fyxmll Hmtraitg pbmg BOUGHT WITH THE* INCOME - - FROM THE SAGE ENDOWMENT FUND THE GIFT OF Heurg IB. Sage 1891 A ., 1.5-.3Q ^ .Vl|l^\\aQ.V. Cornell University Library N7832.L92 M8 Monuments of the early church / olin 3 1924 032 578 399 The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924032578399 f^anoboafea of Stcljtealiigg anb antiquities MONUMENTS OP THE EARLY CHURCH A HANDBOOK OF CHRISTIAN ARCHAEOLOGY HANDBOOKS OF ARCHAEOLOGY AND ANTIQUITIES ROMAN FESTIVALS. — Fowler - $1.25 net GREEK SCULPTURE. -Gardner 2.50 " GREEK CONSTITUTIONAL HISTORY. — Greenidge 1.25 " GREEK AND ROMAN COINS. — Hill 2.25 " THE DESTRUCTION OF ANCIENT ROME. — Lanciani 1.75 " MONUMENTS OF THE EARLY CHURCH. — LOWRIE 1.75 " ROMAN PUBLIC LIFE. -Greenidoe 2.50 " Others to follow MONUMENTS OF THE EARLY CHURCH BY WALTER LOWRIE, M.A. LATE FELLOW OF THE AMERICAN SCHOOL OF CLASSICAL STUDIES AT ROME Nefo fforft THE MACMILLAN COMPANY LONDON : MACMILLAN & CO., Ltd. 1901 K All rights reserved Copyright, 1901, By THE MACMILLAN COMPANY. NortoooS fficm J. S. CuBhing & Co. — Berwick St Smith Norwood Mass U.S.A. PREFACE This volume is-designed to give a general view of the monu- ments of the early Church, comprising in its scope all branches of Christian art and archaeology, and treating each of them as completely as is possible within the limits of a handbook. It is designed also, in behalf of students who may wish to pursue the subject further, to provide a general introduction to the archaeology of the early Christian period. To this end it is hoped that the Bibliography will be found useful. This book deals expressly with the monumental remains of Christian antiquity, and only incidentally with the early literature which illustrates the same period and the same subjects. It gives a detailed exposition of many of the most representative monuments, especially of those which are reproduced in the accompanying illustrations. There is included only so much of a more general nature as is neces- sary to define the point of view from which the monuments of early Christian art must be regarded, or to summarize the historical inferences which may be drawn from them. The author has confined himself scrupulously to the well- defined and soundly assured results of this study, and has endeavored to eliminate questions of a controversial character. The period covered by this study extends from the second to the sixth century inclusive. The interest of this period to the secular student lies in the fact that it represents the last phase of Grseco-Roman art and civilization, and reveals at the same time a new artistic impulse which, after remaining dormant for centuries, was destined to germinate in another soil and appear again in the more familiar art of the Middle Ages. The religious interest of the period is of course still vi PEE FACE greater, whether to those who trace back to it their religious forms and customs through the medium of the Middle Ages, or to those who refer for their inspiration more directly to the early Church. It is obvious that so compendious a work as this must rely upon the conclusions of scholars who have devoted themselves specially to one or another of the departments which are herein treated. To speak only of writers still living, the author is glad to express his obligation to Mgr. Joseph Wil- pert for all that concerns early Christian painting, as also in the matter of civil and ecclesiastical dress; to Professor Orazio Marucchi for the latest study of the Roman catacombs ; to Professor G. Dehio for the study of the Christian basilicas ; to M. Auguste Choisy for the analysis of Byzantine architec- ture ; to Dr. Heinrich Holtzinger for Christian architecture in general; to Dr. R. Forrer for the study of textile art; and to all who have kindly consented to the reproduction of their illustrations. He is under obligation also to Mr. Charles R. Morey, Fellow of the American School of Classical Studies in Rome, for kind assistance in reading the proofs. Princeton, N.J., April, 1901. TABLE OF CONTENTS I. INTRODUCTION PAGE Relation of Christian to Classic Art 1 Formal Decline of Art in the Early Christian Period ... 4 Attitude of the Church toward Art 5 Limits of the Early Christian Period 6 Classification of Monuments 7 Distribution of Monuments 8 Relation to Literary Sources 11 History of the Study 17 II. CHRISTIAN CEMETERIES General Description of the Catacombs 23 Nomenclature 23 Plan 29 The Catacombs and the Arenaria 31 Exaggerated Idea of their Extent 32 Actual Location and Extent 33 The Fossors 36 The Catacombs and the Titles 37 The Christian Mode of Burial 40 Worship in the Catacombs 43 Constructions above Ground 47 Public Character of the Catacombs 47 The Area and its Buildings 48 Surface Burial 60 The Agape 50 Legal Tenure of the Cemeteries 53 The Sepulchre in Roman Law 53 Private Tenure 54 Corporate Tenure — the Burial Societies 56 Inscriptions 62 Epitaphs 63 The Poems of Damasus 74 History - of the Catacombs after the Fourth Century . . 77 vii TABLE OF CONTENTS III. CHRISTIAN ARCHITECTURE Introduction PAGE 83 A. THE BASILICA Meaning of the Name . 89 Various Views of its Origin 91 A Deliberate Invention . 91 Derived from the Civil Basilica . 92 Earlier than Constantine . . 93 Derived from the Schola . 94 Derived from the Private House 94 From the Palace Architecture . 97 From the Common House . 97 Origin in the Atrium of the Late Roman House 98 Peristyle and Atrium . . . . 98 Analogies with the Basilica . 100 Fundamental Characteristics op the Basilica . . 101 Classical Elements . 102 Internal Perspective 104 Description of the Basilica . 105 Plan . 105 Vestibule .... . 106 Nave . ... . 107 Columns and Capitals . 108 Architrave and Archivolt . 109 Pillars . Ill Galleries .... . 112 Windows . 113 Doors .... . 115 Roof . 116 Ceiling .... . 117 Decoration . 119 Presbyterium . 121 Transept .... . 124 Prothesis and Diaconicon 126 Exterior .... . 128 Facade .... . 129 Syrian Architecture . . 130 TABLE OF CONTENTS B. THE CENTRAL TYPE In General PAGE 131 Round or Polygonal Buildings Simple Plan 135 135 Composite Plan 139 The Dome from a Square Base Cross-shaped Plan .... Oblong Plan ... S. Sophia .... . . 147 147 .153 153 C. FURNITURE OF THE CHURCH Altar and Confessio . . . . 159 Ciboeium . 167 Chancels . 168 Iconostasis 170 Cathedra ........ ... 172 Ameons ... 174 Baptistery Furniture . . . . . . 175 D. POSITION AND SURROUNDINGS Orientation Atrium and Peribolos Cantharus . Propylaion . Adjoining Buildings 176 178 179 180 182 IV. PICTORIAL ART Introduction . A. EARLY CHRISTIAN PAINTING 185 (Being the frescos of the catacombs and early Christian symbol- ism in general.) Introduction 187 (Technic, divisions of the subject, character and employment of early Christian fresco. ) Early Christian Symbolism ........ 193 x TABLE OF CONTENTS PAGE Symbols of Deliverance 198 The Orans .201 Old Testament Characters . 204 Adam and Eve 204 Noah 205 Moses 206 Jonah 207 Abraham and Isaac 208 . Daniel 209 Susanna .... 210 The Three Children and the Three Magi 211 The Miracles of Christ 213 Lazarus .... 213 The Blind Man 213 The Woman with the Issue 213 The Paralytic .... ... .214 The Multiplication of the Loaves .... . 214 The Good Shepherd ... . . . . 214 The Celestial Banquet 221 The Eucharistic Banquet 223 The Fish 232 The Cross and the Monogram ... ... 236 The Virgin Mart .... 245 B. SCULPTURE Introduction 247 Character of the Monuments .... . . 247 Sculpture in East and West . 248 Themes New and Old 249 Portrait Types of Peter and Paul 251 Sarcophagi . 252 In General 252 The Jonah Sarcophagus in the Lateran . . 256 The Sarcophagus from S. Paul's 257 Other Sarcophagi in the Lateran ...... 263 Sarcophagus of Junius Bassus 264 The Similar Sarcophagus in the Lateran .... 265 The Sarcophagi of Ravenna 268 The Doors of S. Sabina at Rome 270 The Crucifixion 276 TABLE OF CONTENTS xi PAGE Ivout Carving 279 In General 279 The Box at Brescia ... 281 A Diptych at Florence 285 Cathedra of Maxiinianus 285 Gospel Cover from Murano 288 Statues 289 C. MOSAICS Introduction 292 The Study limited to Church Decoration 292 Technic of Floor, Wall, and Ceiling Mosaics .... 29.3 The Co.nstantinian Mosaics . 297 S. Costanza 297 The Churches of Rome, S. Peter's 300 The Vine Pattern 301 Themes of Apsidal Decoration 302 The Apsidal Mosaic of S. Pudenziana 304 Apsidal Mosaics of the Ftfth and Sixth Centuries . . 311 S. Paul's, SS. Cosma and Damiano, S. Lorenzo, S. Vitale. Historical Subjects 319 S. Maria Maggiore — Arch and Nave, S. Apollinare Nuovo. Mosaics of the Fifth Century at Ravenna .... 329 D. MINIATURES Introduction . . . 333 The Paris Psalter . 335 The Joshua Roll . 335 The Vienna Genesis . 337 Codex Rossanensis . 338 Cosmas Indikopleustes . . ... . 339 The Cotton Bible and the Cambridge Evangeliar . 339 V. THE MINOR ARTS Eucharistic Vessels Lamps ......... Censers Divers Objects Medals, Amulets, Ampullae, Lead and Silver Objects, 343 347 353 354 xii TABLE OF CONTENTS PAGE Glass ... 357 Engraved Glass, Cut Glass, vasa diatreta, Gold Glass. The Textile Art 362 Lack of a Thorough Study of Classical Textiles .... 363 Account of the Discoveries in Egypt ... . 364 Study of the Egyptian Finds 365 Execution, Material, Color. Tapestry 370 Three Periods of Design 371 Early Classical Tapestry 371 Deterioration in the Fourth Century .... 373 The Byzantine Period, Christian Themes, etc. . . 374 Importance of these Designs in Relation to Contemporary Art 377 Altar Cloths and Curtains 377 VI. CIVIL AND ECCLESIASTICAL DRESS In General .... 383 Subordinate Elements of Dress, — Head-dress, Foot-gear, Breeches . . .... . . 386 The Tunic 389 The Dalmatic . . . 394 The Pjenula (Chasuble) 396 The Clamys 399 The Lacerna-Byrrus 400 The Toga . 401 The Pallium and the Pallium Scarf 403 The Stole and the Maniple 410 APPENDICES SELECT BIBLIOGRAPHY 415 INDEX 427 LIST OP ILLUSTRATIONS FIG. 1. 9. 10. 11. 12. 13. 14. 15. 16. Plan of part of the catacomb of Domitilla Marucchi, Elements d' Archeologie Chretienne, Tome 2. A cubiculum in the catacomb of Callistus, one of the so-called sacrament chapels . Reber, Kunstgeschichte. Capella greca in the catacomb of Priscilla . ... Wilpert, Fractio Panis. Section of the catacomb of Callistus De Rossi, Roma Sotterranea. Crypt of S. Csecilia, catacomb of Callistus De Rossi, Roma Sotterranea. A crypt in Cyrene Garrucci, Storia delVArte. Fresco of the arcosolium of the fossor Diogenes, S. Callistus Garrucci, Storia delVArte. A sarcophagus from the catacomb of Priscilla . Wilpert, Fractio Panis. Lead coffin from Phoenicia Garrucci, Storia delVArte. Papal crypt in S. Callistus, restored .... De Rossi, Roma Sotterranea. An arcosolium in the catacomb of Cyriaca Bull, crist. Tombstone with graffito in Museo Kircheriano . Wilpert, Fractio Panis. Fresco in a crypt in Cyrene .... Garrucci, Storia delVArte. Decoration in stucco, ceiling of a crypt Garrucci, Storia delVArte. An orans, detail of a fresco in S. Priscilla . Wilpert, Gottgeioeihten Jungfrauen. Sarcophagus of Livia Primitiva Garrucci, Storia delVArte. PAGE 24 26 28 31 33 34 36 41 42 45 48 51 55 57 60 62 LIST OF ILLUSTRATIONS FIG. 17. Sarcophagus Garrucci, Storia delVArte. 18. Tombstone with graffito in the Lateran Photograph. 19. Tombstone in the catacomb of Domitilla Wilpert, Fractio Panis. 20. Fragment of the monument of Abercius Wilpert, Fractio Panis. 21. Loculus of a martyr in S. Domitilla . De Rossi, Bull, crist. 22. The martyrdom of S. Achilleus, stone relief Garrucci, Storia delVArte. 23. Inscription by Pope Damasus in the papal crypt Grisar, Analecta. 24. Crypt of S. Januarius, Catacomb of Pretextatus Garrucci, Storia delVArte. 25. Plans of typical Greek and Roman houses Dehio and v. Bezold, Baukunst. 26. Plans of basilicas .... Dehio and v. Bezold, Baukunst. 27. Plans of basilicas Dehio and v. Bezold, Baukunst. 28. Plans of basilicas .... Dehio and v. Bezold, Baukunst. 29. Plan of old S. Peter's .... Reber, Kunstgeschichte. SO. View of old S. Peter's Kraus, Oeschichte der Kunst. 31. Church in Babuda, Syria . De Vogu6, Syrie Centrale. 32. Basilica in Turmanin, Syria, (facade restored) . De Vogue, Syrie Centrale. 33. Basilica in Turmanin, Syria, (apsidal end) De Vogii6, Syrie Centrale. 34. Basilica in Ruweha, Syria, (longitudinal section) De Vogue, Syne Centrale. 35. Basilica in Kalb-Luseh, Syria, (section) . De Vogii6, Syrie Centrale. 36. Cathedral of Torcello, (interior) Holtzinger, Altchristliche Architektur. 37. S. Apollinare Nuovo, Ravenna, (interior) . Lubke, Gnindriss. PAGB 63 64 68 70 71 74 75 76 99 102, 103 104, 105 107 109 no in 112 114 116 117 119 121 LIST OF ILLUSTRATIONS FIG. 38. S. Maria in Cosmedin, Rome, (interior restored) Giovenale, S. Maria in Cosmedin. 39. Cathedral of Parenzo Photograph. 40. Apse of the basilica Severiana, Naples De Rossi, Bull, crist. 41. Restored basilica of S. Paul, Rome, (interior) . Photograph. 42. Stone windows Various sources. 43. Plans of buildings of central type, simple plan . Dehio and v. Bezold and others. 44. Plans of buildings of central type, composite plan Dehio and v. Bezold and others. 45. Early Christian capitals Holtzinger, Altchristliche Architektur. 46. Church of S. George, Ezra, Syria, (longitudinal section) De Vogue, Syrie Centrale. 47. Plans of Byzantine churches Dehio and v. Bezold, Baukunst. 48. S. Vitale, Ravenna, (interior) Photograph. 49. The mausoleum of Galla Placidia, Ravenna, (exterior) Photograph. 50. The mausoleum of Galla Placidia, Ravenna, (interior) Photograph. 51. The church of S. Sophia, Constantinople, (interior) Photograph. 52. Plans of S. Sophia and the basilica of Maxentius Dehio and v. Bezold, Baukunst. 53. The mausoleum of Theodoric, Ravenna Lubke, Grundriss. 54. Altar from S. Que"nin De Fleury, La Messe. 55. Altar with confessio in S. Giorgio in Velabro, Rome . Holtzinger, Altchristliche Architektur. 56. Altar with fenestella in S. Alessandro, Rome . De Fleury, La Messe. 57. Chancels of fourth and fifth centuries, Rome Mazzanti, Scultura ornamentale. 58. Chancel in S. Clemente, Rome . De Eleury, La Messe. PAGE 122 125 127 128 130 1.35 138, 139 140, 141 144 LIST OF ILLUSTRATIONS FIG. 59. Ambon in Thessalonica De Pleury, La Messe. 60. Cantharus formerly in atrium of S. Peter's De Rossi, Bull, crist. 61. Basilica with peribolos in Ruweha, Syria . De Vogii6, Syrie Centrale. 62. Complex of church buildings at el Barah, Syria De Vogue, Syrie Centrale. 63. Church of S. Simon Stylites, Kalat^Seman, Syria De Vogue, Syrie Centrale. 64. Orans in palla, fresco in S. Callistus . Wilpert, Qewandung der Christen. 65. Orans in dalmatic, fresco in S. Callistus . Wilpert, Qewandung der Christen. 66 Orans in girdled tunic, fresco in the catacomb of Thrason Wilpert, Qewandung der Christen. 67. The vindication of Susanna, fresco in S. Callistus Wilpert, Sacramentscapellen. 68. The Epiphany, fresco in the catacomb of SS. Petrus and Marcellinus Wilpert, Ein Cyclus christologiseher Geinalde. 69. The Good Shepherd, ceiling fresco in the catacomb of SS. Petrus and Marcellinus . . . Wilpert, Qewandung der Christen. 70. Veneranda and S. Petronilla, fresco in S Domitilla . Wilpert, Gewandung. 71. Christ consecrating the Eucharist, fresco in one of the Sacra- ment Chapels, S. Callistus Wilpert, Sacramentscapellen. 72. Christ and the woman at the well of Samaria, fresco in one of the Sacrament Chapels, S. Callistus Wilpert, Sacramentscapellen. 73. Baptism and the Eucharist, fresco in one of the Sacrament Chapels, S. Callistus . . . Wilpert, Sacramentscapellen. 74. " The breaking of bread," fresco in the Capella greca, S. Pris- cilla . . ... Wilpert, Fractio Panis. 75. The mystic fish, fresco in the crypt of Lucina . . . 230 Marucchi, Elements d' Archeologie Chretienne. 76. Carved gems . . Garrucci, Storia delVArte. PAGE 174 179 181 182 183 202 203 204 210 211 216 222 224 220 228 231 233 LIST OF ILLUSTRATIONS FIG. 77. Carved gems Garrucci, Storia delVArte. 