placed above it, both narthex and gallery are covered by barrel vaults. In the domed cross church (KreutzkÜppelkirche) the central dome rests on barrel vaults which extend to the outer walls of the building and form the arms of the cross, the eastern arm forming the bema. The lighting of the church is by windows in the gable walls which terminate the north, south, and west cross arms. The prothesis and diaconicon open off the side arms, and two small chambers in the western angles of the cross bring the plan externally to the usual rectangular form. The domed cross church may have galleries, as in S. Theodosia (p. 170), or may be without them, as in SS. Peter and Mark (p. 193). Where galleries are present they are placed in the cross arms and are supported by arcades at the ground level. The vaults beneath the galleries are cross-groined. The domed cross church is a centrally planned church, in contrast to the domed basilica, which is oblong, and therefore we should expect that where galleries are used they will be formed in all three arms of the cross, as is the case in S. Theodosia. There are a number of churches which vary from these types, but which can generally be placed in one class or the other by the consideration of two main characteristics: if the dome arches extend to the outer walls the building is a domed cross church; if the galleries are screened off from the central area by arcades the building is a domed basilica. The church at DerÈ Aghsy,16 for instance, if we had only the plan to guide us, would appear to be a typical domed basilica (Fig. 2), but on examining the section we find that the north and south dome arches extend over the galleries to the outer walls and form cross arms (Fig. 3). The building is, in fact, a domed cross church with no gallery in the western arm. Above the narthex at the west end, and separated from the western cross arm, is a gallery of the type usual in the domed basilica, so that DerÈ Aghsy may be regarded as a domed cross church with features derived from the domed basilica. S. Sophia at Constantinople, the highest development of the domed basilica, has a very similar western gallery. Fig. 2.—DerÉ Aghsy (Rott). Fig. 2.—DerÉ Aghsy (Rott). The church of S. Nicholas at Myra 17 (Fig. 4) has a gallery at the west end, but the cross arms do not appear to be carried over the galleries. The plan is oblong and the cross-groined vault is not used. The church, therefore, takes its place as a domed basilica. Fig. 3.—DerÉ Aghsy—Section (Rott). Fig. 3.—DerÉ Aghsy—Section (Rott). The church of the Koimesis at Nicaea 18 (Figs. 5 and 6) has no galleries to the sides. The aisles open into the central area by arcades, above which are triple windows over the aisle vaults. At the western end is a gallery above the narthex. The aisles are barrel-vaulted, and as the church is planned on an axis from east to west, and is not symmetrical on all three sides, it is regarded as a domed basilica. It is such a form as might be developed from a basilica without galleries. Fig. 4.—S. Nicholas, Myra (Rott). Fig. 4.—S. Nicholas, Myra (Rott). In Constantinople there are three churches which seem to constitute a type apart, though resembling in many ways the types just considered. They are S. Andrew in Krisei, (p. 117), S. Mary Pammakaristos (p. 150), and S. Mary Panachrantos (p. 130). In these churches, as originally built, the central dome is carried on four arches which rise above a one-storied aisle or ambulatory, allowing of windows in the dome arches on three sides—the eastern dome arch being prolonged to form the bema. The dome arches have arcades communicating with the ambulatory on the north, south, and west. The vaulting is executed either with barrel or with cross-groined vaults. These churches are evidently planned from a centre, not, like the domed basilicas, from a longitudinal axis. At the same time the absence of any cross arms differentiates them from the domed cross churches. S. Andrew, which still retains its western arcade, dates from at least the sixth century, so that the type was in use during the great period of Byzantine architecture. Indeed, we should be inclined to regard S. Andrew as a square form of SS. Sergius and Bacchus, but without galleries. The type is a natural development from the octagonal domed church with its surrounding ambulatory. Fig. 5.—The Church of the Koimesis, Nicaea (Wulf). Fig. 5.—The Church of the Koimesis, Nicaea (Wulf). The typical late Byzantine church is a development from the domed cross plan. In three examples in Constantinople, S. Theodosia (pp. 170, 172), S. Mary Diaconissa (p. 185), and SS. Peter and Mark (p. 193), we can trace the gradual disappearance of the galleries. S. Theodosia, as has already been mentioned, has galleries in all three cross arms. In S. Mary Diaconissa they are confined to the four angles between the cross arms; SS. Peter and Mark is a simple cross plan without galleries. In later times it became customary to build many small churches, with the result that the chambers at the angles of the cross, of little account even in a large church, were now too diminutive to be of any value, and the question how to provide as much room as possible for the worshippers became paramount. Accordingly the dome piers were reduced to mere columns connected with the outer walls of the building by arches; and thus was produced the typical late Byzantine plan—at the ground level a square, enclosing four columns; above, a Greek cross with a dome on the centre. Fig. 6.—The Church of the Koimesis, Nicaea (Rott). Fig. 6.—The Church of the Koimesis, Nicaea (Rott). From its distinguishing feature this type has been styled the 'four column' plan. It appears in many Constantinopolitan churches, as, for example, S. Theodore (p. 248) and S. Saviour Pantepoptes (p. 214). The cross arms are not always equal, and may be covered with barrel vaults (p. 214 or with cross-groined vaults (p. 198). The bema is usually a bay added to the eastern arm. The angle chambers have either cross-groined vaults or flat dome vaults. In general the churches of this type in Constantinople do not differ from the numerous churches of the same class in the provinces.19 A lobed cruciform plan is found in only one church in Constantinople, that of S. Mary of the Mongols (p. 277). Here the central dome is supported on four piers set across the angles of the square, so that the pendentives do not come to a point as usual, but spring from the face of the piers. Against each side of the square a semidome is set, thus producing a quatrefoil plan at the vaulting level. Both trefoiled and quatrefoiled churches are not uncommon in Armenia, such as the cathedral at Etschmiadzin; 20 trefoiled churches of a later date are found in the western provinces, and examples have been published from Servia,21 Salonica, 22 and Greece. 23 An unusual form of the cross plan is seen in the building known as Sanjakdar Mesjedi (p. 267), where a cross is placed within an octagon. Probably the building was not originally a church. It resembles the octagon near the Pantokrator (p. 270), and may, like it, have been a library. Single Hall Churches.—The plans hitherto considered have all been characterised by the presence of aisles, galleries, or other spaces adjoining the central area. The churches of the present class consist simply of an oblong hall, terminating in an apse, and either roofed in wood, or covered with domes placed longitudinally, and resting to north and south on wall arches. Examples of this plan are found in Monastir Mesjedi (p. 264), S. Thekla (p. 211), Bogdan Serai (p. 284), and in the memorial chapels attached to the Pantokrator (p. 235), and the Chora (p. 309). In the case of these two memorial chapels, their narrow, long-stretched plan is evidently due to the desire to keep their eastern apses in line with the east end of the churches they adjoin, and at the same time to bring the western end to the narthex from which they were entered. They are covered with two domes, a system perhaps derived from S. Irene (p. 94). KefelÉ Mesjedi (p. 257), which at first sight resembles a single hall church roofed, in wood, was a refectory. Its plan may be compared with that of the refectory at the monastery of S. Luke at Stiris. 24 II. Architectural Features and Details Apses.—A fully developed Byzantine church terminated in three apses: a large apse, with the bema or presbytery, in the centre; on the right, the apse of the prothesis where the sacrament was prepared; on the left, the apse of the diaconicon, where the sacred vessels were kept. Although there is proof that the prothesis and the diaconicon were in use at a very early period, yet many churches of the great period, as for example S. John of the Studion, SS. Sergius and Bacchus, and S. Sophia, dispensed with these chambers as distinct parts of the building. They were also omitted in small churches of a late date, where they were replaced by niches on either side of the bema. The three apses usually project from the east wall of the church, but occasionally p. 248) the two lateral apses are sunk in the wall, and only the central apse shows on the exterior. As a rule the apses are circular within and polygonal without. It is rare to find them circular on both the interior and the exterior (p. 203), and in Greece such a feature is generally an indication of late date. An octagonal plan, in which three sides of the octagon appear, sometimes with short returns to the wall, is the most common; but in later churches polygons of more sides are used, especially for the central apse, and these are often very irregularly set out. Some of the churches of Constantinople show five, and even seven sides. Bema.—The bema is rectangular, and sometimes has concave niches on each side (p. 130). It is covered either with a barrel or with a cross-groined vault, and communicates with the prothesis and the diaconicon. Prothesis and Diaconicon.—These chambers are either square (p. 214) or have a long limb to the east resembling a miniature bema (p. 214). They are lower than the central apse and the cross arms, so that the cruciform figure of the church shows clearly above them on the exterior,25 though in some churches with galleries small chapels overlooking the bema are placed above them at the gallery level (S. Theodosia). They have usually a niche on three sides, and are either dome vaulted or have cross-groined vaults. The combination of a cross-groined vault with four niches springing from the vaulting level is particularly effective. In S. Saviour in the Chora (p. 307) these chambers are covered with drum domes, pierced with windows, but this treatment is quite exceptional. The Gynecaeum.—In the development of church building, the gynecaeum, or gallery for women, tends to become less and less important. In S. Sophia, S. Irene, and S. Theodosia, the gallery is a part of the structure. In S. Mary Diaconissa (p. 185) it is reduced to four boxes at the angles of the cross, while in S. Mary Pammakaristos and SS. Peter and Mark it is absent (pp. 149, 193). But though no longer a structural part of the church, a gynecaeum appears over the narthex in the latest type of church (p. 215). It is generally vaulted in three bays, corresponding to the three bays of the narthex below, and opens by three arches into the centre cross arm of the church and into the aisles. The Narthex.—Unlike the gynecaeum, the narthex tends in later times to become of greater importance, and to add a narthex was a favourite method of increasing the size of a church. In basilican churches, like S. John of the Studion, the narthex was a long hall in three bays annexed to the west side of the building, and formed the east side of the atrium. In domed cross churches with galleries the passage under the western gallery was used as a narthex, being cut off from the central area by the screen arcade which supported the gallery. Such a narthex has been styled a 'structural narthex,' as forming an essential part of the central building. It occurs in several of the churches of the city (p. 114). In domed cross churches without galleries, and in churches of the 'four column' type, neither narthex nor gallery was possible within the cross, and accordingly the narthex was added to the west end. It is usually in three bays and opens into the aisles and central area. Frequently the ends of the narthex terminate in shallow niches (p. 198). In many churches a second narthex was added (p. 166) to the first, sometimes projecting an additional bay at each end, and communicating with halls or chapels on the north or south, or on both sides of the church (p. 128). S. Mark's at Venice presents a fine example of such an extension of the narthex. When a church could not be sufficiently enlarged by additional narthexes, a second church was built alongside the first, and both churches were joined by a narthex which extended along the front of the two buildings. S. Mary Panachrantos (p. 128) is a good example of how a church could be thus enlarged from a simple square building into a maze of passages and domes. The Interior.—The natural division, in height, of an early church, whether basilican or domical, was into three stories—the ground level, the gallery level, and the clearstory or vault level. In the West these structural divisions were developed into the triple composition of nave-arcade, triforium, and clearstory. In the East, in conjunction with the dome, these divisions survive in many examples of the later period. Still, Byzantine architecture was more concerned with spaces than with lines. Large surfaces for marble, painting, or mosaic were of prime importance, and with the disappearance of the gallery the string-course marking the level of the gallery also tended to disappear. In churches with galleries, like S. Theodosia (p. 170) and S. Mary Diaconissa (p. 185), the string-courses fulfil their function, the first marking the gallery level, the second the springing of the vault. In SS. Peter and Mark (p. 193), which has no gallery, there is only one string-course, corresponding in level to the original gallery string-course; accordingly the main arches are highly stilted above it. The absence of the second string-course is a faulty development, for a string-course at the vault level would be a functional member, whereas at the gallery level it is meaningless. In the Panachrantos (p. 130), as well as in other churches without a gallery, the gallery string-course is omitted by a more logical development, and the string-course at the springing of the vault is retained. Openings which do not cut into the vault are then frankly arched, without impost moulding of any kind. Simple vaulted halls, narthexes, and passages have usually a string-course at the vaulting level, broken round shallow pilasters as at the Chora, S. Theodosia, and the Myrelaion. Sometimes the string-courses or the pilasters or both are omitted, and their places are respectively taken by horizontal and vertical bands. Decorative pilasters flush with the wall are employed in the marble incrustation of S. Sophia. In churches of the 'four column' type the full triple division is common but with a change in purpose. A gallery in a church of this character is not possible, for the piers between which the gallery was placed have dwindled into single shafts. Hence the first string-course ceases to mark a gallery level and becomes the abacus level of the dome columns, as in the north and in the south churches of the Pantokrator. It is then carried round the building, and forms the impost moulding of the side arches in the bema and of the east window. Sometimes, however, it does not extend round the bema and apse but is confined to the central part of the church, as in the Myrelaion, S. Theodore, and the Pantepoptes. On the other hand, in at least one case, the parecclesion of the Pammakaristos, the central part of the chapel is designed in the usual three tiers, but the apse and bema vaults spring from the lower or abacus string-course, leaving a lunette in the dome arch above pierced by a large window. A corresponding lunette at the west end opens into the gynecaeum of the chapel. In S. John in Trullo the two string-courses coalesce and the arches connecting the columns with the walls cut into the stilted part of the dome arches, with the result that all the structural arches and vaults spring from the same level. Fig. 7. Map of Byzantine Constantinople. Fig. 7. Arches.—Though the pointed arch was known and employed in cisterns, as in the Cistern of the One Thousand and One Columns, Bin-bir-derek, the circular arch is invariably found in work meant to be seen. The difficulty attending this form, in which arches of unequal breadth do not rise to the same height, was overcome, as in the West, by stilting, that is, by raising the smaller arches on straight 'legs' to the required height. The stilted arch, indeed, seems to have been admired for its own sake, as we find it used almost universally both in vaulting and in decorative arches even where it was not structurally required. In windows and in the arches connecting the dome columns to the wall stilting is sometimes carried to extremes. Domes.—The eastern dome of S. Irene, erected about 740 A.D., is generally considered to be the first example of a dome built on a high drum, though S. Sophia of Salonica, an earlier structure, has a low imperfect drum. After this date the characteristics of the Byzantine dome are the high drum divided by ribs or hollow segments on the interior, polygonal on the exterior, and crowned by a cornice which is arched over the windows.26 Drumless domes are sometimes found in the later churches, as in the narthexes of the Panachrantos and S. Andrew, the angle domes of S. Theodosia, and in Bogdan Serai. These are ribless hemispherical domes of the type shown in Fig. 8, and are in all cases without windows. The earlier system of piercing windows through the dome does not occur in the later churches, though characteristic of Turkish work. The three diagrams (Figs. 8, 9, and 10) illustrate the development of the dome: firstly, the low saucer dome or dome-vault in which dome and pendentives are part of the same spherical surface; secondly, the hemispherical dome on pendentives; and thirdly, the hemispherical dome with a drum interposed between it and the pendentives. Fig. 8.—The Saucer Dome Or Dome-Vault. Fig. 8.—The Saucer Dome Or Dome-Vault. Flat external cornices on the dome are not uncommon in the later churches of Byzantine Greece, as in S. Sophia at Monemvasia.27 In Constantinople only one dome with a flat cornice can be regarded as original, that of S. John in Trullo, a church which is exceptional also in other respects. The many other domes in the churches of Constantinople on high drums and with flat cornices are Turkish either in whole or in part. The high ribless domes of the Panachrantos, for instance, circular in plan within and without, with square-headed windows, plain stone sill, and flat cornice in moulded plaster, may be regarded as typical Turkish drum-domes. As will appear in the sequel, the dome over the north church of the Pantokrator and the domes of SS. Peter and Mark, the Diaconissa, and S. Theodosia, are also Turkish. Fig. 9.—The Dome on Pendentives. Fig. 9.—The Dome on Pendentives. It is most unfortunate that the domes of these three domed cross churches have been altered, especially as the domes of S. Mary Diaconissa and S. Theodosia are larger than any of the later domes except the large oval dome on the central church of the Pantokrator which is almost of the same size. It is therefore now difficult to say what was the precise form of the original domes. Most probably they were polygonal drum-domes, and their collapse owing to their size may well have led to the small drum-domes of later times. Though not strictly Byzantine these Turkish domes are of interest as showing the development of Byzantine forms under Turkish rule, and that reversion to the earlier drumless dome which is so marked a feature of the imperial mosques of the city. Fig. 10.—The Drum Dome. Fig. 10.—The Drum Dome. Domes are either eight, twelve, or sixteen sided, and usually have a window in each side. These numbers arise naturally from setting a window at each of the cardinal points and then placing one, two, or three windows between, according to the size of the dome. Internally the compartments are separated by broad, flat ribs, or are concave and form a series of ridges on the dome which die out towards the crown. In sixteen-sided domes of the latter type the alternate sides sometimes correspond to the piers outside, so that the dome which has sixteen sides within shows only eight sides without, as in the narthex of S. Theodore (p. 246). The octagonal dome of the Myrelaion (p. 198) seems to have had only four windows from the beginning. The ribs of a Byzantine dome are not constructive in the same way as are the ribs of a Gothic vault. They were built along with the rest of the dome and of the same material, and are in no way separate from the infilling, though they no doubt strengthened the shell of the dome by their form28. On the outside a circular shaft with a very simple cap is often placed at the angles of the piers, and from these shafts the brick cornice springs in a series of arches over the windows. Sometimes the angle is formed by a point between two half-shafts, as in the domes of the narthex in S. Theodore (p. 246). External Treatment.—In the older churches the exterior seems to have been left in simple masses of brickwork, impressive only by their size and proportion. Probably even this effect was not considered of great importance. In later times a very beautiful system of decoration with slender shallow niches was introduced and was applied in particular to the east end and to the apses. The finest examples of this system on a large scale are seen at the Pantokrator (p. 235) and S. Theodosia (p. 173). Carefully considered or elaborate external compositions are rare, and the only examples in Constantinople are the side chapel of the Pammakaristos (p. 154) and the narthex of S. Theodore (p. 246). External Marble and Mosaic.—Marble and mosaic, we have reason to know, were occasionally used on the exterior of churches,29 though no fragments remain. On the south side of the Pantepoptes (p. 216) the string-course does not correspond to the line of the walls, but projects in a manner which shows that marble must have been employed to line the large windows. A similar projection of the string-course or cornice is not uncommon elsewhere, though not so evident as in the Pantepoptes, and may have been made to receive a marble or mosaic lining. Doors and Windows.—It is a primary rule in Byzantine architecture that all constructive openings are arched. Whatever may be the eventual form of a door or window the opening is first built in brick with a semicircular head, and into this opening the marble jambs and lining are fitted leaving a semicircular lunette above. Doors are square-headed, with heavily moulded architraves and cornice, and the lintel is mitred into the jambs instead of having the more constructive horizontal joint used in the West. The doors made of wood or of wood lined with bronze, swing on top and bottom pivots which turned in bronze-lined sockets in lintel and threshold. They closed with a rebate in the jambs and against the raised threshold. Windows were sometimes filled in a similar manner, as in the palace of the Porphyrogenitus and in the north gallery of S. Saviour in the Chora (Fig. 100). In the latter double windows or shutters were employed, opening inwards in the same way as did the doors. These shutters may perhaps be regarded as domestic, for in the churches, as is still seen in S. Sophia though the arrangement has vanished elsewhere, the entire arched opening was usually filled in with a pierced marble grille. In addition to the simple round-headed windows double and triple windows are found. Double windows were naturally formed by dividing the single arch by a central pier. This method presented two varieties: either the pier was continued up to the containing arch, thus giving two pointed lights, or the two lights were covered by separate arches within the main arch. Both methods are used in the narthex of S. Theodore (p. 247). Another variety was produced by placing two single lights together, with a shaft between them instead of the central pier. But as double windows are not very satisfactory, triple windows are more common. In this case both the methods just described of forming the windows were adopted. A large semicircular opening divided by two piers will give an arched light between two pointed lights, or three arched lights, as in the narthex of S. Theodore. In the former case, if shafts are substituted for the piers, a little adjustment will produce the beautiful form found in the side-chapels of the Pammakaristos (p. 152), and of S. Saviour in the Chora (p. 310), where the two side lights are covered by half-arches whose crowns abut on the capitals of the shafts, while between and above them rises the semicircular head of the central light. The method of grouping three arched windows of the same height is adopted in apse windows, each of them occupying one side of the exterior. As the deep, narrow mullions are set radiating, the arch is narrower inside than outside. But this difficulty was overcome, partly by lowering the inner crowns, so that the arch is conical, partly by winding the surface. In the Pantokrator (p. 238), instead of radiating to the centre of the apse, the side and mullions are placed parallel to the axis of the church, thus obviating all difficulty. Generally the centre to which the mullions radiate is considerably beyond the apse, so that any necessary little adjustment of the arch could easily be made. Triple windows supported on circular columns are not infrequent in the north and south cross arms. Sometimes the central light is larger than the lateral lights, at other times, as in the Pantepoptes, the three lights are equal. The lower part of these windows was probably filled in with a breastwork of carved slabs, as in S. Sophia, while the upper part was filled by a pierced grille. At present the existing examples of these windows have been built up to the abaci of the capitals, but in the church of S. Mary Diaconissa (p. 186) the columns still show the original form on the inside. Vaulting.—All Byzantine churches of any importance are vaulted in brick. The only exception to this rule in Constantinople is the little church known as Monastir Mesjedi (p. 264). The different systems of Byzantine vaulting have been so fully treated by Choisy and other authorities, that in the absence of any large amount of new material it is not necessary to give here more than a few notes on the application of these systems in Constantinople. It should always be kept in view that, as these vaults were constructed with the lightest of centering, the surfaces and curves must have been largely determined by the mason as he built, and would not necessarily follow any definite geometrical development. "Il serait illusoire," remarks Choisy, "d'attribuer À toutes les voutes byzantines un trace gÉomÉtrique rigoureusement dÉfini." 30 PLATE II. The Myrelaion (since it was burned), from the north-west. The Myrelaion (since it was burned), from the north-west. The Myrelaion (since it was burned), from the north-east. The Myrelaion (since it was burned), from the south-east. (By kind permission of H. M. Dwight, Esq.) To face page 20. The vaults commonly found are the barrel vault, the cross-groined vault, and the dome-vault. The first is frequently used over the cross arms and the bema, and sometimes over the narthex in conjunction with the groined vault (Diaconissa). It is the simplest method of covering an oblong space, but it does not easily admit of side windows above the springing. A very beautiful form of cross-groined vault is found in S. Sophia and in SS. Sergius and Bacchus, in which the crown is considerably domed, and the groins, accordingly, lose themselves in the vaulting surface. This form is found in Greek churches of late date, but does not occur in the later churches of Constantinople. A full description of the form and construction is given by Choisy31 and by Lethaby and Swainson.32 The cross-groined vault as found in the Myrelaion and many other churches of the city is level in the crown, with clearly marked groins. It is sometimes used with transverse arches resting on pilasters, or without these adjuncts. One of the most interesting of the vault forms is the dome-vault, a shallow dome with continuous pendentives. It is distinguished in appearance from the groined vault, as found in S. Sophia, by the absence of any groin line, and is completely different in construction. The geometrical construction is that of the pendentives of all domes. The four supporting arches intersect a hemispherical surface whose diameter is equal to the diagonal of the supporting square. The pendentives produce at the crown line of the arches a circular plan which is filled in by a saucer dome of the same radius as the pendentives, constructed of circular brick rings, the joints of which radiate to the centre. If the space to be covered is not square the broader arches intersect at a higher level, while the narrow arches are not stilted, but kept down so as to receive the dome surface, and in this case the narrow arches are not semicircular, but segmental. Where the difference in size between the two sides was not great, the difficulty presented was easily overcome by the Byzantine builder, who in the later buildings, at any rate, rarely built anything within four inches of its geometrical position. Where the difference was too great it was frankly accepted, and we find segmental arches at the narrow ends. Fig. 11. Diagram of Vaulting in Outer Narthex of S. Saviour in the Chora. Fig. 11. The vaulting of the outer narthex of S. Saviour in the Chora illustrates this fully (Fig. 11). Though some of the bays of that narthex are oblong and others almost square all are covered with dome vaults. The almost square bays, although their sides vary considerably, are covered precisely as if their sides were exactly equal. But in two of the oblong bays, which are nearly three times as long as they are broad, such a method could not be applied. Longitudinal arches (AA) were accordingly thrown between the transverse arches (CC) and made to rest on their spandrils. The oblong form of the intervening space was thus very much reduced, and over it flat domes are thrown. Their rings are true circles, and as the space they cover is still somewhat oblong they descend lower, with additional segments of rings (BB), at the ends than at the sides. In the remaining two oblong bays of the narthex, the result of introducing the longitudinal arches is to convert a decidedly oblong space in one direction into a slightly oblong space in the opposite direction, an additional proof, if any were needed, that the exact shape of plan with this form of vault was a matter of comparative indifference to the builder. In S. Sophia the vault springs from the intrados of the transverse arches, that is, from the lower edge. In SS. Sergius and Bacchus it springs from a point so slightly raised as to be hardly noticeable. In the later vaults, however, the transverse arches, when present, are boldly shown, and the vault springs from the extrados or outer edge (e.g. S. Saviour in the Chora, S. Theodore). Construction.—Most of the churches of the city are covered with thick coats of plaster and whitewash, both within and without. Only in a few cases, where these coatings have fallen away through neglect, or in some remote corner of a building to which these coatings were never applied, can the construction and the laying of the brickwork be studied. The two-storied chapel, known as Bogdan Serai (p. 283), is almost denuded of plaster, and is therefore of importance in this connection. The bricks of the wall arches on which its dome rests are laid considerably flatter than the true radiating line, leaving a triangular piece to be filled in at the crown. On the other hand, the bricks of the transverse arches under the dome radiate to the centre. It has been supposed that the method followed in the wall arches was employed in order to economise centering, since bricks could gradually be worked out over the space, each course simply sticking to the one below. This is undoubtedly the case in some examples. But here centering could not have been of any service in the wall arches, and the transverse arches are laid without flattening of the courses, though that arrangement might have been useful in their case. It is therefore more probable that the flattening of the courses in the wall arches is simply a piece of careless workmanship. The pendentives, like all pendentives that could be examined, were formed of horizontal courses corbelled out to the circle. The dome, bema, and the barrel vault in the lower story (p. 285) seem to be laid with true radiating joints. The springing of the barrel vault is formed of four courses of stone laid horizontally and cut to the circle, and above them the entire barrel is of brick. The dome arches of the Sanjakdar Mesjedi (p. 270) are formed of three distinct rings, not bonded into one another. They radiate to the true centre, and the pendentives are, as usual, in horizontal courses. The transverse arches of the outer narthex in S. Saviour in the Chora are also built with true radiating courses. The gynecaeum of the side-chapel of the Pammakaristos (p. 153) has never been plastered, and consequently the laying of the brickwork can be seen there to advantage. The little stair leading up to the gallery is covered with a sloping barrel vault built in segments perpendicular to the slope of the stair and could easily have been built without centering. The same remark applies to the cross vault at the head of the stair, which is similarly constructed in 'slices' parallel to each side (p. 154). The arches of the gynecaeum itself, the vaults, and the two little domes, seem to have true radiating joints. The ribs of the domes are formed in the brickwork, and are not structurally separate. In these last examples, and in all door and window openings, in which the joints invariably radiate from the centre, a certain amount of centering was inevitable. The Myrelaion (since it was burnt). Interior, looking east. (By kind permission of H. M. Dwight, Esq.)