78. Carved rings Garrucci, Storia delVArte. 79. Coin of Apamea Garrucci, Storia delVArte. 80. Various forms of the cross and the Constantinian monogram 81. Coins of Constantine Garrucci, Storia delVArte. 82. Coins of Nepotianus and Galla Placidia Garrucci, Storia delVArte. 83. Crux gemmata, fresco in the catacomb of Pontianus Kraus, Gesehichte der Kunst. 84. The Virgin and Child, fresco in S. Priscilla Wilpert, Gewandung. 85. A virgin taking the veil, fresco in S. Priscilla . Wilpert, Gottgeweihten Jungfrauen. 86. Lid of a sarcophagus in the Lateran . Photograph. 87. Lid of a sarcophagus in the Lateran . Garrucci, Storia delVArte. 88. Fragment of a sarcophagus from S. Callistus Bull, crist. 89. Lid of a sarcophagus in the Lateran . Garrucci, Storia delVArte. 90. SS. Peter and Paul, bronze medal in the Vatican Garrucci, Storia delVArte. 91. SS. Paul and Peter, bronze medal in the Vatican Garrucci, Storia delVArte. 92. An orans, relief on a screen .... Wilpert, Gottgeweihten Jungfrauen. 93. Sarcophagus in the Lateran Museum . Photograph. 94. Sarcophagus in the Lateran Museum . Photograph. 95. Sarcophagus from S. Paul's, now in the Lateran Photograph. 96. End of a sarcophagus in the Lateran Museum . Beissel, Bilder aus der Gesehichte der Altch. Kunst 97. Elijah, fragment of a sarcophagus in the Lateran Photograph. PAGE 234 235 237 . 238 240, 241 242, 243 244 . 245 . 246 . 248 . 249 . 250 . 250 . 251 . 251 . 263 . 254 . 255 . 256 257 . 258 LIST OF ILLUSTRATIONS 98. Sarcophagus in the Lateran Museum Photograph. 99. Sarcophagus in the Lateran Museum Photograph. 100. Sarcophagus of Junius Bassus Bom. Quartalschrift. 101. Details of the sarcophagus of Junius Bassus Beissel, Bilder aus der Geschichte der Altchr. Kunst. 102. Sarcophagus in the Lateran Museum Photograph. 103. Sarcophagus of Theodore, Bishop of Ravenna . Photograph. 104. Sarcophagus in Ravenna Photograph. 105. 106. Parts of the wooden doors of S. Sahina, Rome . Photograph. 107. The Ascension, panel of the doors of S. Sahina Photograph. 108. The Crucifixion, panel of the doors of S. Sabina Grisar, Analecta. 109. Ivory box in the British Museum .... Garrucci, Storia delVArte. 110. Sculpture in ivory at Treves Garrucci, Storia delVArle. 111. Part of ivory box of Brescia Photograph. 112. Part of ivory box of Brescia Photograph. 113. Ivory diptych in Florence Photograph. 114. Ivory cathedra of Maximianus at Ravenna Photograph. 115. Details from the cathedra of Maximianus Photograph. 116. Ivory Gospel cover, Ravenna Photograph. 117. Statue of the Good Shepherd in the Lateran Museum Photograph. 118. Statue of S. Hippolytus in the Lateran Museum Photograph. 119. Bronze statue of S. Peter in the Vatican . Photograph. PAGE . 260 . 261 . 262 264, 265 . 267 . 268 . 269 270, 271 . 274 . 277 . 279 . 280 . 282 . 283 . 284 . 28G . 287 . 288 . 289 . 290 . 291 LIST OF ILLUSTRATIONS FIG. PAQK 120. Mosaic pavement of apse, Ancona 294 Bull, crist. 121. Section of mosaic pavement in Parenzo 295 Lohde. 122. Marble incrustation in S. Sabina 296 De Rossi, Mosaici. 123. Inscription in mosaic, S. Sabina 296 De Rossi, Mosaici. 124. Portrait in mosaic, catacomb of Cyriaca 297 De Rossi, Mosaici. 125. Mosaic decoration of ring vault in S. Costanza . . 298, 299 De Rossi, Mosaici. 126. Sketch of the mosaics of the dome, S. Costanza . . . 800 Garrucci, Storia delVArte. 127. Apsidal mosaic in S. Pudenziana, Rome 306 De Rossi, Mosaici. 128. Plan of Jerusalem, part .of mosaic pavement of a church at Madaba, Palestine . . 307 Marucchi, in Nuovo Bull, crist. 129. Apsidal mosaic in SS. Cosma and Damiano, Rome . . . 312 De Rossi, Mosaici. 130. Mosaic of the arch of S. Lorenzo, Rome 313 De Rossi, Mosaici. 131. Mosaic in S. Apollinare in Classe 314 Photograph. 132. The sacrifices of Abel and Melchisedek, mosaic in S. Vitale, Ravenna 314 Photograph. 133. Abraham entertaining the angels, mosaic in S. Vitale . . 315 Photograph. 134. The Emperor Justinian carrying a votive paten, mosaic in S. Vitale 316 Photograph. 135. The Empress Theodora carrying a votive chalice, mosaic in S. Vitale 318 Photograph. 136. Mosaic of the arch of S. Maria Maggiore, Rome . 320, 321 De Rossi, Mosaici. 137. Mosaics in the nave of S. Maria Maggiore, Rome . 322, 323 De Rossi, Mosaici. 138. Christ enthroned, with angels and saints, mosaic in S. Apolli- nare Nuovo, Ravenna 324, 325 Photograph. LIST OF ILLUSTRATIONS and saints, mosaic in . 326, 327 328 30 FIG. 139. Virgin and Child, with angels, Magi S. Apollinare Nuovo, Ravenna . Photograph. 140. S. Luke, mosaic in S. Vitale, Ravenna Photograph. 141. The Good Shepherd, mosaic in the mausoleum of Galla Pla- cidia, Ravenna . . . . Photograph. 142. S. Ursicinus, mosaic in S. Apollinare in Classe, Ravenna Photograph. 143. Palace of Theodoric, mosaic in S. Apollinare Nuovo, Ravenna Photograph. 144. Mosaic in S. George, Thessalonica Texier and Pullen, L' Architecture Byzantine. 145. Joshua and the Gibeonites, miniature from the Joshua Roll of the Vatican . . Graven, in L'Arte. 146. Joshua and the men of Ai, miniature from the Joshua Roll of the Vatican Graven, in VArte. 147. Pharaoh's feast and Rebecca at the well, miniatures from the Vienna Genesis 336, 337 Hartell and Wickoff, Die Wiener Genesis. 148. David, miniature from a Greek Psalter in Paris Hartell and Wickoff, Die Wiener Genesis. 149. Censers . . De Fleury, La Messe. 150. Terra-cotta lamps Garrucci, Storia deWArte. 151. Martyr exposed to a lion, terra-cotta lamp Bull, crist. 152. Terra-cotta lamp in the form of a fish Garrucci, Storia deWArte. 153. Bronze lamp in the form of a basilica Garrucci, Storia deWArte. 154. Bronze lamp in the form of a sheep . Garrucci, Storia deWArte. 155. Jonah under the gourd, bronze lamp Garrucci, Storia delVArte. 156. Lead cup with reliefs, from Carthage Garrucci, Storia delVArte. 331 332 332 333 334 338 346 348 349 350 350 351 355 356 LIST OF ILLUSTRATIOXS xxl FIG. PAGE 157. Lead cup from Carthage, reliefs displayed .... 356 Garrucci, Storia delVArte. 158. Metal ampulla in Monza 356 Garrucci, Storia dell' Arte. 159. Bronze medal ... 357 Garrucci, Storia delVArte. 160. Fragment of cut-glass vessel representing Biblical subjects . 358 Bull, crist. 161. Silver box from Africa 360, 361 Bull, crist. 162. The Good Shepherd, three gold-glasses 362 Garrucci, Storia delVArte. 163. Daniel and Bel, Adam and Eve, gold-glasses .... 363 Garrucci, Storia delVArte. 164. Jonah, two gold-glasses 364 Garrucci, Storia delVArte. 165. Fish under a gourd — Jonah as a type of Christ, a gold-glass . 365 Garrucci, Storia delVArte. 166. Daniel and Bel, gold-glass . . ... 365 Garrucci, Storia delVArte. 167. Family portraits, gold-glasses . 366 Garrucci, Storia delVArte. 168. A cycle of miracles depicted on a gold-glass . . 367 Garrucci, Storia delVArte. 169. Representation of a master carpenter, a gold-glass . . 368 Garrucci, Storia delVArte. 170. Segments, classical designs in tapestry ... . 371 Forrer, Fruhchristlichen Alterthumer. 171. The Church and the Empire united in the suppression of evil, woven silk design . . . . . 372 Forrer, Fruhchristlichen Alterthumer. 172. "Virgin and Child, segmentum in tapestry 373 Forrer, Fruhchristlichen Alterthumer. 173. The eye, design in tapestry .... . 374 Forrer, Fruhchristlichen Alterthumer. 174. The Egyptian cross, part of a clavus in tapestry . 374 Forrer, Fruhchristlichen Alterthumer. 175. Reconstruction of a curtain 375 Swadoba, Archaologische Ehrengabe. 176. Curtain in the Victoria Albert Museum .... 376 Photograph. xxii LIST OF ILLUSTBATIONS FIG. PAGE 177. Part of a clavus in silk embroidery 378 Forrer, Fruhchristlichen Alterthiimer. 178. Details of a scarf in silk embroidery 379 Forrer, Fruhchristlichen Alterthiimer. 179. Portrait in tapestry ... 380 Forrer, Fruhchristlichen Alterthiimer. 180. Pope Sixtus I. and the Bishop Optatus, fresco in the catacomb of Callistus . . .... 397 Wilpert, Tin Capitolo delta Storia del Vestiario. 181. Mosaic in the chapel of S. Venantius, Lateran Baptistery 404, 405 De Eossi, Mosaici. 182. Linen maniple from Achmim 412 Forrer, Fruhchristlichen Alterthiimer. MONUMENTS OF THE EARLY CHURCH A HANDBOOK OF CHRISTIAN ABCHJEOLOGY INTRODUCTION It would be but a bald definition of archaeology to say that it is the study of ancient monuments. Like every other branch of history its ultimate interest is not the documents with which it deals, but the human story to which they bear wit- ness. The story of many ancient peoples must be traced chiefly, if not solely, in their monumental remains, and the historian can ill afford to dispense with archaeology, be the literary records never so plenty. Far from being the dullest branch of history, archaeology is one of the most interesting, for it serves to illuminate precisely the most intimate and familiar aspects of ancient life, depicting them most vividly and most concretely. The archaeology of the early Christian period is not essen- tially different from any other branch of the study, though it is, of course, marked by problems and peculiarities quite its own. The most distinctive peculiarity which need be noticed is this, that it is not engaged with the totality of the monu- ments of any people or of any period, but solely with those which bear expressly the stamp of Christian inspiration. Christianity was not conterminous with any race, and the existence of the Church within the Empire did not signify the propagation of foreign art and custom in the midst of the regnant Classical civilization. Already with the second century — the period with which the monumental study of Christianity begins — the gentile element was so thoroughly predominant in the Church that we need not look for the tradi- tions of Jewish custom except in matters most intimately associated with religion. The Jew had no art to hand on to the Christian, nor any architecture. The last temple at Jerusalem was a Classic building, as the first had been a Phoenician. Moreover, the Jew throughout the Empire was b 1 2 INTRODUCTION already as much of a cosmopolitan as he is to-day. A certain aloofness from popular gentile life, prescribed by his religion and his moral code, must not blind us to the thoroughness of his tincture with classical civilization — he spoke in its language, he thought in its philosophy, he clothed himself in its dress. The emperors were not slow to recognize in the Hellenistic Jew an apt exponent of the imperial ideal, and they accorded him corresponding privileges. In the case of the Jewish Christians, as we see already in the Apostolic age, the ties of race and the barriers of racial exclusiveness were still further dissolved, and in the case of the gentile Christians they did not exist at all — they stood apart from their fellow- citizens only in matters which regarded religion and morality. In general, one can hardly exaggerate the Koman's success (building, indeed, in part upon the work of the Greek) in im- pressing upon the world, not only the administrative regime of the Empire, but the whole complex of Classical civilization : it was in the fulness of time Christ came and the Apostles preached. The Church was a phenomenon of the Empire, and in time of persecution no less than in the era of triumph belonged essentially to it. To say that the Church grew out of the Graeco-Roman civilization (as though transcending it), or grew up within it (as an isolated entity), would be mislead- ing; the Church must be recognized as a very part of that antique classical civilization, and only by inwardly appropriat- ing — not by adoption but by natural inheritance — the funda- mental terms of its life and thought, was it able in spirit to transform it. If one were bent upon the reconstruction in its totality of Christian life in the Classical period, the monuments which are reckoned to the part of Classical archaeology would con- stitute data no less important than those which bear expressly a Christian character. And though we must here make com- plete abstraction of the common elements of Classical life, we cannot ignore the fact that the art which was the most distinctively Christian in spirit was none the less Classic in form. It was simply inevitable that all forms of decorative and pictorial art, of sculpture and architecture, should be adopted by the Christian as he found them ready to his hand. We see how thoroughly he belonged to his age when we CLASSIC AND CHRISTIAN ART 3 note with how little scruple he accepted them. He was con- cerned in the first instance only to eliminate such current themes as were expressly idolatrous or immoral, or to select such as could be given a Christian interpretation. This was followed, of course, by the development of themes properly Christian ; but even in this work the Christian artist could no more divest himself of the formal traditions of Classic art — nor had he any more reason to do so — than in vising the Classic languages he could dissociate himself from the funda- mental modes of Classic thought. There was no doubt also a formal progression upon the old lines ; but the antique traditions were so soon broken off in the general collapse of Roman civilization that we are left without any sound basis for the speculation as to whether, or in how far, Christian art with a free and interrupted devel- opment might finally have transcended the traditional forms. This speculation, at any rate, does not concern us here. We shall not trace the old traditions through the petrified forms of Byzantine art, where they proved impotent for any scope beyond symbolism and decoration. It shall be left to the reader, with the monuments which are here put before him, to reckon the influence of Classical Christian art upon art Mediaeval : this influence is to be traced rather in the selec- tion of themes and in the arrangement of scenes than in spirit and technic. The Renascence, which returned for its inspira- tion Classic art, had no relation whatever (except through the Mediaeval tradition) with the Christian phase of that art ; its monuments were already hidden or ignored. This Classical Christian art proved no less adequate to the expression of Christian themes, and deserves no less truly the title of Christian art, than did any which was developed later and on Christian soil. This art has long been completely un- known, and not even yet is it popularly appreciated. Yet it remains more intelligible to the modern mind, it stands closer to us in sympathy, than does the art which stands nearer in time, developed during the Middle Ages by men of our own stock and under influences exclusively Christian. Classic art, even in this its Christian phase, reflects, in its lucid ideal and definite presentation, the moderation and balance which were characteristic of Classical civilization and thought. There is 4 introduction: fundamentally but one art and one reason : not a Christian art and a Christian reason. The proof that Christian theology was expressed in forms of Greek philosophy has been accounted to the discredit of the Church. On the contrary, it was the signal advantage of Christianity that it could from the begin- ning express itself in the forms of Greek art and in the terms of Greek thought ; for it was true art and right reason. The normal authority which we ascribe to the early age of the Church is justified not only by the fact that it stood close to the origin of the faith and constituted the first expression of Christianity in the world, but also by the fact that it was the expression of Classical culture. The pagan Classical age is one against which many serious charges may be brought. One may be by no means disposed to recognize the legitimacy of the customs and theologies which sprang up on that soil. But one can hardly refuse to recognize that