Suleiman Aga Mesjedi, beside S. Saviour Pantokrator.. Suleiman Aga Mesjedi, beside S. Saviour Pantokrator.

To face page 24.

On the other hand a little passage in S. Saviour in the Chora between the church and the parecclesion (p. 311), is covered with a barrel vault evidently built without centering. The space is first narrowed by two corbelled courses of stone and, above them, by three projecting courses of brick. From this springs the vault, built from each end in strongly inclined segments. These segments meet in the middle, leaving a diamond-shaped space filled in with longitudinal courses. Like the stairs in the Pammakaristos, this passage is very narrow, some 85 cm., yet the builders thought it necessary to corbel out five courses before venturing to throw a vault without centering.

Near the Pantokrator is an octagonal building, now Suleiman Aga Mesjedi but generally regarded as a Byzantine library, which has on each side a large wall arch strongly elliptical in form (p. 270). Two arches of somewhat similar form and apparently original are found in the south end of the gynecaeum of the Pantokrator (p. 237). These arches may have been built in this manner to economise centering. Still, in the library they are wall arches easily constructed without centering.

Failing the examination of a larger number of buildings in Constantinople we can hardly judge of the later methods of vault and arch construction, but one point may be further noticed. The wall internally is often set back slightly at each spring course, so that with the projection of the course a considerable ledge or shelf is left. On this ledge centering could easily be supported and would have required no further framework to the ground. Centering seems to have been used for dorm, arches, vaults, and door and window openings. It was not used in small vaults. But it is difficult to imagine any method of constructing such groined vaults as those found in the narthexes of the Pantokrator without a very considerable amount of centering.

Ties.—As a general rule tie rods or beams were used, either of iron or wood. In the latter case they were painted with leaf or fret ornaments, and were evidently considered as natural features. But large vaults are often found without such ties as in the narthex of the Pantokrator. Many churches have ties to the dome-arches, and none to the main vault; but it is difficult to lay down a fixed rule. The enormous amount of mortar in the walls must have made them yield to a certain degree when newly built, and some of the larger vaults would have been the better for rods.

Abutments.—The system of abutments in the Byzantine churches of the great period has been carefully studied by M. Choisy. 33 In early examples the dome springs directly from the pendentives on the inside, but is thickened externally over the haunches, producing a double curve and an apparent drum. This is seen very clearly in SS. Sergius and Bacchus. In S. Sophia the numerous windows are cut through this drum, so that it resembles rather a series of small abutments. The object was to support the crown of the dome by adding weight over the haunches. In both these churches the thrust of the dome and its supporting arches is taken by the two-storied galleries, which form, in fact, flying buttresses within the buildings, and are adapted to their architectural requirements. The square plan and the enormous size of the dome in S. Sophia demanded the great buttresses on the sides; while in SS. Sergius and Bacchus the eight buttresses show only on the outside of the dome and are not carried over the aisles as they are in S. Sophia. Below the roof the arches and piers of the galleries and aisles are arranged so as to carry the thrust to the external walls, and following the tradition of Roman vaulting all buttressing is internal. In S. Irene, where the true drum dome first appears, the buttresses between the windows of the dome still remain, though much reduced in size. A dome raised on a drum can evidently no longer exercise a thrust against the dome-arches; its thrust must be taken by the drum, and only its weight can rest on the arches.

The weight of the drum and dome rests on the pendentives and dome-arches. Their thrust is neutralized by the use of ties and by the barrel vaults of the cross arms, and these in their turn depend on the thickness of the walls. The lower buildings attached to the church in the form of side-chapels and the narthex also helped to stiffen and buttress the cross walls. The system is by no means perfect in these late churches. It was apparently found impossible to construct drum domes of any size, except at the extreme risk of their falling in, and probably it is for this reason that many of the larger domes in late churches, like SS. Peter and Mark, S. Theodosia, the Chora, have fallen. No system of chainage appears to have been used for domes in Constantinople.

Flying buttresses probably of the ninth century are used at the west end of S. Sophia. The double-flying buttress to the apse of the Chora does not bond with the building and is certainly not original. It may be set down as part of the Byzantine restoration of the church in the fourteenth century. In any case, such external flying abutments are alien to the spirit of Byzantine architecture, and may be regarded as an importation from the West. Flying buttresses, it may here be noted, are not uncommon in the great mosques of the city. They are found in Sultan Bayazid, Rustem Pasha, Sultan Selim, the Suleimanieh, and the ShahzadÉ. But they are generally trifling in size, and are rather ornaments than serious attempts to buttress the dome.

Walls.—The walls of the earlier churches are built of large thin bricks laid with mortar joints at least as thick as the bricks, and often of greater thickness. Stone is used only in special cases, as in the main piers of S. Sophia, but monolithic marble columns are an important part of the structure. In the later churches stone is used in courses with the bricks to give a banded effect, and herring-bone, diamond, and radiating patterns are frequently introduced. The palace of the Porphyrogenitus, the parecclesion of the Pammakaristos, and Bogdan Serai, exhibit this style of work. As illustrations of the method adopted in the construction of walls the following measurements may be given, the sizes being in centimetres:

Brick.

Joint.

Parecclesion of the Pammakaristos		.08	.04
4 courses brick, 5 joints		.46	—
S. John in Trullo		.03	.07 to .09
Refectory of the Monastery of Manuel		.04	.04 to .06
4 course stone, 3 joints		.78	—
4 courses brick, 5 joints		.30	—
	paranthesis	.0375	.052
Bogdan Serai		.035	.035
		.04	.04
4 courses stone, 8 joints		—	.55 to .60
4 courses brick, 5 joints		—	.43 to .47
Sanjakdar, brick		.045	—

 

Building Procedure.—The first step in the erection of a building was to obtain the necessary marble columns with their capitals and bases. These seem to have been largely supplied ready made, and Constantinople was a great centre for the manufacture and export of stock architectural features. Then the main walls were built in brick, the columns were inserted as required, the vaults were thrown, and the whole building was left to settle down. Owing to the enormous amount of mortar used this settling must have been very considerable, and explains why hardly a plumb wall exists in Constantinople, and why so many vaults show a pronounced sinking in at the crown or have fallen in and have been rebuilt. After the walls had set the marble facings, mosaic, and colour were applied and could be easily adapted to the irregular lines of the walls.

Byzantine architecture made little use of mouldings. The great extension of flat and spacious decoration rendered unnecessary, or even objectionable, any strong line composition. External cornices are in coursed brick, the alternate courses being laid diagonally so as to form the characteristic dentil. The richest form is that found in the Pammakaristos, S. Theodosia, and S. Thekla, where the small dentil cornice is supported on long tapering corbels, a design suggested by military machicolations.

PLATE IV.

Bracket in S. Saviour in the Chora. S. Saviour in the Chora. Bracket in the Inner Narthex Interior, looking east.	Sculptured Slab in S. Theodore. S. Theodore. Sculptured Marble Slab Built into the Minaret of the Mosque
S. Mary Diaconissa. Heads of Windows in south arm. S. Mary Diaconissa. Heads of Windows in South Arm Interior, looking east.	Sculptured Slab on the West Wall. S. Mary Diaconissa. Sculptured Slab on the West Wall.

To face page 28.

The stone ogee, cavetto, or cavetto and bead cornice is common, but seems in every case to be Turkish work and is very common in Turkish buildings. Internal cornices and string-courses are in marble, and are all of the same type, a splay and fillet. The splayed face is decorated with upright leaves or with a guilloche band, either carved (in the Pantepoptes) or painted (in the Chora), the carving as in classic work, serving only to emphasise the colour. The splay is sometimes slightly hollowed, sometimes, as in the Chora, worked to an ogee.

Doors.—Doors often have elaborately moulded architraves and cornice. In S. John of the Studion (p. 61), the oldest example, the jamb-moulding has a large half-round on the face, with small ogees and fillets, all on a somewhat massive scale. The doors of S. Sophia are very similar. The later mouldings are lighter but the half-round on the face remains a prominent feature. It is now undercut and reduced in size, and resembles the Gothic moulding known as the bowtell. This is combined with series of fillets, small ogees, and cavettos into jamb-moulds of considerable richness. The cornices are often simply splayed or are formed of a series of ogees, fillets, and cavettos. The jamb-mouldings are cut partly on a square and partly on a steep splayed line. In some, the portion forming the ingo seems to have been regarded as a separated piece though cut from the solid. If in the doors of the Pantokrator or the Pantepoptes the line of the inner jamb be continued through the rebate, it will correspond on the outside with the bowtell moulding, as though the inner and outer architrave had been cut from one square-edged block, placing the bowtell at the angle and adding the rebate. This formation is not followed in S. John of the Studion.

Carving.—Carving is slight, and is confined to capitals, string-courses, and the slabs which filled in the lower parts of screens and windows. Fragments of such slabs are found everywhere. They are carved with geometrical interlacing and floral patterns, often encircling a cross or sacred monogram, or with simply a large cross. Such slabs may be seen still in position in S. Sophia and in the narthex of S. Theodore. In the latter they are of verd antique, and are finely carved on both sides. In later times the embargo on figure sculpture was considerably relaxed. Little figures are introduced in the cornices of the eikon frames in the Diaconissa (p. 186), and both in the parecclesion and the outer narthex of the Chora are found many small busts of angels, saints, and warriors carved with great delicacy. The carving in the Chora is the finest work of the kind excepting that in S. Sophia.

Capitals.—The development of the capital from the Roman form, which was suitable only for the lintel, to the impost capital shaped to receive an arch has been well explained by Lethaby and Swainson. According to these authors Byzantine capitals exhibit seven types.

I. The Impost capital.—It is found in SS. Sergius and Bacchus, the outer narthex of the Chora, the inner narthex of S. Andrew and elsewhere. A modification of this type is used in windows. It was employed throughout the style but especially in early times up to the sixth century, and again in the twelfth, thirteenth, and fourteenth centuries.

II. The Melon type.—This is seen on the columns of the lower order in SS. Sergius and Bacchus and on the columns of the narthex of S. Theodore, where they have been taken from an older building. The melon capital was probably not in use after the sixth century.

III. The Bowl capital.—This type is used in the great order of S. Sophia at Constantinople. It has been thought peculiar to this church, but the capitals from S. Stephen at Triglia in Bithynia resemble those of S. Sophia closely. Only the peculiar volutes of the S. Sophia capitals are absent.34

IV. The Byzantine or 'Pseudo-Ionic.'—This is found in the upper order of SS. Sergius and Bacchus, and in the narthex of S. Andrew. It is an early type, not used after the sixth century, and its occurrence in S. Andrew favours the early date assigned to that church.

V. The Bird and Basket.—Found in Constantinople, only in S. Sophia.

VI. The Byzantine Corinthian.—This is the commonest form of capital in the later churches, and must have been in continuous use from the earliest date. It occurs in S. John of the Studion, the Diaconissa, the Chora, and in many other churches. Here the classic form is accurately adhered to, but, as the curved abacus was unsuitable to the arch, a large splayed abacus or impost block is placed above the capital. It is a general feature of the Byzantine capital that it projects at no point beyond the impost line of the arch, thus differing both from the classic and the Gothic forms.

VII. The Windblown Acanthus.—This is found in the churches of Salonica and Ravenna. Three examples are mentioned as seen in Constantinople, two near the Diaconissa, forming bases for the posts of a wooden porch to a house; one is the cistern commonly known as the cistern of Pulcheria.

Window Capitals.—In shafted window of several lights, the impost piers between the arches are of the full thickness of the wall, but are very narrow from side to side. Similarly the shafts are almost slabs placed across the wall, and sometimes, as in the Pammakaristos, are carved on their narrow faces. The capitals are cubical, of slight projection at the sides, but spreading widely at the ends, while the bases closely resemble capitals turned upside down. As with columns, the joints at base and necking are bedded in sheet lead.

Floors.—The floors are usually of thick red brick tiles, some .31 cm. square, or, as in S. Theodore, hexagonal, .34 cm. across by 45 cm. from point to point. Marble floors were used when possible, inlaid with patterns, or in slabs surrounded by borders of coloured marbles, as is still seen in a portion of the floor in the Pantokrator (Fig. 76).

Decoration.—Of the churches of Constantinople only S. Sophia, S. Mary Diaconissa, the South Church of the Pantokrator, and the Chora, retain any considerable part of their original decoration. The first is beyond our present scope, but from the general tone and atmosphere which still linger there we are able to appreciate the effect of the same style of decoration where it survives in less complete form.

The accepted method, as may be observed in the Chora and the Diaconissa, was to split marble slabs so as to form patterns in the veining, and then to place them upright on the wall. It is probable that the finest slabs were first placed in the centre points of the wall, and that other slabs or borders were then arranged round them. The centre slabs in the Chora are of exceptional beauty. The usual design consists of a dado of upright slabs surmounted by panelling to the cornice level, the panels being outlined with plain or carved beads. In the Diaconissa the notched dentil form is used for the beads; in the Chora, a 'bead and reel.' The arches have radiating voussoirs, or, in the Diaconissa, a zigzag embattled design, found also in S. Demetrius of Salonica, though two hundred years must have separated the buildings. In the Chora the arch spandrils and cornice are inlaid with scroll and geometrical designs in black, white, and coloured marbles.

The surfaces above the cornice and the interior of the domes gleamed with mosaic, representing, as seen in the Chora, figures on a gold background. The mosaic cubes are small, measuring 5 mm. to 7 mm., and are closely set. This is about the same size as the mosaic cubes in S. Sophia, but smaller than those at Ravenna, which measure about 10 mm.

Painting.—In the majority of churches this full decoration with marble and mosaic must have been rendered impossible by the expense, and accordingly we find examples like the parecclesion at the Chora decorated with painting, following exactly the tradition of marble and mosaic. This painting is in tempera on the plaster, and is executed with a free and bold touch.

Conclusion.—Byzantine architecture is essentially an art of spaces. 'Architectural' forms, as we are accustomed to think of them, are noticeably absent, but as compensation, colour was an essential and inseparable part of the architecture. The builder provided great uninterrupted spaces broken only by such lines and features as were structurally necessary—capitals, columns, string-courses, and over these spaces the artist spread a glittering robe of marble or mosaic. No school has ever expressed its structure more simply, or given fuller scope to the artist, whether architect or painter.

Byzantine architecture is not only a school of construction, it is also a school of painting. Most of the churches of Constantinople have unfortunately lost the latter part of their personality. They are mere ghosts, their skeletons wrapped in a shroud of whitewash. Still the Greek artist retained his skill to the last, and the decorative work of S. Saviour in the Chora will stand comparison even with the similar work in S. Sophia.

In Byzantine times the greatness of S. Sophia tended to crush competition. No other ecclesiastical building approached the 'Great Church.' But structural ability was only latent, and displayed its old power again in the erection of the imperial mosques of the early Turkish Sultans, for they too are monuments of Greek architectural genius.

The origins of Byzantine architecture have been discussed at great length by Strzygowski, Rivoira, and many other able writers. Much work still remains to be done in the investigation of the later Roman and early Byzantine work; nor does it seem probable that the difficult questions of the Eastern or the Western origin of Byzantine art will ever be finally settled.

The beginnings of Byzantine architecture have never been satisfactorily accounted for. With S. Sophia it springs almost at once into full glory; after S. Sophia comes the long decline. It may, however, be noted that the 'endings' of Roman architecture are similarly obscure. Such buildings as the Colosseum, in which the order is applied to an arched building, are evidently transitional, the Roman construction and the Greek decoration, though joined, not being merged into one perfect style. Even in the baths and other great buildings of Imperial Rome the decoration is still Greek in form and not yet fully adapted to the arched construction. At Spalatro, in such parts as the Porta Aurea, a developed style seems to be on the point of emerging, but it is not too much to say that in no great Roman building do we find a perfect and homogeneous style.

There is nothing in either the planning or the construction of S. Sophia which cannot be derived from the buildings of the Roman Imperial period, with the exception of the pendentive, a feature which had to be evolved before the dome could be used with freedom on any building plan on a square. The great brick-concrete vaulted construction is that of the Roman baths, and with this is united a system of decoration founded on the classic models, but showing no trace of the Greek beam tradition which had ruled in Rome.

S. Sophia then may be regarded as the culminating point of one great Roman-Byzantine school, of which the art of classic Rome shows the rise, and the later Byzantine art the decline. This view is in accord with history, for Constantinople was New Rome, and here, if anywhere, we should expect to find preserved the traditions of Old Rome.

The division of Western Mediaeval Architecture into the two schools of Romanesque and Gothic presents a parallel case. It is now realised that no logical separation can be made between the two so-called styles. Similarly we may continue to speak of the Classic Roman style and of the Byzantine style, although the two really belong to one great era in the history of art.

10 Eastern Palestine Memoirs, p. 172. A similar dome is given by Choisy, L'Art de bÂtir chez les Byzantins, Plate XV.

11De VogÜÉ, Syrie centrale, i. p. 45, Plate VII.

12 DÜrm, Handbuch, Part II. vol. iii. pp. 115, 149. A restored plan is given in Lethaby's Mediaeval Art, p. 47.

13 Orient oder Rom, p. 19.

14 Kleinasien, p. 152.

15 Kleinasien, p. 121 et seq.

16 Oskar Wulf, Die Koimesiskirche in Nikaea, p. 71.

17 H. Rott, Kleinasiensche DenkmÄler, p. 329.

18 Wulf, op. cit. p. 23.

19 For local variations in late churches in Greece, see Traquair's 'Churches of Western Mani,' Annual of British School at Athens, xv. 1908.

20 Strzygowski, 'Das Etschmiadzin Evangeliar,' Byzant. DenkmÄler, i., 1891.

21 Ravanica, F. Kanitz, Serbiens byzantische Monumente, Wien, 1862.

22 Pullan and Texier, S. Elias.

23 G. Lampakis, Les AntiquitÉs chrÉtiennes de la GrÈce, Athens, 1902.

24 Schultz and Barnsley, The Monastery of S. Luke at Stiris, p. 13, fig. 6.

25 See, however, North Church in S. Mary, Panachrantos, p. 128.

26 Strzygowski's views as to the early date of the drum-dome are not universally accepted. The examples he produces seem rather octagons carried up from the ground to give a clearstory under the dome than true drums interposed between the dome and its pendentives.

27 Annual B.S.A. xii. 1905-6. See also Schultz and Barnsley, Monastery of S. Luke at Stiris.

28 See p. 154.

29 Dome of the Rock at Jerusalem. S. Mary Peribleptos; see Vida de Gran Tamorlan y itinerario del Ruy Gonzalez de Clavijo, p. 52.

30 L'Art de bÂtir chez les Byzantins, p. 57.

31 Ibid. p. 99.

32 Sancta Sophia, p. 219.

33 L'Art de bÂtir chez les Byzantins, p. 135.

34 Hasluck, 'Bithynica,' Annual B.S.A. XIII. 1906-7.

---

CHAPTER II

THE CHURCH OF S. JOHN THE BAPTIST OF THE STUDION, EMIR AHOR JAMISSI

The mosque Emir Ahor Jamissi, situated in the quarter of Psamathia, near the modern Greek church of S. Constantine, and at short distance from the Golden Gate (Yedi KoulÉ), is the old church of S. John the Baptist, which was associated with the celebrated monastery of Studius, ἡ μονὴ τοῦ Στουδίου. It may be reached by taking the train from Sirkiji Iskelessi to Psamathia or Yedi KoulÉ. 35

In favour of the identification of the building, there is, first, the authority of tradition, 36 which in the case of a church so famous may be confidently accepted as decisive. In the next place, all indications of the character and position of the Studion, however vague, point to Emir Ahor Jamissi as the representative of that church. For the mosque presents the characteristic features which belonged to the Studion as a basilica of the fifth century, and stands where that sanctuary stood, in the district at the south-western angle of the city,37 and on the left hand of the street leading from S. Mary Peribleptos (Soulou Monastir) to the Golden Gate.38 Furthermore, as held true of the Studion, the mosque is in the vicinity of the Golden Gate,39 and readily accessible from a gate and landing (Narli Kapou) on the shore of the Sea of Marmora.40

According to the historian Theophanes,41 the church was erected in the year 463 by the patrician Studius, after whom the church and the monastery attached to it were named. He is described as a Roman of noble birth and large means who devoted his wealth to the service of God,42 and may safely be identified with Studius who held the consulship in 454 during the reign of Marcian.43

If we may trust the Anonymus, 44 the church erected by Studius replaced a sanctuary which stood at one time, like the Chora, outside the city. Seeing the territory immediately beyond the Constantinian fortifications was well peopled before its inclusion within the city limits by Theodosius II., there is nothing improbable in the existence of such extra-mural sanctuaries, and as most, if not all, of them would be small buildings, they would naturally require enlargement or reconstruction when brought within the wider bounds of the capital. According to Suidas, 45 the building was at first a parochial church; its attachment to a monastery was an after-thought of its founder.