it was an age which was marked, as no other has been, by the general dissemination of just criteria of reason. It is a matter of no slight advantage that the philosophic claims of Christianity were first expressed in the Greek language, and, therefore, in terms intelligible to the world — in terms which then were comprehensible to dis- ciples of Plato, of Aristotle, and of Zeno, and which remain intelligible for all times. One must recognize in early Christian art two qualities which are not commonly thought of as compatible in the same subject : it was an art at once nascent and decadent. In many respects it was full grown when born, and its decline dates from its beginning. The highest expression which art attained in the Roman Empire was reached in the time of Trajan and Hadrian. Its decline during the next two centuries is meas- ured upon the Arch of Constantine by the difference between the sculptured reliefs taken from the demolished Arch of Trajan and those which the first Christian emperor had exe- cuted expressly for his own monument. Within two centuries more art lay almost prostrate. We have not, therefore, in the case of Christian art, to look for the rude essays which we associate with all beginning, the effort, with little skill, to match the idea with the appropriate form. On the other hand, we cannot trace in it the keen progress upward which THE CHURCH AXD ART 5 constitutes the special charm of the study of early periods of art. Christian art began — so far as we can trace its monu- ments — with the end of the first century, that is with the most flourishing period of imperial art. It found its forms furnished ready to hand. The earliest Christian artists were technically the best; their skill declined with the general collapse of Roman art and culture. The monuments reveal just such use and appreciation of art in the Church as might be expected of gentiles who were bred to a taste for it and found in their new religion no express veto against it. Apart, however, from the extant monuments, and relying solely upon the scant references in literature, one would hardly form a just conception of the Christian employ- ment of art in the second and third centuries. One might be disposed rather to credit the opinion that the early Church • set its face resolutely against its religious use, and that the artistic activity of the fourth century, to which the literature as well as the monuments bear witness, was a sign of the worldly corruption of the Church under imperial favor. This view was so confidently held in Protestant circles, and espe- cially in England, that for a long time credence was refused to the discoveries of early Christian art in the Roman cata- combs. It was a view which rested fundamentally upon a strong religious prejudice ; but it seemed none the less to have some justification in history. For it seemed not unrea- sonable to presume that the Mosaic prohibition of pictorial art must have been observed in the Church as it was in Islam, there being the same reason for it in the keen struggle with surrounding idolatry. And, in fact, the few literary references to art which we have (in Tertullian, Clement of Alexandria, and Irenseus), though they do not express a condemnation of it, suggest distinctly the suspicion of its abuse. The Spanish Synod of Elvira (306 a.d.), with its prohibition of the pictorial representation of the divine Persons in the churches (can. 36), has been taken for conclusive evidence of the general attitude of the Church. Eusebius of Csesarea, — in other respects no enemy of art, — writing, in reference to a picture of Christ, to Constantia, sister of Constantine the Great, suggests the scruple that strangers, seeing such repre- sentations, might suppose that the Christians carry about their 6 INTBODUCTION god in a picture like the idolaters. It is likely that the same fear of misunderstanding prompted the canon of Elvira ; and for the faithful themselves there may have been no little dan- ger in the painting of " that which is worshipped and adored " upon the walls of the churches. This utterance, however, does not represent the general atti- tude of the Church, which continued with little or no scruple (so far as monuments reveal) to employ pictorial art for the representation of religious themes and of the divine Persons. The monuments show us, too, that we have not to think of a strict observance of the Second Commandment even on the part of the Hellenistic Jews. One of the Jewish catacombs in Rome contains decoration in fresco representing animal forms such as are found in the Christian cemeteries. There is similar decoration in the Jewish cemetery at Carthage ; and in Tunis the mosaic pavement of a synagogue is entirely in Classic design, with garlands, fishes, and birds. We may admit that such Jewish Christian communities as the Ebionites probably respected the mosaic prohibition. It is not unlikely that among the Christians a severe sect like the Montanists "held aloof from art. There were voices raised against its religious use, or abuse, both before and after the iconoclastic Eastern emperors. But the enthusiasm with which art was cultivated by the Catholic Church is sufficiently demonstrated by the monuments themselves, or even by the examples which are illustrated in this book. It is from no point of view an easy matter to fix the limits of the early Christian period, although the question is simpli- fied when we regard it with exclusive reference to the history of art, and recognize that the theme of our study is Classic art "as it was employed by Christians for the expression of Christian ideals. From this point of view it is plain that early Chris- tian art ends long before Mediaeval art begins. Between the two there intervenes a period of about two centuries, a true middle age, in which Classic art was no longer and Mediaeval art was not yet. It adds little to our information to follow through this period the merely static survival of Classic forms in the Byzantine Empire ; and it is in no sense germane to our topic to study the purely barbaric art of the Anglo-Saxons, LIMITS OF THE EARLY CHRISTIAN PERIOD 7 Celts, or Lombards. Practically, our study begins with the earliest extant monuments of a strictly Christian character, that is with the end of the first century, and it closes with the development of Byzantine architecture before the end of the sixth. It must be remarked, however, that no single depart- ment of our study covers the whole of this period : each branch of art had in some respects an independent history and a separate fate, originating later or terminating earlier than the limiting dates which are here assigned to early Christian art in general. The precise limits of the period are of the less importance to our present study because it is rather on the plane of descrip- tion than of narrative, being interested not so much to trace the origins of Classical Christian art, nor to follow its decay, as to portray it in its perfection. There is needed here, by way of introduction, no more than the briefest survey of the various classes of monuments with which we have to deal. The early Christian painting which has been preserved to us is found almost exclusively in the Itoman catacombs : it there- fore belongs to the earliest part of our period, and may be pre- sumed to have a more or less immediate reference to death — or rather to the Christian hopes which illuminated it. The themes of later painting — especially in the decoration of the churches — can be inferred from the mosaics. On the other hand, light is thrown upon the study of the mosaics by the latest branch of painting which we have to consider, namely, the miniature illustrations of the Bible text. The catacombs themselves, quite apart from the monuments of art which they contain, constitute a testimony of the highest importance both for the early age during which they were in practical use and for the later in which they had become objects of veneration. It lay obviously in the nature of the case that the art of sculp- ture was not utilized as soon as painting for the representation of Christian themes. It is represented principally by the sarcophagi, which are most abundant for the fourth and fifth centuries, and were produced until the seventh, when artis- tic skill sank below the level demanded for figured work. The same period is represented, and the same history of decline is repeated, in ivory and metal work. In general the minor 8 INTRODUCTION arts followed the Peace of the Church, but the manufacture of gold-glass commenced toward the end of the third century. The discoveries in Egyptian burying-grounds enable us to follow the technic of the textile industry from the second century to the Mahometan conquest; but there were no dis- tinctively Christian designs till the age of the Church's pros- perity. In architecture there remain but few monuments earlier than the Constantinian age. The churches built during the reign of Constantine have been destroyed or remodelled. The following century, however, furnishes many complete examples. With the reign of Justinian in the sixth century there was prodigious activity in the building of churches, and many edifices have been preserved. Specially notable in this period is the development of the central type of architecture. The history of mosaic art follows closely that of architecture, of which it was the handmaid. The mosaic decoration has, however, often been preserved when the strictly architectural features of the church have been transformed out of all recog- nition. Such, in the most general terms, are the monuments which we have to study. But it must be remembered that to the archaeologist each monument has a many-sided interest ; a mosiac painting, for instance, may be studied as a part of church decoration, or as an example of pictorial art, or for its illustration of ancient costume, of early church doctrine, or of ecclesiastical use. Local differences in art due to racial traditions can be but barely noticed in this book. They have not as yet been suffi- ciently studied ; in many cases where we might expect to note them the monuments are too rare to furnish the necessary evidence. In general the monuments do not allow us to forget that they all belong — in spite of incidental peculiarities — to the broadly disseminated culture of the Roman Empire. It is in the East especially we might expect to notice the influence of earlier art ; it is there, too, we might expect the greater abundance of monuments; for it was the East — especially Asia Minor, Syria, and Egypt — which was the most flourishing seat of Christianity in the third century and even later. But, as a matter of fact, it is particularly in the East DISTRIBUTION OF MONUMENTS 9 that the monuments have been either most completely destroyed or least explored. Christian art in the East fell victim finally to the iconoclastic spirit of Islam. Valuable monuments of early architecture have been here and there preserved by their adoption as mosques, particularly in Constantinople, Thessa- lonica, and Jerusalem. But already pictured art had suffered much from iconoclastic zeal in the Church itself. Many of the pictures which are venerated in Italy as original portraits or miraculous images (acheiropoeta) of Christ and of the Virgin belong to the Byzantine art of the seventh century and were, doubtless, transported for preservation. To this general destruction Syria forms an important excep- tion, especially in view of its architectural remains. The con- quest of Islam left the greater part of it a desert, and owing to the admirable character of its masonry (and to the fact that little or no wood was used) it furnishes the most numerous and the most perfect examples of the domestic and ecclesiastical architecture of the third and fourth centuries ; and not only in its architecture, but in its carved reliefs, it bears witness to a national Syrian school, which probably had a wider influence than any other local art upon the surrounding peoples. The Holy Land — especially Jerusalem and Bethlehem, but in general the whole of Palestine — was a resort of pious pil- grimage and the site of many churches of imperial foundation, due first to Constantine and his mother Helena, and later to Justinian. Scanty remains are still left of some of these churches, though the Crusaders in their zeal for rebuilding showed little respect for ancient monuments. We- might expect in Egypt a very strong influence upon art from the ancient national traditions and the ancient monu- ments. As a matter of fact, the Coptic churches have been either so thoroughly destroyed or so fundamentally altered that hardly more of the original than the ground plan can now be descried and the textiles and small objects found in the graves present very little which is distinctive. Asia Minor has as yet been insufficiently explored. Archaeo- logical study will naturally follow in the wake of the German commercial exploitation of the country, and there is reason to hope for finds of great importance. How much may survive Mahometan occupation we see in the case of North Africa, 10 INTRODUCTION where French colonization has brought to light many important monuments of architecture, sculpture, and the minor arts. But this brings us already to the West, where the monuments are in general more common, or better known, and where de- struction has been due chiefly to perverse activity in restora- tion. On the borders of the Empire monuments belonging strictly to our period are for the most part very scarce. In England they are almost exclusively confined to mosaic pave- ments, especially at Trempton and Horkstow. In Spain they are represented chiefly by sarcophagi ; in France by sarcophagi (at Aries), by a few churches which are almost entirely rebuilt, and by rare specimens of the minor arts. In Germany, Treves is specially interesting for three cemeteries (one of them under ground) ; numerous small objects and inscriptions have been found also in the Rhine country. Outside of Italy it is the northern coast of the Adriatic (Dalmatia and Istria), with its purely Latin colonies, which proves the greatest activity in artistic production, especially toward the sixth century. Here are found architectural monuments which have lately been well studied by scholars native to the region who are peculiarly zealous for their classical traditions. But it is Italy itself, and preeminently Rome and its vicinity, which furnishes the greatest number of monuments — • indeed, the majority of all which have been found. In Rome the very activity of ecclesiastical life, the fact that no period has been altogether lacking in additions and reconstructions, has con- tributed to the destruction of the monuments ; but it has also served to present an almost unbroken sequence representing nearly every department of Christian art. If the monuments of the Middle Ages (both early and late) have been almost totally annihilated by the scorn of the Renascence for what it counted an age of barbarism, the earlier monuments have fared somewhat better. The student is, indeed, likely to be disap- pointed at first, for what is left requires much patient hunting out and piecing together; but, after all, there are sufficient monuments in the way of architecture and mosaics to repre- sent at least intelligibly the period of Byzantine influence, the more strictly Roman style of the fifth century, and even the Constantinian foundations. Sculpture is fully represented throughout this period by sarcophagi, by the wooden doors of LITERARY SOURCES 11 S. Sabina, and by minor objects. For the earlier period it is Rome alone, -with its catacombs, which, provides the key for the appreciation of the development of Christian art, and of the spirit with which it was employed by the Church. Sicily is interesting for its cemeteries, especially the cata- comb of S. Giovanni at Syracuse. Naples also has its primi- tive catacomb of S. Gennaro, and in the Campagna there are other cemeteries of less interest. In the North, Milan furnishes sarcophagi, ivory carvings, and churches of the sixth century (very much altered) ; and Ra- venna, with its churches, mosaics, and sarcophagi, splendidly testifies to the artistic interest and skill which marked both the Gothic and the Byzantine rule at the turn of the fifth and sixth centuries. In enumerating the monuments which specially distinguish each place and period, it has not been convenient to speak of inscriptions, because they are to be found throughout the whole field. It is, however, in Rome and Italy that they are by far the most numerous ; the inscriptions of Gaul stand next in importance. In the East they still, for the most part, await discovery. All but a small fraction of our Christian inscrip- tions are sepulchral. It need not be said that the study of the early history of the Church is not dependent solely upon monuments, nor that the study of archaeology itself is chiefly illuminated by docu- mentary evidence. It was not a barbarous, but a literary age ; and whatever of the literature of the Church — or more espe- cially its official records — may have been destroyed by perse- cution, the stable basis of its history is still a literary one. But the archaeological study of the monumental remains is only the more fruitful for being less than indispensable; its problems are more readily and more surely solved, since it is hardly ever left without some word of interpretation from contemporary writers. The literary sources which contribute to fix our conceptions of Christian archaeology cannot, how- ever, be mentioned in this work ; it is enough that we present here the sources of monumental evidence. It goes without saying that the works of the Fathers and the History of Eusebius, for example, are indispensable for any study of our 12 INTRODUCTION subject, and even for the appreciation of its monuments. The comparatively few passages which may be quoted in this book do not by any means reveal the extent to which the early authors must be depended upon in formulating the conclusions of this science. But there is one class of writings which stands in so close a relation to archaeological study, and is of so little importance apart from it, that the student may need a special introduction, if but a brief one. It is necessary to pass over with a mere word a class of documents which were of inestimable importance to de Rossi and to earlier students for the location and exploration of the catacombs, because the discovery to which they conduce has been all but definitely accomplished. I refer to the Roman martyrologies (the most important is that which is associated with the name of S. Jerome), which are liturgical lists giving the day of the annual celebration of the death, deposition, or translation of the saints who were commemorated by the Church, naming also the cemetery in which the body reposed. It is the last item which gives to the martyrologies their topo- graphical value in the work of locating the catacombs and ascribing to them their proper names. The Lives of the Saints had a like use, notwithstanding that of the few which are strictly authentic none belong to Rome. It was de Rossi's generous faith which enabled him to draw even from late tra- ditions a profit which most of the scholars of his time would have been too sceptical to look for. The results of his work justified his belief, and he was able by monumental evidence to reconstruct in some measure the genuine tradition contained in the Lioes with which he had to deal {e.g. that of S. Ceeilia\ subjecting them to a criticism which was impossible on purely literary lines. Comparatively few of the Lives, of course, could be subjected to this test, and for the most of them we must remain without any such criteria for distinguishing the false from the true. But in the proof that in the Lives which have been accounted most hopelessly corrupt we may look for some substantial basis of fact, lies one of the most important general results of de Rossi's work, for it cannot but modify the general sceptical attitude toward ecclesiastical tradition. Another class of documents, the pilgrim Itineraries, proved LITERARY SOURCES 13 of still more value for the topography of the catacombs. They were intended in most cases, not only as a record of the pil- grim's visit, but as a guide to others who might follow him ; and it was with this view the pilgrims were so precise in de- scribing the order and in marking the location of the monu- ments which they visited. In Rome the pilgrims were interested chiefly in the tombs of the saints and the memorial which were erected over them, all of which (with the exception of the house of the fourth- century martyrs John and Paul) were without the walls. In Palestine interest centred, of course, in the sites connected with our Lord's life and the Constantinian structures which marked them. For Rome the itineraries are furnished us exclusively by German and English pilgrims. There are two in a Ms. of Salzburg : the first of them recounts a pilgrimage made shortly after the pontificate of Honorius I. (625-638), and the second belongs approximately to the same date. About the end of the same century, but hardly less valuable topographically, comes the Itinerary of William of Malmes- bury; and early in the century following, one which was found in a codex of the monastery of Einsiedeln. Earlier and of still greater interest — though they describe later monuments — are the accounts of pilgrimage to Palestine. The itinerary of the pilgrims from Bordeaux to Jerusalem (fifth century) has been long known. The Peregrinatio Silvice (supposed to be S. Sylvia of Aquitaine) has been recently dis- covered, and is the most interesting of its class. This is a description by a lady of Southern Gaul of her visit to the Holy Places of Palestine in the fourth century. It was writ- ten for the instruction of her "sisters," that is, doubtless, the members of her religious community. These itineraries describe the Holy Land as it was embellished by the construc- tions of Constantine ; there are two others of later date which describe the buildings of Justinian. To this must be added the description of the holy places of Palestine by the abbot Adamnanus, of the early part of the eighth century. Even more closely related to archaeology is the Roman Liber Pontificcdis. It is a brief record of the official acts of the popes, — ceasing however to be brief when it reaches the popes of the latter part of the sixth century. It was evidently not 14 INTRODUCTION all of it written at one time ; the later lives at least were added one by one by contemporary scribes. The first part of the work, as Duchesne would have it, probably extended only to the middle of the short rule of Silverius (536-537), and was therefore composed by a contemporary writer. Some would date the original composition a century later; but they would not thereby seriously affect its value to the archaeologist. For, serious as are the historical blunders of the first part, it con- tains one element which could have been derived only from the official records of the Church, that is, the account of the con- struction of churches by popes and emperors and of the gifts which were made for their adornment. This is all given with such precision of detail — with number, weight, and measure — that it argues an author who, however ill equipped for the task of an historian, had at his service official records which extended back to the time of Constantine. Among the most interesting notices are those of the gifts of that emperor. The Liber Pontificalis of Ravenna is the work of a single author, the Abbot Agnellus, in the early part of the ninth cen- tury. Agnellus had no such minute records at his disposal as had the author of the Roman work, but on the other hand he had a special talent for his task. He may be said to be the first Christian archaeologist ; it was by the study of the monu- ments of his city that he was able to piece out its history, and the documents which assisted him in his labor were chiefly the metrical inscriptions which were preserved in the churches. The inscriptions which Agnellus records — in part dedica- tory, in part descriptive of works of art — belong to a class of documents more strictly literary than those we have hitherto considered. The Christian poets, like the Greek, often chose for their theme the famous work of some artist, and exhibited their skill in its description. This sort of poetry was current both in the East and in the West, but with marked differences of type. In the East it fol- lowed the Classical models. The Epigrams of Paulus Silentiarius are copied after the Greek form, as is also his poem descriptive of S. Sophia. This latter work was highly lauded by contem- poraries, and it was claimed that a reader might get from it as vivid an impression of the church as if he had actually seen it. Procopius of Caesarea described in prose the buildings of LITERARY SOURCES 15 Justinian, giving us a more exact and detailed account of the building operations of that emperor than Eusebius gives us (in his Life and Eulogy) of the churches of Constantine. In the West, we have, first, the metrical inscriptions which Pope Damasus (366-384) erected in the catacombs in honor of the martyrs. They were frequently copied by the pilgrims, and in that way many of them have been preserved to us, though it seems that they never were gathered into one edition as were the poems of l'rudentius. They belong so properly to our monumental study that they must be treated more at length elsewhere, for in the exploration of the catacombs a number of the original inscriptions have been brought to light. With Prudentius, at the end of the fourth century, we come to a type of descriptive poetry which belongs exclusively to the West and constitutes a development altogether peculiar to Christianity and highly significant of the spirit of Christian art. It is the so-called titulus, a type of poetry which is related to the Greek epigram, but at the same time sharply contrasted with it. Classic art in its perfection strove after a purely formal effect upon the beholder, that is to say, its chief inter- est lay, not in its substantial import, but in its external form. This aim was completely summed up in the expression of the beautiful. Intimately related to the spirit of Greek art was the Greek epigram, which sought in language a clever and pointed expression of the effect which the work of art was intended to produce. The epigram was often inscribed upon the base of a statue, but the statue was none the less essentially independent of it, for its effect was purely formal and it needed no inter- preting word. On the other hand we see in Christian art a spirit which was ever intent upon seeing behind the veil. In Christian art from the very first — in its early symbolical themes no less than in its later and more obviously didactic development — it was not the form but the thought which predominated ; and the Christian work was characteristically accompanied not by the epigram with its clever expression of the formal effect, but by the title (titulus) which summed up no less tersely the substantial point of the narrative or dogma which the picture would represent. The titulus was usually inscribed under a picture, but it had precisely the same char- acter when it was inscribed over a door, upon the altar, or 16 INTRODUCTION referred to the whole church — in any case it was intended to express the essential significance of the object. The titulus had a long history in the Church, and it was continued finally in a prose form quite through the Middle Ages, or until with the Renascence the interest in the formal effect of art became predominant, and — it must also be said — art became better able to tell its own tale. In the Peristephanon of Prudentius we have descriptions of the pictures of the Martyrium of 8. Cassianus in Imola (the ancient Forum Cornelii) and of the Martyrium of S. Hippolytus at Home near S. Lorenzo. In his Dittochceon we see exempli- fied a characteristic of Christian art which we shall have sev- eral occasions to notice in the monuments. The name is com- pounded of two Greek words (Surds — o^i?) and means double nutriment. The titles were evidently composed for actual in- scription, and they represent a parallel series of pictures taken from the Old and New Testaments. A similar typological cycle from both the Old and the New Testaments is that of Helpidius Rusticus of the sixth century. Even more interesting are the poems of Paulinus of Nola (353-431). In his letter to Sulpicius Severus of Priinuliacum in Gaul he rebukes his friend for his too friendly act in paint- ing his (Paulinus's) portrait in the refectory together with that of S. Martin, and he proposes the choice of two titles which would obviate the sacrilege by pointing the contrast between the saint and the sinner. The same letter contains dedications for the baptistery and the basilica which his friend was build- ing, and gives the tituli of his own ehurch of S. Felix at Nola. We have also from S. Paulinus a poem descriptive of the basil- ica of S. Felix ; and another letter contains a poetical descrip- tion of the basilica which he was building at Fundi. We have literary record of the tituli of some of the churches of North Italy of the fifth century : those ascribed to S. Am- brose, and those referred falsely to Claudian. Also in Gaul for the basilica of S. Martin at Tours (about 460), and for the Cathedral of Tours as recorded by Venantius Fortunatus about 565. S. Gregory of Tours (+ 594), in his Historia Francoriun, gives a prose description of the basilica of S. Martin and of the basilica of Namantius in Clermont. The literature which has here been referred to as especially HISTORY OF THE STUDY 17 illustrative of Christian archaeology can most of it be found in the relative volumes of the Quellenschriften fiir Kunstgeschichte (cf. Bibliography). • The history of the study of Christian archaeology need be given here only in the briefest terms : its present status is fairly represented by the Bibliography which is given in the Appendix. The beginnings of a truly monumental study of Christian history are indissolubly associated with the exploration of the Boman catacombs. In 1578 a general interest was awakened by the accidental discovery of a single chamber of one of the catacombs which had lain buried for centuries. This furnished an impetus to study, and early in the seventeenth century two Dutchmen, L'Heureux and de Winghe, wrote works upon the catacombs and upon early Christian art in general which showed careful observation and a fairly just appreciation of the monu- ments. These works would have constituted a hopeful begin- ning for the study, if only they could have gained the attention of scholars, but as a matter of fact they were not even pub- lished. The real beginning of the exploration and of the apprecia- tion of the catacombs was made by Bosio, who after many dis- coveries, after many years of familiarity with the catacombs and a close and fruitful study of their contents, left his work incomplete and unpublished at his death in 1614. It was pub- lished, however, in 1632, under the title Roma Sotterranea, by Severano and in a Latin edition by Aringhi in 1651-1659. How great was Bosio's merit one can best judge from de Kossi's frequent expressions of obligation to him. He had recognized the