PLATE V.

The Ruined Interior, seen from the minaret of the mosque.

S. John of the Studion.
Ruined Interior, seen from the Minaret of the Mosque.

The West Side of the Church.

S. John of the Studion, from the west.

To face page 36.

The monastery was large and richly endowed, capable of accommodating one thousand monks. 46 Its first inmates were taken from a fraternity known as the Akoimeti, 'the sleepless'; so named because in successive companies they celebrated divine service in their chapels day and night without ceasing, like the worshippers in the courts of heaven.

'Even thus of old Our ancestors, within the still domain Of vast cathedral or conventual church Their vigils kept: where tapers day and night On the dim altar burned continually. In token that the House was ever more Watching to God. Religious men were they; Nor would their reason, tutored to aspire Above this transitory world, allow That there should pass a moment of the year When in their land the Almighty's service ceased.'

But this devout practice does not seem to have been long continued at the Studion; for we never hear of it in any account of the discipline of the House. The monks of the Studion should therefore not be identified with the Akoimeti who took up such a determined and independent attitude in the theological conflicts under Zeno, Basiliscus, and Justinian the Great.47

In the course of its history the church underwent noteworthy repairs on two occasions. It was first taken in hand for that purpose, soon after the middle of the eleventh century,48 by the Emperor Isaac Comnenus (1057-58), who was interested in the House because he and his brother had received part of their education in that 'illustrious and glorious school of virtue.' 49 What the repairs then made exactly involved is unfortunately not stated. But, according to Scylitzes, they were so extensive that 'to tell in detail what the emperor and empress did for the embellishment of the church would surpass the labour of Hercules.' 50 Probably they concerned chiefly the decoration of the edifice.

The next repairs on record were made about the year 1290, in the reign of Andronicus II., by his unfortunate brother Constantine Porphyrogenitus. Owing to the neglect of the building during the Latin occupation the roof had fallen in, the cells of the monks had disappeared, and sheep grazed undisturbed on the grass which covered the grounds. Constantine, rich, generous, fond of popularity, did all in his power to restore the former glory of the venerated shrine. The new roof was a remarkable piece of work; large sums were spent upon the proper accommodation of the monks, and the grounds were enclosed within strong walls. 51

Like other monastic institutions, the Studion suffered greatly at the hands of the iconoclast emperors. Under Constantine Copronymus, indeed, the fraternity was scattered to the winds and practically suppressed, so that only twelve old members of the House were able to take advantage of the permission to return to their former home, upon the first restoration of eikons in 787 by the Empress Irene. Under these circumstances a company of monks, with the famous abbot Theodore at their head, were eventually brought from the monastery of Saccudio to repeople the Studion, and with their advent in 799 the great era in the history of the House began, the number of the monks rising to seven hundred, if not one thousand.52

Theodore had already established a great reputation for sanctity and moral courage. For when Constantine VI. repudiated the Empress Maria and married Theodote, one of her maids of honour, Theodore, though the new empress was his relative, denounced the marriage and the priest who had celebrated it, insisting that moral principles should govern the highest and lowest alike, and for this action he had gladly endured scourging and exile. The Studion had, therefore, a master who feared the face of no man, and who counted the most terrible sufferings as the small dust of the balance when weighed against righteousness, and under him the House became illustrious for its resistance to the tyranny of the civil power in matters affecting faith and morals. When the Emperor Nicephorus ordered the restoration of the priest who had celebrated the marriage of Constantine VI. with Theodote, not only did Theodore and his brother Joseph, bishop of Thessalonica, and their venerable uncle Plato, endure imprisonment and exile, but every monk in the Studion defied the emperor. Summoning the fraternity into his presence, Nicephorus bade all who would obey his order go to the right, and all who dared to disobey him go to the left. Not a single man went to the right. Under the very eyes of the despot all went to the left, and in his wrath Nicephorus broke up the community and distributed the monks among various monasteries. Upon the accession of Michael I. the exiled monks and Theodore were allowed indeed to return to the Studion, peace being restored by the degradation of the priest who had celebrated the obnoxious marriage. But another storm darkened the sky, when Leo V., the Armenian, in 813, renewed the war against eikons. Theodore threw himself into the struggle with all the force of his being as their defender. He challenged the right of the imperial power to interfere with religious questions; he refused to keep silence on the subject; and on Palm Sunday, in 815, led a procession of his monks carrying eikons in their hands in triumph round the monastery grounds. Again he was scourged and banished. But he could not be subdued. By means of a large and active correspondence he continued an incessant and powerful agitation against the iconoclasts of the day. Nor would he come to terms with Michael II., who had married a nun, and who allowed the use of eikons only outside the capital. So Theodore retired, apparently a defeated man, to the monastery of Acritas53; and there, 'on Sunday, 11 November 826, and about noon, feeling his strength fail, he bade them light candles and sing the 119th psalm, which seems to have been sung at funerals. At the words: "I will never forget Thy commandments, for with them Thou hast quickened me," he passed away.' He was buried on the island of Prinkipo, but eighteen years later, when eikons were finally restored in the worship of the Orthodox Church, his body was transferred to the Studion, and laid with great ceremony in the presence of the Empress Theodora beside the graves of his uncle Plato and his brother Joseph, in sign that after all he had conquered.54 Tandem hic quiescit.

Note

His remains were interred at the east end of the southern aisle, where his uncle Plato and his brother Joseph had been buried before him, and where Naucratius and Nicholas, his successors as abbots of the Studion, were laid to rest after him. πρὸς τῷ δεξιῷ μέρει ἐν τῷ κατ' ἀνατολὰς τοῦ Προδρομικοῦ τεμένους πανδόξῳ καὶ ἱερῷ τῶν μαρτύρων σηκῷ, ἔνθα δὴ καὶ τοῦ ὁσίου πατρὸς ἡμῶν Θεοδώρου ἡ πανευκλεὴς καὶ πανσέβαστος τιμία θήκη καθίδρυται (Vita S. Nicolai Studitae, Migne, P.G. tome 105).

There, in fact, during the recent Russian exploration of the church, three coffins were discovered: one containing a single body, another four bodies, and another three bodies. The grave had evidently been disturbed at some time, for some of the bodies had no head, and all the coffins lay under the same bed of mortar. No marks were found by which to identify the persons whose remains were thus brought to view. But there can be no doubt that five of the bodies belonged to the five persons mentioned above. To whom the three other bodies belonged is a matter of pure conjecture. They might be the remains of three intimate friends of Theodore, viz. Athanasius, Euthemius, Timotheus, or more probably of the abbots, Sophronius (851-55), Achilles (858-63), Theodosius (863-64). Cf. Itin. russes, p. 100.

It would be a mistake, however, to think of Theodore only as a controversalist and defier of the civil authority. He was a deeply religious man, a pastor of souls, and he revived the religious and moral life of men, far and wide, not only in his own day, but long after his life on earth had closed. He made the Studion the centre of a great spiritual influence, which never wholly lost the impulse of his personality or the loftiness of his ideal. The forms of mediÆval piety have become antiquated, and they were often empty and vain, but we must not be blind to the fact that they were frequently filled with a passion for holy living, and gave scope for the creation of characters which, notwithstanding their limitations, produced great and good men.

S. John of the Studion.
FaÇade of the Narthex.

To face page 40.

Speaking of Eastern monks and abbots, especially during the eighth and ninth centuries, Mr. Finlay, the historian, justly remarks that 'the manners, the extensive charity, and the pure morality of these abbots, secured them the love and admiration of the people, and tended to disseminate a higher standard of morality than had previously prevailed in Constantinople. This fact must not be overlooked in estimating the various causes which led to the regeneration of the Eastern Empire under the iconoclast emperors. While the Pope winked at the disorders in the palace of Charlemagne, the monks of the East prepared the public mind for the dethronement of Constantine VI. because he obtained an illegal divorce and formed a second marriage. The corruption of monks and the irregularities prevalent in the monasteries of the West contrast strongly with the condition of the Eastern monks.' Certainly to no one is this tribute of praise due more than to the brotherhood in the monastery of Studius.

The monks of the Studion, like most Greek monks, lived under the rules prescribed by S. Basil for the discipline of men who aspired to reach 'the angelic life.' Theodore, however, quickened the spirit which found expression in those rules, and while inculcating asceticism in its extremest form, showed greater consideration for the weakness of human nature. The penalties he assigned for transgressions were on the whole less Draconian than those inflicted before his time.

According to the moral ideal cherished in the monastery, the true life of man was to regard oneself but dust and ashes, and, like the angels, to be ever giving God thanks. If a monk repined at such a lot, he was to castigate himself by eating only dry bread for a week and performing 500 acts of penance. The prospect of death was always to be held in view. Often did the corridors of the monastery resound with the cry, 'We shall die, we shall die!' The valley of the shadow of death was considered the road to life eternal. A monk could not call even a needle his own. Nor were the clothes he wore his personal property. They were from time to time thrown into a heap with the clothes of the other members of the House, and every monk then took from the pile the garment most convenient to his hand. Female animals were forbidden the monastery. A monk was not allowed to kiss his mother, not even at Easter, under penalty of excommunication for fifty days. Daily he attended seven services, and had often to keep vigil all night long. There was only one set meal a day; anything more in the way of food consisted of the fragments which a monk laid aside from that meal. No meat was eaten unless by special permission for reasons of health.

If a brother ate meat without permission he went without fish, eggs, and cheese for forty days. The ordinary food consisted of vegetables cooked in oil. Fish, cheese, and eggs were luxuries. Two, sometimes three, cups of wine were permitted. If a brother was so unfortunate as to break a dish, he had to stand before the assembled monks at dinner time with covered head, and hold the broken article in view of all in the refectory.55 It was forbidden to a monk to feel sad. Melancholy was a sin, and was to be overcome by prayer, one hundred and fifty genuflexions, and five hundred Kyrie Eleisons a day. The monks were required to read regularly in the monastery library.56 The task of copying manuscripts occupied a place of honour, and was under strict regulations. Fifty genuflexions were the penalty prescribed for not keeping one's copy clean; one hundred and fifty such acts of penance for omitting an accent or mark of punctuation; thirty, for losing one's temper and breaking his pen; fasting on dry bread was the fate of the copyist guilty of leaving out any part of the original, and three days' seclusion for daring to trust his memory instead of following closely the text before him.57

Ignatius of Smolensk58 found Russian monks in the monastery employed in transcribing books for circulation in Russia. Stephen of Novgorod59 met two old friends from his town busy copying the Scriptures. A good monastic scriptorium rendered an immense service; it did the work of the printing-press.

Yet, notwithstanding all restrictions, men could be happy at the Studion. One of its inmates for instance congratulates himself thus on his lot there, 'No barbarian looks upon my face; no woman hears my voice. For a thousand years no useless (ἄπρακτος) man has entered the monastery of Studius; none of the female sex has trodden its court. I dwell in a cell that is like a palace; a garden, an oliveyard, and a vineyard surround me. Before me are graceful and luxuriant cypress trees. On one hand is the city with its market-place; on the other, the mother of churches and the empire of the world.'60

Hymnology was likewise cultivated at the Studion, many hymns of the Greek Church being composed by Theodore and his brother Joseph.

Two abbots of the monastery became patriarchs: Antony (975),61 and Alexius (1025),62 the latter on the occasion when he carried the great relic of the Studion, the head of John the Baptist, to Basil II. lying at the point of death.63

At least as early as the reign of Alexius I. Comnenus, the abbot of the Studion held the first place among his fellow-abbots in the city. His precedence is distinctly recognised in a Patriarchal Act of 1381 as a right of old standing.64

The spirit of independence which characterized the monastery did not die with the abbot Theodore. The monks of the Studion were the most stubborn opponents of the famous Photius who had been elevated to the patriarchal throne directly from the ranks of the laity, and in the course of the conflict between him and the monks during the first tenure of his office for ten years, the abbots of the House were changed five times. Indeed, when Photius appointed Santabarenus as the abbot, a man accused of being a Manichaean, and who professed to be able to communicate with departed spirits, many of the monks, if not all of them, left their home. Nor was this the last assertion of the freedom of conscience for which this monastery was distinguished, and which makes it memorable in history.