right method for the study and exploration of the cata- combs, and he only just failed to make the important discov- eries which the world now owes to his successor de Kossi. During the seventeenth century Christian archaeology con- tinued to be chiefly occupied with the catacombs and their con- tents, but without being able to avail itself of the impetus of Bosio's work for new discovery. No substantial progress was made in this line even in the eighteenth century, nor indeed until the discoveries of de Bossi in our own time. For the eighteenth century, however, the works of Boldetti and Bottari served to keep alive an interest in the subject, and already by 18 INTBODUCTION the end of the preceding century Ciampini had widened the field by studying early mosaics and the churches of Constan- tinian foundation. Up to the nineteenth century the field of * discovery was almost exclusively confined to Rome, and the origin and significance of the catacombs and their contents was a fertile ground of dispute between Roman Catholics and Protestants. With the nineteenth century the development of the study has been significant in many respects. A real appreciation of early Christian art was perhaps impossible before Classical art itself was understood, that is before the development of Classical archaeology. At all events, our study has followed the development of that science and has had the advantage of its well-established method. The field of investigation has at the same time been broadened so that substantial contributions are made from all sides and from many lands, though Rome, as the chief storehouse of monuments, remains also the most important centre of study. The study in this its later phase has been marked not only by a scientific method, but by a scientific spirit which has for the most part risen quite above the interests of sectarian polemics. It was through the work of Giovanni Battista de Rossi that Christian archaeology first assumed its modern form and im- portance : it was he who first made it a science. Equipped with everything which was needed to furnish him for the work, with solid learning, and with zeal and ingenuity no less re- markable, he followed out a rigid method which led not only to the discovery of the catacombs, but to a just appreciation of their historical significance. It was a work of new discovery, inasmuch as a great part of the limited region of the cata- combs known to Bosio had been again lost track of. Even the catacombs which were already known had not been correctly identified by their ancient names ; they could not, therefore, be brought into relation with the history and traditions of the Church, and their testimony was almost exclusively limited to the obvious facts which met the eye. There were defects in Bosio's method which, small as they were, quite baffled his pursuit; and for all his patient investigation he found none of the sepulchres of the historically renowned heroes of the faith. On principle he avoided all regions of the catacombs which HISTORY OF THE STUDY 19 showed traces of late masonry and painting : this seemed to him an obvious sign that the region and its tombs must be of late date. De Rossi, on the other hand, recognized that the decoration and construction which belonged to a time when the' catacombs had ceased to be places of burial must indicate the neighborhood of tombs which were venerated from the fourth to the seventh century. It was at that time new stair- ways were constructed and the corridors were enlarged to give easy access to the historic chambers, which themselves "were amplified and decorated as an expression of the popular cult which was accorded the martyrs. In locating and distinguishing the catacombs, in assigning to the different regions their date, and in prescribing the method which must be followed in their study, de Rossi accomplished an indispensable work which perhaps no one else was fitted to do. This, however, was but the ground work ; the most interesting part of his study, the important fruit which he drew from it, consists in the lines of connection which he traced — ingeniously combining many mere hints into a sound argument — between the monuments of the cata- combs and the general history of the Church and of the Em- pire. His peculiar talent for the work is especially shown by the genealogical puzzles which he unravelled, connecting the different catacombs, or their individual tombs, with historical families of Christian or pagan fame. Such arguments, how- ever, are in their very nature so dependent upon minute de- tail that not even a specimen can be presented in this book. And in general it must be said that the soundness of de Eossi's results, the accuracy of his method, and the candor of his judgment, cannot be fairly judged except from his own works in full — not even from the English condensation of his work by Northcote and Brownlow. The disparagement and dis- trust which de Rossi encountered in England and America was due to a merely second-hand acquaintance with his work which gave an inadequate idea of the cogency of his argu- ments. In all that is fundamental to the study of the catacombs his work deserves the highest confidence. The catacombs have not yet been completely explored ; but what remains may be discovered (it is chiefly a question of expense) by the light which he has furnished. The four great volumes 20 INTB OD UC TION of Roma Sotterranea hardly represent the half of his published studies of the subject. The studies and discoveries in the catacombs since the date of that publication are about to be edited by his surviving disciples as a continuation of Roma Sotterranea. As regards the frescos of the catacombs de Rossi's opinion has less value. In the reproduction of the pictures he did not avail himself of the accurate photo-mechanical processes which were at his disposal, and his copyists were not always trust- worthy. He himself lacked the keen appreciation of art which is requisite in a critic, and even in the interpretation of the subjects his judgment was often astray and has been cor- rected by his disciples. This lack in de Rossi's work is com- pletely supplied by Joseph Wilpert, who has already published some works on the subject, and is about to give a complete presentation of the pictures of the catacombs in a work which for the accuracy of its reproductions of the frescos and for the scholarly worth of the studies upon them is sure to be recog- nized as the final work on the subject. Marucchi, another of the immediate disciples of de Rossi, is likewise engaged in carrying on his master's work ; two others whose work was very important, Armellini and Stevenson, have recently died. But de Rossi's influence was not confined to the circle of his personal scholars, nor to the department of study with which his name is specially associated. Even in the study of the catacombs he was obliged to deal with almost every subject which interests the archaeology of the early Christian cen- turies. The Roman monuments which he discovered and illustrated presented for the first time such a sequence as made it possible to understand early Christian art in the order of its development, and they constitute as it were a standard scale for the judgment of the date and value of all kindred objects wherever they may be found. De Rossi's great work on the Christian mosaics of Rome is only a strik- ing example of the breadth of his activity in this field. To the study of almost every branch of Christian archaeology he imparted a new impulse; in some departments he first laid the foundation. None of the scholars who have followed him can fail to recognize their debt. HISTORY OF THE STUDY 21 But fundamental as de Rossi's work was, it was only a symp- tom of a general revival of the study of early Christian monu- ments and art. Already by the end of the last century there was an awakening to the worth of Mediaeval art, especially of architecture ; and the interest gradually extended back to the early period. In Italy and contemporary with de Rossi, Raffaelle Garrucci planned a corpus of the monument of early Christian art and carried it out very creditably. In France, nearer the beginning of the century, Seroux d'Agincourt made a similar publication of Christian monuments, covering less completely a more extended period. Early Christian art found in Raoul Rochette a sharp but mistaken critic who stimulated research by the very temerity of his views. Tow- ard the end of the century again, Rohault de Fleury labored at the reproduction of Christian monuments, illustrating how- ever only selected themes, and following them out through the Middle Ages. His son, Georges de Fleury, is engaged in revis- ing his father's work and in continuing it on the same lines. Edmond le Blant, working under the inspiration of de Rossi, not only furnished new material by the publication of the inscriptions and sarcophagi of Gaul, but suggested new points of view for the appreciation of early Christian art in general. The abbe Duchesne has done valuable work in this field, though he is better known for his liturgical studies. Franz Xaver Kraus, though he has added nothing material to the study, has done very much to popularize it among the Roman Catho- lics of Germany. Among German Protestants the first really monumental study of the early Christian. period begins with Ferdinand Piper, who, quite independent of de Rossi, inaugurated in Berlin an interesting museum of Christian antiquities, and in several large works laid down a programme for the study which has had great influence in Germany and some effect even in England and America. Of the newer school Victor Schultze has been the most prominent leader. He has done a considerable ser- vice in stimulating interest and work in this field. A number of students have devoted themselves to the study, and in gen- eral the students of early Christian litearture have showed appreciation of the importance of archaeology. The predomi- nance of a polemic interest, and the zeal for novel interpreta- 22 INTR OD UCTION tions, has not affected favorably the value of their results. Fieker has been editing a series of studies by younger scholars: monographs which present a detailed account of a limited theme and possess sometimes a certain value. Finally, it is to Germany we owe the most solid as well as the most numerous studies upon early Christian architecture ; Gr. Dehio and H. Holtzinger are especially to be mentioned in this connection. It is natural that the British archaeologists should be chiefly interested in the monuments which are furnished by their own land — belonging to a period and a type quite distinct from the Classical art we are here studying. But how generously the Continental studies have been appreciated and utilized in Eng- land, may be seen conspicuously in the production of Smith and Cheetham's Dictionary of Christian Antiquities, and par- ticularly in the fact that so many competent scholars could be enlisted in the work. "With the widening of the field of discovery Roman scholar- ship is losing something of the preponderance of authority which it has hitherto enjoyed; and British and American ex- plorers are among the first in exploiting the outlying regions which may be expected to yield valuable monuments for our study. For example, the Palestine Exploration has incident- ally proved very valuable also for the discovery of Christian monuments. Strictly in the line of archaeology are the valuable researches of Professor Ramsay, which are published in his Cities and Bishoprics of Phrygia ; and it is to be regretted that his geo- graphical studies have too little relation to art to be noticed in this book. In his St. Paul the Roman Citizen he shows what may be done through archaeology to illuminate even the history of the Apostles — a service which must be of special interest to Protestant theology. He has brought to li°-ht however, a number of inscriptions of which we can take cognizance, for among them is the most interesting of all Christian epitaphs. TI CHRISTIAN CEMETERIES GENERAL DESCRIPTION OF THE ROMAN CATACOMBS The word "catacomb" has a curious history and a very doubtful etymology ; de Rossi takes it to be a hybrid word, half Greek and half Latin, meaning " next the sepulchres " — accubitorium is a late Latin word for tomb. The great number of tombs on the via Appia is supposed to account for this appellation, although they were so frequent along all the prin- cipal roads that one hardly sees why this name should have been used to distinguish one very limited region. But what- ever was the original significance of the name, its further his- tory is sufficiently clear. The word does not appear before the end of the third century, and then not as the name of a cemetery, but as a topographical designation (ad catacumbas) — like many of the popular names for different quarters of the suburbs (ad! ursum pileatum, ad clivum cucumeris, etc.) — for a special region along the via Appia about the second mile from the present walls. In process of time the name became expressly associated with one of the cemeteries along that road, that of S. Sebastian. This cemetery was one of the very few which were known and accessible after the eighth century. With it the pilgrims confused the traditions of the more important cemeteries near the Appia, and in the ninth century the name which was proper to this cemetery began to be employed for all underground burying places indiscriminately. The name which was original to, and distinctive of, the Christian burial places was cemetery (ccemeterium — KOL/xrjTrj- piov). It is sufficiently indicative of the Christian hope, for it signifies a sleeping place. It denotes a whole congeries of tombs rather than a single sepulchre, though it has this latter meaning in the very rare instances of its pagan use. It was 23 24 CEMETERIES equally apt to denote a surface burying place or one beneath the ground. There was no special designation for the whole complex of the underground cemetery ; particular areas, espe- cially the chambers which included the tomb of a martyr, were called cryptce; ordinary chambers, cubicula; the Classical name hypogenm denoted especially an isolated and private sepulchral chamber. The name ambulacrum is given by archaeologists to the corridors, though there is no evidence of its ancient use. Fig. 1. — Plan of part of the catacomb of Domitilla. It is these narrow corridors or galleries, extending often to great length and crossing at every angle, which are the charac- teristic feature of the Roman catacombs, and give the impres- sion not only of boundless extent, but of hopeless intricacy (Fig. 1). In Rome the average width is less than a yard ; the height is almost always ahove that of a man, and not infre- quently very much greater. In the catacomb of S. Gennaro at Naples the harder quality of the tufa permitted a very much greater width (from three to ten yards), and the limestone in NOMENCLATURE 25 which S. Giovanni at Syracuse is excavated permitted corridors and cubkula of a width which has no example in Rome. It is along the walls of these corridors that the tombs were excavated ; the name given them was .wpulcrum or locus, though the archaeologists have adopted the diminutive term loculus. The tomb is merely an oblong niche cut in the wall, just wide and deep enough to receive the body. As the whole wall space was- utilized to the uttermost, the general appearance is that of a series of shelves (Figs. 2, 4, 5). Graves were sometimes made under the floor itself. Odd spaces too short for adults were often utilized for the loculi of children. The body was sometimes embalmed in Oriental fashion ; but in general it was simply clothed in a tunic and winding sheet, and covered with a coating of plaster. This covering of plaster was probably more effective than anything else in preserving the air of the catacombs from contamination, though the locus itself was also tightly sealed with a plate of stone (Fig. 18), or by several tiles (Fig. 21). It was upon this covering the inscription was cut or painted. In case there was no inscription some common article belonging to the deceased or to his family was pressed - into the soft plaster and served to identify the tomb. A more distinguished place of burial was afforded by the arcosolium (Fig. 11), in which the opening was made, not from the side, but from above, and usually under an arch. This style of tomb permitted the burial of several bodies side by side ; such a locus was described as bisomus, trisomus, etc., according as there were places for two, three, or more bodies. The arch and the wall above and below the tomb being plastered, afforded room for painting, and apart from this there is rarely any decoration in the corridors. Here and there, generally in groups, there open out from the corridors chambers of various size and shape (cubicula), which were designed as family burying places (Figs. 2, 5). In general they were intended to be readily accessible ; the attach- ments for the wooden doors are sometimes to be seen. But in one case, at S. Domitilla, the entrance is by a hole at some elevation from the ground, which was permanently closed by a stone, after the likeness of our Lord's tomb. These cham- bers are of very various shape and size : square, rectangular, polygonal. There is a very important round cubiculum, 26 CEMETERIES crowned by the shaft of a luminarium, in the region of S. Sotere in S. Callistus. The ceiling is sometimes flat, sometimes domed. The sepulchres — either the common locus or the arcosolium — are excavated in the walls of the cubiculum in the same way as in the corridors; or, if sarcophagi are used, they are ranged against the walls or accommodated in niches. Other crypts, though used likewise for burial, seem to have been designed expressly for the celebration of Christian wor- Fig. 2. — A cubiculwrn in the catacomb of Callistus, one of the so-called sacrament chapels Third century. ship — particularly, one must suppose, of the eucharist. Not- withstanding their limited size, they are rightly regarded as subterranean chapels intended for a more or less public use. They open generally with a broad entrance upon the corridor, so that a number of worshippers might have been accommodated in the latter. The accommodation was sometimes still further increased by a second chapel opening directly opposite the principal one ; it is supposed that this was occupied by the women. An example of such a double chapel is the crypt of Miltiades in S. Callistus ; it retains traces of the stone bench which' ran around the wall. One of the most ample and com- N OMEN "C LA T URE 27 plete of these subterranean basilicas is a crypt in the Ostrian cemetery which is provided with a presbytery architecturally separated from the nave, ending with an apse, an elevated seat for the bishop, and a lower bench for the presbyters. Here there are also lateral niches for the sacred utensils, and on the other side of the corridor there is a corresponding chamber for the women. The earliest of these chapels (ascribed to the beginning of the second century), and the most interesting in point of decoration, is the so-called Capella greca (Fig. 3) in S. Priscilla. These crypts and chapels were not infrequently adorned architecturally with columns, capitals and cornices worked in the tufa walls or done in marble ; but more com- monly with painted decoration alone. These early chapels, which are marked by great simplicity in the means of decoration, must be distinguished from those which were constructed after the peace of the church, when the crypt of every famous martyr was transformed into a resort for pilgrims and a place of public worship. To this end they were sometimes enlarged (Fig. 5), often adorned with marble and furnished with an episcopal chair, with altar and ciborium, and with everything necessary for the celebration of the eu- charist (Fig. 10). The neighboring corridors were enlarged and strengthened with brickwork, and not infrequently a special stairway was constructed to give ready access to the sacred spot. With the increasing desire for burial near the saints — ad martyres — corridors were cut and loculi accommodated on all sides, so that the primitive aspect of the region was seri- ously altered. This transformation of the historical crypts was due above all to Pope Damascus, who embellished them also with his metrical inscriptions. Burial in the catacombs ceased in the year 410 with the in- vasion of Alaric. This fact, which has been proved by de Rossi, affords an important limit for the dating of both frescos and inscriptions. The frescos which were added in the fifth, sixth and seventh centuries adorn exclusively the tombs of the martyrs, and they stand in sharp contrast with the more primitive work. All of these signs of a late age in the recon- struction and decoration of the martyrs' tombs led Bosio and other early explorers to avoid the very regions which histori- cally and artistically constitute the points of greatest interest, 28 CEMETERIES and often prove to be the centre of the original excavation. The pilgrims have left records of their visits in the graffiti — Fig. 8. — The Cape/la greea in the catacomb of Priscilla. Middle of the second century. names or exclamations scratched upon the plaster — which are often of great importance, and may serve even to identify the crypts in which they are found. PLAN 29 Many of the catacombs are quite simple in plan. Those at Syracuse have ample and regular galleries flanked by large cubicula; they are excavated on a single level, which lies only slightly below the siirface. Hardly more complex in construc- tion is the catacomb of S. Gennaro at Naples. One broad corri- dor runs from end to end of the hill under which the catacomb is constructed; close to it along its whole length runs a nar- rower passage ; and from these there diverge at right angles and at close intervals short galleries or cubicula. The smallest of the Roman catacombs, that of S. Valentine, is even more simple. It is excavated in a hill which is practically the ex- tension of the Pincian along the via Flaminia. The galleries were cut straight back from the steep hillside. The crypt of S. Valentine (martyred in the reign of Claudius II., Gothicus) opened directly upon the road, and behind it there was exca- vated later a large chamber for religious services in honor of the martyrs who were buried on the spot, and a few broad and regular galleries for the accommodation of the tombs of those who would be buried ad sanctos. On a higher level there is a more extensive excavation, but one which is still quite simple in plan : the low, narrow galleries cross one another usually at right angles. One usually speaks of a descent into the cata- combs, and an epitaph reveals that an early name for the stairways was catabaticum (descent) ; but in this case one must ascend along the face of the hill. And the case is by no means unique, for the cemeteries were commonly excavated in a hillside, or at least under a distinct rise of ground. This choice was fixed by the necessity of avoiding dampness and serious inundation from rain water. It was therefore possible to enter the cemeteries by a passage sloping gradually back from the surface. The long steep stairs which descend directly from above belong for the most part to the period after Con- stantine and were constructed to furnish direct access to the crypts of the martyrs. But after all, for the Roman catacombs simplicity of struc- ture is the rare exception. Not only to one who is wandering in their mazes, but even to one who studies their plan, they seem involved in an inextricable confusion. This apparent confusion is, however, explained, if not entirely resolved, when one detects the fact that several cemeterial areas originally 30 CEMETERIES distinct were later united in a single complex. The primary excavation was probably obvious enough in plan ; the irregu- larity of the galleries which meet at various angles was due to the later necessity of utilizing all available space. For the catacombs were not (according to the popular idea) illicit ex- cavations, unknown to the law, and therefore unrestricted in extent. They were distinctly circumscribed by the boundaries of the superficial area allotted to them ; they were not free, either in the earlier or the later period, to extend under neigh- boring property whether public or private. Hence there were express limitations to the agglomeration of separate cemeteries ; the catacombs of Pr.etextatus and Callistus had only the via Appia between them, but they were never joined in any way. No less absolute were the natural restrictions; for the gal- leries could be conveniently constructed only in a special qual- ity of tufa, and in case the catacombs lay under the slopes of different hills it was likely that communications could not be made without descending to a level which would be frequently inundated by water. There are obvious difficulties iu the strict observation underground of the legal limits imposed : one knows little of the Roman methods of mensuration ; but it seems probable from the plans of the catacombs even as we now have them that the first galleries along the level selected were constructed all the way around the area which was at the disposal of the cemetery. The limits were thus plainly marked, and subsequent galleries ran roughly parallel to them. The level was easily maintained by following the stratifications of the tufa. The impression of complexity in the Eoman catacombs is very greatly increased by the fact that they are seldom con- fined to one level; in the catacomb of Callistus there are as many as six different levels of excavation (Fig. 4). This, of course, was prescribed by the necessity of utilizing all the available room within the limits imposed by the superficial area. But to this vertical extension of the excavation there were also natural limits due to the water level and the nature of the soil. The possibility of carrying out such excavations as surprise us in the catacombs is due to the geological char- acter of the Roman campagna. PLAN 31 The great plain which surrounds Rome is almost entirely composed of a volcanic tufa — a mixture of pumice stone and sand. It exhibits different characters according as these ele- ments appear in varioxis proportions ; but in general the differ- ent sorts of tufa are sufficiently distinguished by three classes : the lithoid tufa (peperino), which was employed for the wall of Servius Tullius and is still used for construction ; the gran- ular tufa; and the sand {pozzolana), which gives to Eoman mortar and cement its tenacious character. It was in the Fig. 4. — Section of the catacomb of Callistus. granular tufa the catacombs were excavated ; the lithoid pre- sented insuperable difficulties to the excavation of narrow passages, and in the sand a vertical wall could have been maintained only by a facing of brick. This difference in the character of the soil imposed natural limits to the extension of the catacombs either laterally or vertically; and it explains in part why the excavations on one level were more extensive than on another, for not all qualities even of the granular tufa were equally apt for the purpose. We see from the above how unreasonable was the denial of 32 CEMETERIES the Christian origin of these huge excavations, or the opinion that they are merely Roman sand pits (arenaria) utilized by the Christians ; for of the different sorts of tufa the granular alone had no value as a material, and the Romans had, there- fore, no object in quarrying it. The fact is that the literary sources of history had not prepared us to accept such testi- mony as the catacombs bore to the numerical and also to the material and organized strength of the Church during the age of persecution. It must be confessed that there was a certain suggestion for the sand-pit theory in the fact that crypta are- naria was not an uncommon designation for the underground cemeteries. There was, of course, a certain similarity between the two, but the differences also are very plainly marked. It was a practical requisition that the galleries of the arenaria should be broad enough for several laborers to work abreast, as broad, in fact, as the safe support of the roof would allow : the galleries of the catacombs on the contrary were broad enough only for passage in single file. It is true, however, that arena- ria are frequently found in connection with the cemeteries — whether they were accidentally entered, or constructed on pur- pose to provide room for the worthless material excavated from the catacombs. It is evidently for this latter purpose that the floors of the arenaria were often pierced by wells which descended to the galleries of the catacomb which lay below. For the removal of such material and for the provi- sion of light and air the stairways were generally insufficient, and perpendicular shafts (luminaria) were opened directly from the surface and pierced sometimes through several floors of the catacomb. There is no doubt that in times of acute persecution — as for example in the third century when the catacombs were confiscated and their ordinary entrances watched — the Christians availed themselves of the arenaria for secret access to their cemeteries, and, on occasion, for wor- ship and refuge. Equally at fault is the popular exaggeration of the extent and significance of these Christian excavations. The idea is still current in Italy amongst the people that the catacombs surrounded the city with an uninterrupted network; and that they were connected by underground passages, on the one hand EXAGGERATION OF THEIR EXTENT 33 with the basilicas within the walls, and on the other, with the mountains and the coast. Underlying all such representations is the notion that the whole system of excavation was secret, that it was intended not only for purposes of burial, but as a refuge for a hunted sect, as the ordinary place of meeting and worship and even of abode. That such notions conflict with the facts is evident from what has already been said ; it remains now to observe more in detail the actual ex- Flo. 5. — Crypt of S. Ca'cllla, catacomb of CalliBtus. Third century. tent, the true use, and the legal status of the Christian ceme- teries. The location and distribution of the Koman catacombs was determined in part by natural, in part by historical considera- tions. As to the historical factor, we know only in general that the cemeteries of the first, and even of the second century must have been located within the suburban properties and usually in connection with the family tombs of noble or wealthy converts to the faith. Pagan tombs were built along all the public roads which diverged from Rome ; but they were especially numerous and especially notable along the via 34 CEMETERIES Appia, and it was along this road, or generally in this region, that a number of the greatest of the Christian cemeteries were located. The location of the Christian cemeteries in this region may have been due in part to the popularity of burial along the via Appia ; but it is to be noted also that the Jewish colony was from the first established in the corresponding quarter of the city, and that Jewish catacombs have been found in this same region. A considerable majority of the catacombs were located along the roads which left the city on the southwest and west ; Fig. 6. — A crypt in