Like other monasteries the Studion often served as a place of correction for offenders whom it was expedient to render harmless without recourse to the extreme rigour of the law. Santabarenus, who has just been mentioned, was sent in his wild youth, after narrowly escaping a sentence of death at the hands of the Caesar Bardas, to this monastery in the hope of being reformed in the orthodox atmosphere of the House. In the reign of Leo VI. (826-912), an official named Mousikos was sent hither to be cured of the propensity to accept bribes.65 In 912, Gregoras and Choirosphacta were obliged to join the brotherhood to repent at leisure for having favoured the attempt of Constantine Ducas, domestic of the Scholae, to usurp the throne of Constantine VII. Porphyrogenitus when seven years of age.66

To face page 44.

Several emperors sought the shelter of the Studion as a refuge from danger, or as a retreat from the vanity of the world. Thither, in 1041, Michael V. and his uncle Constantine fled from the popular fury excited by their deposition of the Empress Zoe and the slaughter of three thousand persons in the defence of the palace. The two fugitives made for the monastery by boat, and betook themselves to the church for sanctuary. But as soon as the place of their concealment became known, an angry crowd forced a way into the building to wreak vengeance upon them, and created a scene of which Psellus has left us a graphic account. Upon hearing the news of what was going on, he and an officer of the imperial guard mounted horse and galloped to the Studion. A fierce mob was madly attempting to pull down the structure, and it was with the utmost difficulty that the two friends managed to enter the church and make their way to the altar. The building seemed full of wild animals, glaring with eyes on fire at their victims, and making the air resound with the most terrible cries. Michael was on his knees clasping the holy table; Constantine stood on the right; both were dressed like monks, and their features were so transformed by terror as to be almost beyond recognition. The spectacle of greatness thus brought low was so pathetic that Psellus burst into tears and sobbed aloud. But the crowd only grew more fierce, and drew nearer and nearer to the fugitives as though to rend them in pieces. Only a superstitious dread restrained it from laying hands upon them in a shrine so sacred and venerated. The uproar lasted for hours, the mob content meanwhile with striking terror and making flight impossible. At length, late in the afternoon, the prefect of the city appeared upon the scene, accompanied by soldiers and followed by large crowds of citizens. He came with instructions to bring Michael and Constantine out of the church. In vain did he try the effect of mild words and promises of a gentle fate. The fallen emperor and his uncle clung to the altar more desperately. The prefect then gave orders that the two wretched men should be dragged forth by main force. They gripped the altar yet more tightly, and in piteous tones invoked the aid of all the eikons in the building. The scene became so heartrending that most of the spectators interfered on behalf of the victims of misfortune, and only by giving solemn assurance that they would not be put to death was the prefect allowed to proceed to their arrest. Michael and Constantine were then dragged by the feet as far as the Sigma, above S. Mary Peribleptos (Soulou Monastir), and after having their eyes burnt out were banished to different monasteries, to muse on the vanity of human greatness and repent of their misdeeds.67

The Studion appears in the final rupture of the Eastern and Western Churches.68 The immediate occasion was a letter sent by the Archbishop of Achrida, in 1053, to the Bishop of Trani, condemning the Church of Rome for the use of unleavened bread in the administration of the Holy Communion, and for allowing a fast on Saturday. Nicetas Stethetos (Pectoratus), a member of the House renowned for his asceticism, and for his courage in reproving the scandalous connection of Constantine IX. with Sklerena, wrote a pamphlet, in Latin, in which, in addition to the charges against Rome made by the Archbishop of Achrida, the enforced celibacy of the clergy was denounced. The pamphlet was widely circulated by the Patriarch Kerularios, who wished to bring the dispute between the Churches to an issue. But the emperor not being prepared to go so far, invited the Pope to send three legates to Constantinople to settle the differences which disturbed the Christian world. Cardinal Humbert, one of the legates, replied to Nicetas in the most violent language of theological controversy, and to bring matters to a conclusion an assembly, which was attended by the Emperor Constantine, his court, and the Papal legates, met at the Studion on the 24th of June 1054. A Greek translation of the pamphlet composed by Nicetas was then read, and after the discussion of the subject, Nicetas retracted his charges and condemned all opponents of the Roman Church. His pamphlet was, moreover, thrown into the fire by the emperor's orders, and on the following day he called upon the Papal legates, who were lodged at the palace of the PegÉ (Baloukli), and was received into the communion of the Church he had lately denounced. But the patriarch was not so fickle or pliant. He would not yield an iota, and on the 15th of July 1054 Cardinal Humbert laid on the altar of S. Sophia the bull of excommunication against Kerularios and all his followers, which has kept Western and Eastern Christendom divided to this day.

When Michael VII. (1067-78) saw that the tide of popular feeling had turned against him in favour of Nicephorus Botoniates, he meekly retired to this House, declining to purchase a crown with cruelty by calling upon the Varangian guards to defend his throne with their battle-axes. Michael was appointed bishop of Ephesus, but after paying one visit to his diocese he returned to Constantinople and took up his abode in the monastery of Manuel(p. 257).69

To the Studion, where he had studied in his youth and which he had embellished, the Emperor Isaac Comnenus retired, when pleurisy and the injuries he received while boar-hunting made him realize that he had but a short time to live. In fact, he survived his abdication for one year only, but during that period he proved a most exemplary monk, showing the greatest deference to his abbot, and besides performing other lowly duties acted as keeper of the monastery gate. How thoroughly he was reconciled to the exchange of a throne for a cell appears in the remark made to his wife, who had meantime taken the veil at the Myrelaion, 'Acknowledge that when I gave you the crown I made you a slave, and that when I took it away I set you free.' His widow commemorated his death annually at the Studion, and on the last occasion surprised the abbot by making a double offering, saying, 'I may not live another year,' a presentiment which proved true. According to her dying request, Aecatherina was buried in the cemetery of the Studion, 'as a simple nun, without any sign to indicate that she was born a Bulgarian princess and had been a Roman empress.'70

On the occasion of the triumphal entry of Michael Palaeologus into the city in 1261, the emperor followed the eikon of the Theotokos Hodegetria, to whom the recovery of the Empire was attributed, on foot as far as the Studion; and there, having placed the eikon in the church, he mounted horse to proceed to S. Sophia.71

One of the sons of Sultan Bajazet was buried at the Studion.72 The prince had been sent by the Sultan as a hostage to the Byzantine Court, and being very young attended school in Constantinople with John, the son of the Emperor Manuel. There he acquired a taste for Greek letters, and became a convert to the Christian faith; but for fear of the Sultan's displeasure he was long refused permission to be baptized. Only when the young man lay at the point of death, in 1417, a victim to the plague raging in the city, was the rite administered, his schoolmate and friend acting as sponsor.

A tombstone from the cemetery of the monastery is built into the Turkish wall at the north-eastern corner of the church. It bears an epitaph to the following effect:—'In the month of September of the year 1387, fell asleep the servant of God, Dionysius the Russian, on the sixth day of the month.' The patrician Bonus, who defended the city against the Avars in 627, while the Emperor Heraclius was absent dealing with the Persians, was buried at the Studion.73

On the festival of the Decapitation of S. John the Baptist, the emperor attended service at the Studion in great state. Early in the morning the members of the senate assembled therefore at the monastery, while dignitaries of an inferior rank took their place outside the gate (Narli Kapou) in the city walls below the monastery, and at the pier at the foot of the steep path that descends from that gate to the shore of the Sea of Marmora, all awaiting the arrival of the imperial barge from the Great Palace. Both sides of the path were lined by monks of the House, holding lighted tapers, and as soon as the emperor disembarked, the officials at the pier and the crowd of monks, with the abbot at their head, swinging his silver censer of fragrant smoke, led the way up to the gate. There a halt was made for the magistri, patricians, and omphikialioi (ὀμφικιάλιοι) to do homage to the sovereign and join the procession, and then the long train wended its way through the open grounds attached to the monastery (διὰ τοῦ ἐξαέρου), and through covered passages (διὰ τῶν ἐκεῖσε διαβατικῶν),74 until it reached the south-eastern end of the narthex (εἰσέρχονται διὰ τοῦ πρὸς ἀνατολικὴν δεξιοῦ μέρους τοῦ νάρθηκος). Before the entrance at that point, the emperor put on richly embroidered robes, lighted tapers, and then followed the clergy into the church, to take his stand at the east end of the south aisle. The most important act he performed during the service was to incense the head of John the Baptist enshrined on the right hand of the bema. At the conclusion of the Office of the day, he was served by the monks with refreshments under the shade of the trees in the monastery grounds (ἀναδενδράδιον); and, after a short rest, proceeded to his barge with the same ceremonial as attended his arrival, and returned to the palace. 75

To face page 48.

The church was converted into a mosque in the reign of Bajazet II. (1481-1512) by the Sultan's equerry, after whom it is now named.

The church of S. John the Baptist of the Studion is a basilica, and is of special interest because the only surviving example of that type in Constantinople, built while the basilica was the dominant form of ecclesiastical architecture in the Christian world. It has suffered severely since the Turkish conquest, especially from the fire which, in 1782, devastated the quarter in which it stands, and from the fall of its roof, a few winters ago, under an unusual weight of snow. Still, what of it remains and the descriptions of its earlier state given by Gyllius, Gerlach, and other visitors, enable us to form a fair idea of its original appearance. The recent explorations conducted by the Russian Institute at Constantinople have also added much to our knowledge of the building.

It is the oldest church fabric in the city, and within its precincts we stand amid the surroundings of early Christian congregations. For, partly in original forms, partly in imitations, we still find here a basilica's characteristic features: the atrium, or quadrangular court before the church; on three of its sides surrounded by cloisters; in its centre, the marble phialÉ or fountain, for the purification of the gathering worshippers; the narthex, a pillared porch along the western faÇade, where catechumens and penitents, unworthy to enter the sanctuary itself, stood afar off; the interior area divided into nave and aisles by lines of columns; the semicircular apse at the eastern extremity of the nave for altar and clergy; and galleries on the other sides of the building to provide ample accommodation for large assemblies of faithful people.

The northern wall of the atrium is original, as the crosses in brick formed in its brickwork show. The trees which shade the court, the Turkish tombstones beneath them, and the fountain in the centre, combine to form a very beautiful approach to the church, and reproduce the general features and atmosphere of its earlier days.

The narthex is divided into three bays, separated by heavy arches. It is covered by a modern wooden roof, but shows no signs of ever having been vaulted. The centre bay contains in its external wall a beautiful colonnade of four marble columns, disposed, to use a classical term, 'in antis.' They stand on comparatively poor bases, but their Corinthian capitals are exceptionally fine, showing the richest Byzantine form of that type of capital. The little birds under the angles of the abaci should not be overlooked.

The entablature above the columns, with its architrave, frieze, and cornice, follows the classic form very closely, and is enriched in every member. Particularly interesting are the birds, the crosses, and other figures in the spaces between the modillions and the heavy scroll of the frieze. The drill has been very freely used throughout, and gives a pleasant sparkle to the work.

In the second and fourth intercolumniations there are doorways with moulded jambs, lintels, and cornices, but only the upper parts of these doorways are now left open to serve as windows.

The cornice of the entablature returns westwards at its northern and southern ends, indicating that a colonnade, with a smaller cornice, ran along the northern and southern sides of the atrium, if not also along its western side. The cloisters behind the colonnades, were connected at their west end with the narthex by two large and elaborately moulded doorways still in position.

Five doors lead from the narthex into the church; three opening into the nave, the others into the aisles.

The interior of the church, now almost a total ruin, was divided into nave and two aisles by colonnades of seven columns of verd antique marble. But only six of the original columns have survived the injuries which the building has sustained; the other columns are Turkish, and are constructed of wood with painted plaster covering.

The colonnades supported an entablature of late Corinthian type, which, as the fall of the Turkish plaster that once covered it has revealed, had the same moulding as the entablature in the narthex. The architrave was in three faces, with a small bead ornament to the upper two, and finished above with a small projecting moulding. The frieze was an ogee, bellied in the lower part. Of the cornice only the bed mould, carved with a leaf and tongue, remains.