Cyrene. that is, from the Aventine and Trastevere quarters, which were the poorest and most crowded of the city. Along the Appia were the cemeteries of Balbina, C'allistus, Pretextatus, and ad Catacumban (8. Sebastian) ; on the via Ardeatina, the whole complex which is represented by the cemetery of Domi- tilla, the most extensive of all the catacombs ; on the via Osti- eti.sis, the cemetery of Lucina (the burial place of S. Paul), as well as those of Commodilla, of Thecla, and of Zeuo ; on the via Portiieiisis, of Pontianus and of Pope Felix ; on the via Amelia, the cemeteries of Pancratius, of Processus and Mar- ACTUAL LOCATION AXD EXTENT 35 tinianus, and of Calepodius. The Vatican cemetery, the burial place of S. Peter, was on the via Cornelia, which, though it was directed to the north, was in immediate communication with the Trastevere. One must be careful, however, not to exaggerate the predominance of the cemeteries which were accessible from this side of the city. To the east of the Appia we find on the via Latina three cemeteries which are of but small importance ; on the via Labicana, the important cemetery of Petrus and Marcellinus, and two others. On the slope which faces the Sabine hills we find some of the most important of the catacombs: on the via Tiburtina, that of Cyriaca (S. Lorenzo), and on the opposite side of the road the cemetery of Hippoly tus ; on the via Nomentana, the cemetery of S. Agnes, and the Ostrian cemetery; on the via Salaria nova the cemetery of Priscilla, and three of less importance; and there are three also on the Salaria vetus. On the via Flaminia there is only the third century cemetery of S. Valen- tine. Of these cemeteries that of S. Callistus alone is com- pletely excavated ; the sites of some are not precisely known ; and the cemeteries of the Vatican and of Lucina have been completely destroyed by the great basilicas which have been erected over them in honor of the Apostles. The Roman law forbade burial within the walls of the city. It was the Servian wall which marked the limit during the period when most of the catacombs were founded. The Aure- lian wall was built about a mile beyond it, and the fact that no Christian cemeteries are found within this circuit proves that by the end of the first century the suburbs were too closely inhabited to leave room for burial within the first mile. The outside limit was determined by considerations of con- venience, and as a matter of fact none of the Roman catacombs are found beyond the third mile. Those which lie at a greater distance belonged to the villages of the Campagna. We have already seen in brief what were the natural re- quirements to be sought in choosing the site of a catacomb ; the conditions of a well-drained surface and of a proper quality of tufa were so obvious that Michele Stefano de Rossi, who brought to the aid of his brother's work his own knowledge as geologist and engineer, was often able, by determining where a catacomb ought to be, to discover where it actually was. To 36 CEMETEBIES Michele Stefano de Kossi is also due a careful computation of the total length of the galleries of the catacombs ; it is founded upon a calculation of the average development of the galleries under a given square area of surface, and it gives as result something over 550 miles. So great an extension of the gal- leries was rendered possible by the several levels of excavation which were worked at depths of from eight to twenty-five yards beneath the surface. THE FOSSOES Such vast works as are represented by the Roman catacombs were not executed at haphazard ; they demanded some skilled direction, not only for the selection of sites, but for the constant ex- tension of the excava- tion. And in fact the excavators — fossores — constituted a sort of guild. To them was committed the prepara- tion of the dead for burial, as well as their actual interment ; but to their office there attached none of the ignominy which made contemptible the name of those who performed similar functions in the pagan community — the vespillones. On the contrary, they were proud of their title and inscribed it upon their tombs as a mark of dignity and merit. In the third century the fossors were actually reckoned among the clergy as the lowest grade ; and in the fourth it appears (from several inscriptions which speak of tombs bought from fossors) that they had the general control of the cemeteries. On the tombstones are often represented the tools of the fossor, and in the frescos, the fossor at work (Fig. 7). How far the fos- sors of the several cemeteries were united by their craft we have no clear indication. It is to be noted, however, that the Fig. 7. — Arcosolium of the fossor Diogenes, a fresco in the catacomb of Callistus. Third century. THE CATAC03IBS AND THE TITLES 37 different catacombs, for all their similarity, are marked by certain differences which indicate not only a separate history, but an independent tradition in the style of decoration. THE CATACOMBS AND THE TITLES Besides the fossors, there must have been a higher authority charged with the general superintendence of the catacombs. During the earliest period, while the cemeteries were still private property and represented an extension of the family tomb to include those who were of the same family in the faith, we can only represent to ourselves that they were held at the disposition of the noble owners. But when, with the beginning of the third century, they became virtually, if not legally, the corporate property of the Church, we cannot but think of them in some strict relation to the regular ecclesiasti- cal organization. And, in fact, Hippolytus, in his Philosopliu- mena, reports that about the year 197 the Roman Bishop Zephyrinus appointed his deacon Callistus to the charge of the "cemetery." This is evidently the cemetery which still bears the name of Callistus, though he himself was not buried there ; and it is not difficult to see why this cemetery in par- ticular should depend directly from the bishop, for it had become — after the cemetery of the Vatican ceased to be used for this purpose — the ordinary place of burial for the heads of the Roman Church. We have evidence for the continuance of this relation about a century later in an inscription which recounts that a deacon of Marcellinus, with permission of that pope, prepared for himself in the cemetery of Callistus a double cubiculum with arcosolia and lucernariam. On the other hand, in the neighboring cemetery of Domitilla there was found the following inscription : ALEXIVS ET CAPRIOLA FECERVNT SE VIVI IVSSV ARCHELAI ET DVLCITI PRESBB; that is to say, a certain husband and wife con- structed this tomb in their lifetime with permission of the Presbyters Archelaus and Dulcitus, who evidently exercised the chief authority over this cemetery. They were probably presbyters of the Titulus Fascioke, the parish church within the walls, with which a number of epitaphs distinctly connect this cemetery, the same church which has borne the name of 38 CEMETERIES SS. Nereus and Achilleus, since in the ninth century the bodies of these saints were removed thither from the cemetery of Domitilla. De Rossi has been able from inscriptions to establish the connection between several of the cemeteries and particular titles of the city, and the suggestion is obvious enough that each parish church had its corresponding cemetery. This arrangement may very well have been observed from the beginning. Some of the titles are as old as the earliest use of a private house for Christian worship. The Domus Ecclesice took its special title from the name of the owner : — Domus or Titulus Pudentis (Ecclesia Pudentiana as it was called later), Praxedis, Equitii, etc. While the church was accommodated in the house of a wealthy benefactor, the cemeteries may have been excavated under his suburban villa, and the community which worshipped in that title would naturally be the one to use the cemetery associated with it. In the Liber Pontificalis it is recorded of Fabianus (shortly after Callistus) : Hie regiones divisit diaconibus et midtas fabri- cas per coemeteria fieri iussit. Of the cemeterial buildings we shall have to speak later ; as to the deacons, it is well known that the Roman Church maintained the number seven, in imi- tation of the " seven " who were originally appointed in Jerusa- lem. By reason of the fewness of their number and their close relation to the bishop they enjoyed an authority which exceeded that of the presbyters, and the successor to the epis- copate was usually taken from their number. The civil regions of Rome were fourteen ; to each of the seven diaconal estab- lishments, which were designed to administer to the temporal wants of the Church, there must have been allotted two of the civil regions. The titles of the city were twenty-five in num- ber in the third century — roughly two of them to every civil region.' That they were charged not only with spiritual care of the people and with the more obvious parochial duties, but with the administration of the cemeteries, we can infer from a notice in the Liber Pontificalis concerning Pope Marcellus, at the very end of the period of persecution : XXV titidos in urbe Romana const Unit quasi dioceses, propter baptismum et poiniten- tiam muUorum qui convertebantitr ex paganis, et propter sepul- turas martyrum. " He instituted in the city of Rome THE CATACOMBS AND THE TITLES 39 twenty-five titles as parishes, for the baptism and penitence of the many who were converted from the pagans, and for the burial of the martyrs." There are reckoned thirty-two public cemeteries for Rome, — a number which precisely corresponds with the sum of the twenty-five titles and the seven deacon- ries. It is not necessary to suppose a strict uniformity in the distribution of the cemeteries, in such wise that the smallest congregations had each one and the largest had no more ; but there is sufficient proof of a general distribution amongst the titles. Dionysius reorganized the cemeterial administration (according to the Liber Pontificalia) after Gallienus had revoked the edict of his father which had prohibited meetings in the cemeteries: Hie presbyteris ecclesias divisit et ccemeteria et paro- chias et dioceses constituit. We learn that there were commonly two priests attached to each title. Of these, one alone was the proper titular, and the other his subordinate; — as we learn from a letter of S. Cyprian : Felix qui presbyterium submit dstrabat sub Decimo. The same system was maintained in Rome in the time of Pope Damasus (as we learn from a passage by S. Ambrose, on 1 Tim. 3) : Nunc autem septem diaconos esse oportet et aliquantos presbyteros ut bini sint per ecclesias — that there be two priests for each church. This dual number of priests for each parish is explained by the relation of the titular basilica to the basilica which was connected with the cemetery. The number of Roman churches early in the third century is indicated in a passage of Optatus Milevitanus, in which he says that among the forty or more basilicas the Donatists had no place where they might meet. According to the testimony of Eusebius the Roman Church had in the time of Cornelius forty-six priests — a number which very well corresponds with the " forty or more basilicas." Besides the twenty-five titles there were, therefore, something like twenty-one cemeterial basilicas which were in the charge of the parochial clergy, who took their titles, however, not from the cemeteries, but from the basilicas within the city. So much for the administration of the cata- combs; the question of their legal tenure (a matter of even greater interest) must be treated at large in another place. The light which the study of the catacombs throws upon the administration of the Roman Church has more than a merely 40 CEMETERIES local interest, for from the first it was the Church in the Capi- tol of the Empire which constituted the norm of ecclesiastical government. THE CHRISTIAN MODE OP BURIAL There is one question which poses itself almost at the very mention of the catacombs : Why did the Christians resort to a mode of burial so different from the prevalent custom of the Graeco-Roman world ? There is certainly a problem here which demands some explanation ; but before all, one's first idea of the singularity of the Christian mode of burial must suffer considerable reduction: all the catacombs are not Christian, nor were all the Christian cemeteries subterranean. Even in Rome the Vatican cemetery, which must have been used by the Church since the burial of S. Peter, was chiefly on the sur- face ; the nature of the rock did not admit of deep excavation, the bodies were deposited in trenches after the same manner as, is customary to-day, or the sarcophagi were left exposed above the surface of the soil. Throughout the Roman cemeteries, burial on the surface rapidly overtook during the fourth century, and by the beginning of the fifth it had quite super- seded, burial in the catacombs. In the East generally, where inhumation was commonly practised, the Christian converts found in their creed no reason to modify essentially the mode of interment or the type of cemetery which was traditional to their particular race or nation. In Egypt during the Roman rule the mode of burial for pagan and Christian alike was of the simplest possible : the body was merely buried in the dry sand, dressed in common garb, and without sarcophagus, coffin, or shroud. In Syria, according to national tradition, the character of the tombs was very various. The commonest was a plain sarcophagus, all of it buried except its rooflike lid, or a similar lid covering a rock-hewn trench. But there were also chamber tombs hewn in the rock and faced with a Greek porch or other architectural ornament. The most characteristic Syrian type of sepulchre was the tegurinm, a pyramidal roof supported above a sarcopha- gus by four pillars. In North Africa Christian burial was always, so far as we know, in the surface soil. Surface ceme- teries were not uncommon in Italy ; the most extensive remains CHRISTIAN MODE OF BURIAL 41 are at Portogruaro (Julia Concordia) ; curious is the arrange- ment of the great stone sarcophagi in groups of ten and twelve. In the Northern countries the open air cemeteries greatly pre- dominated, — in France, in the Rhine country, in Dalmatia, and in Istria. It would be going too far afield to describe the various forms which the Christian tomb exhibits in different lands. Such an account would almost constitute a treatise upon ancient modes of sepulture, for the Christians generally continued without scruple the customs of burial to which they were used as gentiles. Fig. 8. — Sarcophagus in the catacomb of Priscilla. Second century. On the other hand, catacombs are not confined to Rome and Italy ; they are found in the Rhine country, e.g. at Cologne and in other lands. Subterranean burial was probably regarded as the most specifically Christian mode of interment. It is particularly in the centres of Graeco-Roman civilization, where cremation prevailed, that the Christian mode of burial seems in sharpest contrast with gentile custom. There is no doubt that the practice of inhumation was prescribed by the new faith and obligatory upon all the faithful ; with a naive conception of the doctrine of the Resurrection cremation seemed irreconcilable. There was also a strong repulsion to burial amongst heathen. It was therefore in the nature of the case that Christians must be buried, and that they must be buried together. As for the matter of burial, it was by no means unknown even among the Romans. In the Laws of the Twelve Tables burial and burning were both contemplated, and burial was mentioned first. Some of the great Roman families — like the Scipios — maintained the early tradition, burying their dead in sarcophagi in the family crypt surmounted by a monument. 