Above each colonnade stood another range of seven76 columns connected, probably, by arches. Along the northern, southern, and western sides of the church were galleries constructed of wood. Those to the north and south still exist in a ruined condition, and many of the stone corbels which supported the beams remain in the walls. Only scanty vestiges of the gallery above the narthex can be now distinguished. Its western wall, the original outer wall of the upper part of the church, has totally disappeared. Its eastern arcade has been replaced by the Turkish wall which constitutes the present outer wall of that part of the church. But beyond either end of that wall are visible, though built up, the old openings by which the gallery communicated with its companion galleries; while to the west of the wall project the ragged ends of the Byzantine walls which formed the gallery's northern and southern sides. The nave rose probably to a greater height than it does now, and had a roof at a higher level than the roofing of the aisles. It doubtless resembled the basilican churches at Salonica, either with clearstory windows, as in S. Demetrius, or without such windows, as in Eski Juma Jamissi.

To face page 48.

The nave terminates in a large apse, semicircular within and showing three sides on the exterior. Only the lower part is original; the Turkish superstructure is lower and on a smaller scale than the Byzantine portion it has replaced. There are no side chapels. Under the bema the Russian explorers discovered a small cruciform crypt. The large quantity of mosaic cubes found in the church during the recent Russian excavations proves that the church was decorated with mosaics, while the remains of iron plugs in the western wall for holding marble slabs show that the building had the customary marble revetment. But what is curious is to find the mortar pressed over the face of the stones, and broad decorative joints formed by ruled incised lines and colour. Mr. W. S. George suggests that this was a temporary decoration executed pending some delay in the covering of the walls with marble. He also thinks that the importance given to the joint in late Byzantine work and in Turkish work may be a development from such early treatment of mortar.

The floor of the church was paved with pieces of marble arranged in beautiful patterns, in which figures of animals and scenes from classic mythology were inlaid. Gerlach77 noticed the beauty of the pavement, and Salzenberg78 represents a portion of it in his work on S. Sophia. But the members of the Russian Institute of Constantinople have had the good fortune to bring the whole pavement to light.

A noticeable feature is the number of doors to the church, as in S. Irene. Besides the five doors already mentioned, leading into the interior from the narthex, there is a door at the eastern end of each aisle, and close to each of these doors is found both in the southern and northern walls of the building an additional door surmounted by a window. The latter doors and their windows have been walled up.

The exterior is in two stories, corresponding to the ground floor and the galleries. It has two ranges of eight large semicircular-headed windows in the northern and southern walls, some of them modified, others built up, since the building became a mosque. The five windows in the gable of the western wall are, like the wall itself, Turkish. Pilasters are placed at the angles and at the apse.

On the south side of the church is a cistern, the roof of which rests on twenty-three columns crowned by beautiful Corinthian capitals.

53 The modern Touzla at the northern head of the gulf of Nicomedia. See the articles by Mr. Siderides and Mr. Meliopoulos in the Proceedings of the Greek Syllogos of Constantinople, vol. xxxi., 1907-8.

CHAPTER III

THE CHURCH OF SS. SERGIUS AND BACCHUS, KUTCHUK AYA SOFIA

On the level tract beside the Sea of Marmora, to the south of the Hippodrome, and a few paces to the north-west of Tchatlady Kapou, stands the ancient church of SS. Sergius and Bacchus. It is commonly known as the mosque Kutchuk Aya Sofia, Little S. Sophia, to denote at once its likeness and its unlikeness to the great church of that name. It can be reached by either of the two streets descending from the Hippodrome to the sea, or by taking train to Koum Kapou, and then walking eastwards for a short distance along the railroad.

There can be no doubt in regard to its identity. For the inscription on the entablature of the lower colonnade in the church proclaims the building to be a sanctuary erected by the Emperor Justinian and his Empress Theodora to the honour of the martyr Sergius. The building stands, moreover, as SS. Sergius and Bacchus stood, close to the site of the palace and the harbour of Hormisdas.81 When Gyllius visited the city the Greek community still spoke of the building as the church of SS. Sergius and Bacchus—'Templum Sergii et Bacchi adhuc superest, cujus nomen duntaxat Graeci etiam nunc retinent.'82

To face page 63.

The foundations of the church were laid in 527, the year of Justinian's accession,83 and its erection must have been completed before 536, since it is mentioned in the proceedings of the Synod held at Constantinople in that year.84 According to the Anonymus, indeed, the church and the neighbouring church of SS. Peter and Paul were founded after the massacre in the Hippodrome which suppressed the Nika Riot. But the Anonymus is not a reliable historian.85

The church did not stand alone. Beside it and united with it, Justinian built also a church dedicated to the Apostles Peter and Paul,86 so that the two buildings formed a double sanctuary, having a common court and a continuous narthex. They were equal in size and in the richness of the materials employed in their construction, and together formed one of the chief ornaments of the palace and the city. There was, however, one striking difference between them; SS. Sergius and Bacchus was a domical church, while SS. Peter and Paul was a basilica. Styles of ecclesiastical architecture destined soon to blend together in the grandeur and beauty of S. Sophia were here seen converging towards the point of their union, like two streams about to mingle their waters in a common tide. A similar combination of these styles occurs at Kalat-SemÂn in the church of S. Symeon Stylites, erected towards the end of the fifth century, where four basilicas forming the arms of a cross are built on four sides of an octagonal court.87

The saints to whom the church was dedicated were brother officers in the Roman army, who suffered death in the reign of Maximianus,88 and Justinian's particular veneration for them was due, it is said, to their interposition in his behalf at a critical moment in his career. Having been implicated, along with his uncle, afterwards Justin I., in a plot against the Emperor Anastasius, he lay under sentence of death for high treason; but on the eve of his execution, a formidable figure, as some authorities maintain,89 or as others affirm, the saints Sergius and Bacchus, appeared to the sovereign in a vision and commanded him to spare the conspirators. Thus Justinian lived to reach the throne, and when the full significance of his preservation from death became clear in the lustre of the imperial diadem, he made his deliverers the object of his devout regard. Indeed, in his devotion to them he erected other sanctuaries to their honour also in other places of the Empire.90 Still this church, founded early in his reign, situated beside his residence while heir-apparent, and at the gates of the Great Palace, and withal a gem of art, must be considered as Justinian's special thankoffering for his crown.

With the church of SS. Sergius and Bacchus was associated a large monastery known, after the locality in which it stood, as the monastery of Hormisdas, ἐν τοῖς Ὁρμίσδου. It was richly endowed by Justinian.91

A remarkable scene was witnessed in the church in the course of the controversy which raged around the writings known in ecclesiastical history as 'The Three Chapters,' the work of three theologians tainted, it was alleged, with the heretical opinions of Nestorius. Justinian associated himself with the party which condemned those writings, and prevailed upon the majority of the bishops in the East to subscribe the imperial decree to that effect. But Vigilius, the Pope of the day, and the bishops in the West, dissented from that judgment, because the authors of the writings in question had been acquitted from the charge of heresy by the Council of Chalcedon. To condemn them after that acquittal was to censure the Council and reflect upon its authority. Under these circumstances Justinian summoned Vigilius to Constantinople in the hope of winning him over by the blandishments or the terrors of the court of New Rome. Vigilius reached the city on the 25th of January 547, and was detained in the East for seven years in connection with the settlement of the dispute. He found to his cost that to decide an intricate theological question, and above all to assert 'the authority of S. Peter vested in him' against an imperious sovereign and the jealousy of Eastern Christendom, was no slight undertaking. Pope and Emperor soon came into violent collision, and fearing the consequences Vigilius sought sanctuary in the church of S. Peter100 as he styles it, but which Byzantine writers101 who record the scene name S. Sergius.

Justinian was not the man to stand the affront. He ordered the praetor of the city to arrest the Pope and conduct him to prison. But when that officer appeared, Vigilius grasped the pillars of the altar and refused to surrender. Thereupon the praetor ordered his men to drag the Pope out by main force. Seizing Vigilius by his feet, holding him by his beard and the hair of his head, the men pulled with all their might, but they had to deal with a powerful man, and he clung fast to the altar with an iron grip. In this tug-of-war the altar at length came crashing to the ground, the Pope's strong hands still holding it tight. At this point, however, the indignation and sympathy of the spectators could not be restrained; the assailants of the prostrate prelate were put to flight, and he was left master of the situation. Next day a deputation, including Belisarius and Justin, the heir-apparent, waited upon Vigilius, and in the emperor's name assured him that resistance to the imperial will was useless, while compliance with it would save him from further ill-treatment. Yielding to the counsels of prudence, the Pope returned to the palace of Placidia,102 the residence assigned to him during his stay in the capital.

Probably at this time arose the custom of placing the churches of SS. Peter and Paul, and SS. Sergius and Bacchus at the service of the Latin clergy in Constantinople, especially when a representative of the Pope, or the Pope himself, visited the city. The fact that the church was dedicated to apostles closely associated with Rome and held in highest honour there, would make it a sanctuary peculiarly acceptable to clergy from Western Europe. This, however, did not confer upon Roman priests an exclusive right to the use of the building, and the custom of allowing them to officiate there was often more conspicuous in the breach than in the observance. Still the Roman See always claimed the use of the church, for in the letter addressed in 880 by Pope Julius VIII. to Basil I., that emperor is thanked for permitting Roman clergy to officiate again in SS. Sergius and Bacchus according to ancient custom: 'monasterium Sancti Sergii intra vestram regiam urbem constitutum, quod sancta Romana Ecclesia jure proprio quondam retinuit, divina inspiratione repleti pro honore Principis Apostolorum nostro praesulatui reddidistis.'103

The most distinguished hegoumenos of the monastery was John Hylilas, better known, on account of his learning, as the Grammarian, and nicknamed Lecanomantis, the Basin-Diviner, because versed in the art of divination by means of a basin of polished brass. He belonged to a noble family of Armenian extraction, and became prominent during the reigns of Leo V., Michael II., and Theophilus as a determined iconoclast. His enemies styled him Jannes, after one of the magicians who withstood Moses, to denote his character as a sorcerer and an opponent of the truth. Having occasion, when conducting service in the imperial chapel to read the lesson in which the prophet Isaiah taunts idolaters with the question, 'To whom then will ye liken God, or to what likeness will ye compare him?' John, it is said, turned to Leo V., and whispered the significant comment, 'Hearest thou, my lord, the words of the prophet? They give thee counsel.' He was a member of the Commission charged by that emperor to collect passages from the Holy Scriptures and the Fathers of the Church that condemned the use of images in worship. Prominent iconodules were interned in the monastery of Hormisdas in the hope that he would turn them from the error of their ways by his arguments and influence. He directed the education of Theophilus and supported the iconoclastic policy pursued by that pupil when upon the throne. Theophilus appointed his tutor syncellus to the Patriarch Antony, employed him in diplomatic missions,104 and finally, upon the death of Antony, created him patriarch. The name of John can still be deciphered under somewhat curious circumstances, in the litany which is inscribed on the bronze doors of the Beautiful Gate at the south end of the inner narthex of S. Sophia. When those doors were set up in 838, Theophilus and his empress had no son, and accordingly, in the threefold prayer inscribed upon the doors, the name of John was associated with the names of the sovereigns as a mark of gratitude and esteem. But in the course of time a little prince, to be known in history as Michael III., was born and proclaimed the colleague of his parents. It then became necessary to insert the name of the imperial infant in the litany graven on the Beautiful Gate of the Great Church, and to indicate the date of his accession. To add another name to the list of names already there was, however, impossible for lack of room; nor, even had there been room, could the name of an emperor follow that of a subject, though that subject was a patriarch. The only way out of the difficulty, therefore, was to erase John's name, and to substitute the name of the little prince with the date of his coming to the throne; the lesser light must pale before the greater. This was done, but the bronze proved too stubborn to yield completely to the wishes of courtiers, and underneath Michael's name has kept fast hold of the name John to this day. The original date on the gate also remains in spite of the attempt to obliterate it.

SS. Sergius and Bacchus was one of the sanctuaries of the city to which the emperor paid an annual visit in state.105 Upon his arrival at the church he proceeded to the gallery and lighted tapers at an oratory which stood in the western part of the gallery, immediately above the Royal Gates, or principal entrance to the church. He went next to the chapel dedicated to the Theotokos, also in the gallery, and after attending to his private devotions there, took his place in the parakypticon (ἐν τῷ παρακυπτικῷ τοῦ θυσιαστηρίου), at the north-eastern or south-eastern end of the gallery, whence he could overlook the bema and follow the public service at the altar.106 In due course the Communion elements were brought and administered to him in the chapel of the Theotokos; he then retired to the metatorion (a portion of the gallery screened off with curtains), while the members of his suite also partook of the Communion in that chapel. At the close of the service he and his guests partook of some light refreshments, biscuits and wine, in a part of the gallery fitted up for that purpose, and thereafter returned to the palace.

In the description of the architectural features of the church and for the plans and most of the illustrations in this chapter I am under deep obligation to Mr. A. E. Henderson, F.S.A. The information gained from him in my frequent visits to the church in his company, and from his masterly article on the church which appeared in the Builder of January 1906, has been invaluable.

In design the church is an octagonal building roofed with a dome and enclosed by a rectangle, with a narthex along the west side. This was a favourite type of ecclesiastical architecture, and is seen also in another church of the same period, San Vitale of Ravenna, in which Justinian and Theodora were interested. There, however, the octagonal interior is placed within an octagonal enclosure. The adoption of a rectangular exterior in the Constantinopolitan sanctuary is a characteristic Byzantine feature.107 S. Vitale was founded in 526, a year before SS. Sergius and Bacchus.

To face page 70.

As an examination of the plan will show, the architect's design has not been followed with strict accuracy, and the result is that both the enclosing square and the interior octagon are very irregular figures. Furthermore, the two portions of the building have not the same orientation, so that the octagon stands askew within its rectangular frame. How this lack of symmetry should be explained, whether due to sloven work or the result of the effort to adapt the church to the lines of the earlier church of SS. Peter and Paul, with which it was united, is difficult to decide.

The court which stands before the Turkish portico in front of the west side of the building represents the old atrium of the church, and to the rear of the portico is still found the ancient narthex. At the south end of the narthex is a stone staircase leading to the gallery. The arch at the foot of the staircase is built of fragments from the old ciborium or eikonostasis of the church. The great height (0.24 metre or 9 inches) of the steps is found, according to Mr. Antoniadi, also in S. Sophia.

The exterior walls, which are mostly in brick and rubble masonry, exhibit poor workmanship, and have undergone considerable repair, especially on the east. On the south there are two thicknesses of walling. The outer thickness has arched recesses at intervals along its length, corresponding to openings in the inner thickness, and thus while buttressing the latter also enlarges slightly the area of the church. The length of the rectangular enclosure from west to east is 101 feet, with an average breadth of 77¹⁄₂ feet from north to south, excluding the recesses in the latter direction.

All the windows of the church have been altered by Turkish hands, and are rectangular instead of showing semicircular heads.

The passage intervening between the rectangular enclosure and the octagon is divided into two stories, thus providing the church with an ambulatory below and a gallery above.

The domed octagon which forms the core of the building stands at a distance of some 18¹⁄₂ feet from the rectangle within which it is placed. It measures 53¹⁄₂ feet by 50¹⁄₂ feet. The eight piers at its angles rise to a height of 33¹⁄₂ feet from the floor to the springing of the dome arches. The archways thus formed, except the bema arch, are filled in with two pairs of columns in two stories set on the outer plane of the piers. The lower colonnade is surmounted, after the classic fashion, by a horizontal entablature profusely carved while the upper columns are bound by arches, thus making seven sides of the octagon a beautiful open screen of fourteen columns and as many triple arcades, resplendent with marbles of various hues and rich with carved work. The mass of the piers is relieved by their polygonal form, a fluted cymatium along their summit, and a repeating design of a flower between two broad leaves below the entablature. Though the flower points upwards it has been mistaken for a cluster of grapes.108 At the four diagonal points the sides of the octagon are semicircular, forming exhedrae, an arrangement which gives variety to the lines of the figure, widens the central area, secures more frontage for the gallery, and helps to buttress the dome. The same feature appears in S. Sophia, whereas in San Vitale all the sides of the octagon, excepting the eastern side, are semicircular. The extension of the interior area of a building (square or octagonal) by means of niches at the angles or in the sides, or both at the angles and in the sides, was a common practice.109

There is considerable difference in the size of the piers and the dome arches. The eastern piers stand farther apart than their companions, and consequently the arch over them, the triumphal arch of the sanctuary, is wider and loftier than the other arches. The bays to the north-east and the south-east are also wider than the bays at the opposite angles. The apse is semicircular within, and shows three sides on the exterior. As in S. Sophia and S. Irene, there is no prothesis or diaconicon.

The pairs of columns, both below and above, are alternately verd antique and red Synnada marble, resting on bases of the blue-veined white marble from the island of Marmora. The capitals on the lower order are of the beautiful type known as the 'melon capital,' a form found also in San Vitale at Ravenna and in the porch of S. Theodore in Constantinople (p. 246). The neckings are worked with the capitals, and enriched by 'egg-and-dart' pointing upwards. In the centre of the capitals was carved the monogram of Justinian or that of Theodora. Most of the monograms have been effaced, but the name of the empress still appears on the capital of the western column in the south bay, while that of Justinian is found on the first capital in the south-western bay; on both capitals in the north-western bay, accompanied by the title Basileus; and, partially, on the last capital in the north-eastern bay.

In the soffit of the architrave are sunk panels of various patterns, the six-armed cross occurring twice. The beadings of the fasciae are enriched with the designs commonly known as 'rope,' 'bead-and-reel,' 'egg-and-dart,' and again 'bead-and-reel.'

The frieze is in two heights. The lower portion is a semicircular pulvinar adorned with acanthus leaves, deeply undercut; the upper portion is occupied by a long inscription in raised ornamental letters to the honour of Justinian, Theodora, and S. Sergius. The cornice is decorated with dentils, 'bead-and-reel,' projecting consols, 'egg-and-dart,' and leaves of acanthus.

The inscription (Fig. 20) may be rendered thus: Other sovereigns, indeed, have honoured dead men whose labour was useless. But our sceptred Justinian, fostering piety, honours with a splendid abode the servant of Christ, Creator of all things, Sergius; whom nor the burning breath of fire, nor the sword, nor other constraints of trials disturbed; but who endured for the sake of God Christ to be slain, gaining by his blood heaven as his home. May he in all things guard the rule of the ever-vigilant sovereign, and increase the power of the God-crowned Theodora whose mind is bright with piety, whose toil ever is unsparing efforts to nourish the destitute.

The inscription is not mere flattery to the founders of the church. Justinian and Theodora were devout after the fashion of their day, and took a deep interest in the poor. The empress erected an asylum for fallen women, hostels for strangers, hospitals for the sick, and homes for the destitute. 'On the splendid piece of tapestry embroidered in gold which formed the altar cloth of S. Sophia, she was represented with Justinian as visiting hospitals and churches.'110

To the rear of the southern straight side of the octagon two columns stand under the gallery, with wide fillets worked on both sides of their bases, shafts, and capitals, showing that a frame of stone or wood was once affixed to them. The capitals are of the ordinary cushion type and bear on opposite faces the monograms Justinian, Basileus.

To face page 74.

Two feet above the cornice, or twenty feet from the floor of the church, the level of the gallery is reached.111 Here the columns are smaller than those below, and are bound together by arches instead of by an architrave. Their capitals represent the type known as the 'Pseudo-Ionic' or cushion capital, in view of its broad head. It appears appropriately here as the form of capital required to carry the impost of an arch upon a capital. At one time, indeed, that demand was met by placing upon the capital a distinct block of stone, a fragment, so to speak, of the horizontal architrave. It is the device adopted in San Vitale at Ravenna, S. Demetrius of Salonica, and elsewhere, but never it would seem in Constantinople, except in the underground cisterns of the city. It was, however, too inartistic to endure, and eventually was superseded by capitals with a broad flattened head on which the wide impost of an arch could rest securely.112

A free form of acanthus, deeply undercut on the face towards the central area of the church, covers the capitals, and in the centre of that face, on all the capitals except the eighth (counting from the north-east) is carved the monogram of the title Basileus, or of Justinian, or of Theodora.

In the south side of the gallery stand two columns corresponding to the two columns in the aisle below. They are poor in design and not original. The western capital is 'Pseudo-Ionic,'113 with a plain cross on the northern face. The eastern capital is in the basket form with roundels on the four faces. Two additional columns are found in the western portion of the gallery. They are of verd antique and larger than the other columns in this story of the church, and have sunk crosses in them. The splendour of the interior decoration has certainly been dimmed, for the walls of the edifice once gleamed with marbles and glittered with mosaics. 'By the sheen of its marbles,' says Procopius,114 'it was more resplendent than the sun, and everywhere it was filled profusely with gold.' When Ferguson examined the building, remains of frescoes or of mosaics, which have disappeared since his time, could be distinguished in the narthex. The soffit, both of the upper and of the lower cymatium on the piers, projects sufficiently to admit the application of the customary marble incrustation. The proportions of the building are marred by the boarded floor which rises seventeen centimeters above the original pavement, disguising the real elevation of the dome and of the columns in the lower colonnade. But notwithstanding all changes for the worse the building is still a beautiful structure. Very effective especially is the happy combination of the various lines and forms here brought together—the rectilinear and the semicircular sides of the octagon, the octagonal fabric and the round dome that crowns it, the horizontal entablature stretched along the summit of the lower story of columns and the arches that leap from column to column in the gallery. This harmonious variety of form has also a historical significance. An old order in architecture and a new order here meet and embrace before the earlier, having served its age, passes away and the later comes triumphant to fill another era of the world with fresh beauties. Here in the tide of time we look before and after.

To the student of architecture the dome of this church is specially interesting. In the application of the dome to the octagon no pendentives are employed. The octagon is carried up to the base of the dome, which is built in sixteen longitudinal compartments that impinge upon one another and form groins giving to the dome its strength and sweep. On the groins is a plaster moulding, probably Byzantine. The eight compartments directly above the dome arches are flat, and flush with the inner face of the octagon, and in each of them is a semicircular-headed window. They rise perpendicular to a point a little above the windows, and then curve with a radius to the centre of the dome.

Fig. 21.—View of the Exterior of the Dome of SS. Sergius and Bacchus..

On the other hand the eight compartments directly above the angles of the octagon are narrower than the preceding compartments; they have no windows, and, what is of special importance to note, they are deeply concave.115 Such marked hollowness is found in later domes as a decorative feature, but here it is primarily and supremely a constructive device. By its means the concave compartments are set slightly back from the octagon's inner face, leaving, at the springing line, portions of the wall-head to appear as little flat ledges on each side of the angles. This is a most skilful expedient, and compares favourably with the methods employed elsewhere to apply the dome to the octagon.116 In the octagonal church of S. Lorenzo at Milan the octagon is turned into the circle by the introduction of squinches. In San Vitale a considerable walling is built between the line of the octagon and the springing line of the dome, while the bed for the dome is formed by introducing, in the space over the angles of the octagon, niches which are worked above to the circle on plan. On the other hand, it is interesting to compare with these methods the method employed in the baptistery of S. Sophia, now a Sultan's TurbÉ, near the southern entrance to the inner narthex. Although the walls of the building describe a square on the exterior, they form an octagon on the interior with semicircular bays at the diagonals, as in SS. Sergius and Bacchus. But in the application of the dome the true pendentive is used. The baptistery was erected shortly before S. Sophia, and in view of the erection of the great church.

The curvature of the dome of SS. Sergius and Bacchus has three zones, which have respectively a radius of m. 8, (drawn from the centre of the octagon), m. 3¹⁄₄, and m. 9¹⁄₂, (centre about m. 2, below the springing of the dome). The first extends to a point a little above the heads of the dome windows; the second about m. 2 higher; the third to the crown of the dome. The groins stop short a little below the dome's apex, where they are arched into one another, leaving a saucer-shaped crown now capped by a Turkish finial. The dome is covered with lead, and presents an undulating surface owing to the protuberance of its eight concave compartments.117

The system of weighting and buttressing the dome displays great skill, and will be best understood by studying Mr. Henderson's geometrical and constructive sections of the systems (Figs. 28, 29).