42 CEMETERIES In the third century the custom of inhumation rapidly spread — whether through the influence of Christian example, or through the popular adoption of Oriental cults and fashions. From this period date the richly carved pagan sarcophagi. Rome therefore pre- served in some measure the traditions of underground burial which were de- rived immediately from the Etruscans. The Etruscan tombs, with their numer- ous chambers and their sculptured and painted decoration, present a most strik- ing analogy to the Christian catacombs. But these tombs were derived in turn from the Phoenician custom, which was disseminated widely — and no doubt especially in Judea — by this active merchant people. The nearest analogy to the Christian catacombs is furnished by the under- ground cemeteries of the Jews, which have been discovered at Rome. One of these on the via Appia is accessible to the public and may conveniently be visited along with the catacombs of Callistus and Domitilla. One will notice certain differences of construction, such as distinguish even the different Chris- tian catacombs from one another ; but there is the same arrangement of gal- leries and cubiada, the same method of accommodating the tombs, the same decoration in paint and plaster ; and in general, except for the presence of Jewish formulas and symbols and for the absence of Christian, one would see no reason to distinguish them from Christian cemeteries of the Roman type. They can none of them claim so ancient a foundation as the Christian ; but they suggest nevertheless, and plainly enough, that the origin of the Christian mode of burial is to be sought in Jewish custom. It Fig. 9. — Lead coffin from Sa\da. Phoenicia. Fourth century. \Y OB SHIP IN THE CATACOMBS 43 is not unreasonable to credit in particular a considerable influ- ence to the memory of the rock-hewn tomb in which the body of our Lord was laid. We must recognize that the Christian catacombs when com- pared with pagan monuments of similar character present no essential difference of type : it is only by reason of their great extent that they strike us as something altogether strange and unparalleled. The original Christian hypogeum (the nucleus of all the catacombs) consisted in a group of crypts connected by short galleries, and its likeness to Etruscan and Oriental tombs is so obvious that the question of origin need hardly be raised. The way of subterranean burial having once been struck out, the subsequent complexity and extent followed as a matter of course upon the rapid increase of the Church. De Rossi has observed that no essential distinction can be drawn between the Christian catacombs on the one hand, and Phoenician, Etruscan, and Jewish tombs on the other ; except that in the former the tombs were always sealed so as to permit visits to the cemetery, while in the latter they were always open. WORSHIP IN THE CATACOMBS This single peculiarity of the Christian cemetery is to be referred to the custom of holding religious services within the crypts, not only at the time of the deposition of the body, but on the anniversary of the death or burial. S. Augustine says expressly that the eucharistic sacrifice was offered at the tomb, and he mentions it in particular at the deposition of his mother. Among the crypts of the catacombs there are many which cannot but be recognized as veritable underground chapels, which were evidently constructed expressly for wor- ship, even though they were incidentally utilized for burial, or primarily represented the tomb of a saint. Mention has already been made of a chapel of the third cen- tury in the Ostrian cemetery which presents substantially the scheme of the church edifice as it appeared after the Peace of the Church. There we see the apse and the apsidal arch, with the episcopal chair in the centre and the presbyters' bench on each side. We have to suppose that a wooden altar was placed before the chair. The presbytery is distinctly sepa- 44 CEMETEBIES rated from the nave by the pillars which support the arch. The chair carved in tufa is a peculiarity of the ostrianum, and it is repeated there very frequently. In case there is no apse, or the apse is occupied by a tomb of the arcosolium type, the chair is found beside the tomb (the stone covering of which probably served for the holy table), or it is located at the middle of the side wall. The arrangement of these chapels seems to indicate that the custom of celebrating the Eucharist was already substantially developed as it is at present; that the presbyter alone stood (or perhaps sat) at the holy table, and the communicants approached it only at the moment of communion. In some of the earlier chapels, however, we seem to have a witness to more primitive custom. For example, the so-called capella greca (Fig. 3), which belongs to the beginning of the second century, though it terminates in an apse and has two deep lateral apses in the form of a transept, has no special seat for the priest or bishop. One will hardly be inclined to lay any stress upon the cross shape which the chapel actually presents, for the apses are nothing more than broad tombs ad arcosolium, and there is no distinct place indicated for altar, presbytery, or nave. Besides the rich decoration there is only one thing which distinctly marks it as a place for public worship ; that is, the tufa bench which runs along one side and is practically continued on another by the surface of the tomb under the central apse. One must imagine a wooden table carried in upon occasion, and one cannot but see in this whole disposition an arrangement for the seating of the com- municants at the eucharistic table according to the earliest Christian custom at the breaking of bread. And, in fact, on the arch of the central apse, Wilpert has lately discovered, by removing the calcarious deposit which covered the wall, an eucharistic scene (Fig. 74) which may very well represent what the artist himself had witnessed in this chapel (see p. 227). When the custom which we see portrayed in this picture had to be abandoned on account of the multitude of communicants, and the congregation could no longer sit about a common table, it is natural to suppose that the priest retained his old position, seated beside the tomb (as is indicated by the chairs of the WORSHIP IX THE CATACOMBS 45 Fig. 10. — Papal crypt in the catacomb of Callistus. Eestored as it was in the fourth century. 46 CEMETERIES ostrianum),a.nd not standing, according to the later fashion, with his back to the people. It has been commonly claimed, and too readily allowed, that all the tombs ad arcosolium were expressly constructed, or at least commonly used, for the cele- bration of the Eucharist. In olden time they were supposed to be the mark of a martyr's grave. In view of the modern practice which has reduced the altar of the church to a mere shelf against the wall, this supposition seems natural enough. But in an age when the primitive conception of the altar as the common table was still preserved, when the very position of the altar between the priest and the people still expressed this idea, the eucharistic use of these so-called table tombs, in the manner which is commonly supposed, is very far from obvious. Furthermore, the frequent and practically private Eucharists which this view implies are hardly to be recon- ciled with the jealousy of the Koman bishop for his preroga- tive as the proper dispenser of the Eucharist — a jealousy which went so far that the consecration of the Eucharist was not allowed even in the churches, without the presence of the express mandate of the bishop, and the congregations were obliged to wait for the consecrated elements which were car- ried to them by the deacons from the episcopal altar. There were doubtless private services in the catacombs, — in the family cubicula, and at the burial of the humblest disciple, — but it is to be supposed that the services in the chapels, and more particularly at the graves of the martyrs, had distinctly a public character, and was representative of the whole Church, even if but few worshippers could convene. The episcopal chair (for example in such a chapel as that we have described in the Ostrian cemetery) clearly implies the bishop's presence. The catacombs were therefore used, and in a measure expressly constructed, for public as well as for private wor- ship. But the extent of their public use has been singularly exaggerated, in clear contradiction to the witness of the monu- ments themselves. It is popularly supposed that they were a regular resort for the assembly of the Church during the centuries of persecution. But as a matter of fact few of the chapels could accommodate as many as fifty persons, and even if we take account of the neighboring corridors and cubicula the THE AREA AND ITS BUILDINGS 47 number of worshippers could hardly exceed one hundred, while the catacombs themselves prove that the membership of the Roman Church in the third century was counted by the tens of thousands. We shall see in another place that the pictured decoration of the catacombs referred predominantly, if not exclusively, to the themes which were associated with death or rather with the Christian hope which illuminated it. And it must be recognized that the religious services which were ordinarily held in the subterranean chapels were expressly associated with the funeral offices and with the memory of the departed. One can, indeed, readily imagine that the catacombs proved often a safe and welcome resort in seasons of severe persecution when public worship may have been impossible elsewhere, and there is good reason to credit the accounts of such meetings, as they are reported in the lives of the saints, though this use was doubtless exceptional. CONSTRUCTIONS ABOVE GROUND The notion that the Church adopted subterranean burial to protect the tombs from the desecration they must have suf- fered had they been publicly known, or to hide from the state the names or numbers of its members, rests upon a complete ignorance of the sacred protection which was accorded by the Roman law to every sepulchre, and it stands opposed to the fact that the catacombs were, in fact, not secret. The confisca- tion of the Christian cemeteries under Valerian (258) and Dio- cletian (303), and their subsequent restitution under Gallienus and Maxentius, as well as the edict of Milan (313), which restored to the Christians (" ad jus corporis eorum ") the churches and cemeteries which belonged to them, proves that the catacombs were not only known to the authorities, but also clearly recognized as the corporate property of the Church. Nor was there, indeed, any attempt to hide the catacombs ; the original entrances (a number of which have been rediscovered) faced the public roads, and each faqade doubtless bore an inscription indicating the extent of the cemetery and the name of the owner of the property. The area must have been marked by distinct boundaries and devoted to the cemeterial structures which Christian as well 48 CEMETEBIES as pagan custom required as adjuncts to the tomb. There is nothing to inform us by what term the Christians in Rome dis- tinguished the superficial area from their subterranean excava- tions ; the word "cemetery " referred equally to both. It is this broad significance of the word which has given rise to the notion that the Christians sometimes actually retired to live in the catacombs. The Liber Pontificalis, in the life of Libe- rius, says : eodem tempore (Constantius) misit et revocavit Libe- rium de cmmeterio beatw Agnes ubi sedebat; and in the life of FiA. CABEINOC KAI TITIANH AAEASOI. That is, "Flavius Sabinus and Titiana, brother and sister." All of this renders plausible the form in which de Rossi completes a mere fragment which appears to have belonged to the inscription placed over the entrance of the cemetery : — Sepulc EVM Flavi R V M At all events, there is no doubt that as early as the first cen- tury this was the burial-place of the Christian members of the imperial Flavian house. These mere names suffice to connect this cemetery with the illustrious converts of the gens Flavia whom the Church could already count within the Apostolic age. It has been suspected, from the language in which Tacitus describes him (mitem virum abhorrentem a sanguine et ai'diMis), that Titus Flavius Sabinus, elder brother of the Emperor T. Flavius Vespasianus, was the first of the family to be converted to Christianity. He was for the first time Praetor in 64 under Nero, and it is certain that as a duty of 66 CEMETERIES his office lie must have examined into the causes of the Chris- tians who were executed for their religion. During the thirty- years of absolute peace and tranquillity which the Church enjoyed after the death of Nero there is no mention of Chris- tians of this name. The relation of the family to Christianity becomes first publicly known by reason of the persecution of Domitian, and it is attested by pagan as well as by Christian historians. The first to fall a victim was the Consul Titus Flavius Clemens, son of the above-mentioned T. Flavius Sabi- nus and first cousin of the Emperor. "While Clemens was beheaded, his wife, Flavia Domitilla, niece of Domitian, and another Flavia Domitilla, who was a niece of Clemens, were exiled to the islands of Pandataria and Ponza. In expla- nation of these harsh measures, it must be supposed that Domitian considered the profession of this strange religion by members of his own family a proof of political disaffection. It suggests food for the imagination to reflect that but for this outbreak of suspicion a Christian emperor might have occupied the throne of the Caesars before the end of the first century ; for it was the two sons of Clemens and Domitilla whom Domi- tian had adopted as his successors, changing their names to Vespasianus and Domitianus. The memory of the Flavian converts and martyrs has been preserved in the Church and hardly needed the confirmation of the monuments. But another illustrious convert and martyr of the first century is known as such only through inscriptions discovered in the cemetery of Priscilla. Manius Acilius Gla- brio, Consul in 91 with Trajan and head of one of the noblest Roman families, was also put to death by Domitian. He was made to fight with a bear or a lion, and, proving victorious in this contest, was beheaded. Though no memory was preserved in the Church that he died a Christian, yet the terms in which Suetonius records the charge which was brought against him and other members of consular and senatorial rank who suffered with him (molitores rerum uorarum) has led several historians to suspect that they were martyrs for the Faith. That the Acilii Glabriones were Christians was put beyond a doubt in 1889 when, in the central and primitive region of the cemetery of Priscilla, there was discovered an extensive and richly orna- mented hypogeum which contained fifteen inscriptions in Latin INSCXIPTIO.VS 67 and Greek of members of this family. Originally there must have been more, for the epitaph of the Consul himself is miss- ing ; the very richness of the marble decoration specially marked this crypt for destruction, and only fragments of the sarcophagi and their inscriptions remain. One of them reads : — aKIAIOC POY*INOC £HCHC EN ©EO Acilius Bufinus live in God — a sure sign of the Christian character of the sepulchre. Another reads: — M ACILIUS V . • C • V • et PRISCILLA C • • Manius Acilius vir clarissimus (et) Priscilla darissima (femina). The title clarissimus vir leaves no doubt that this personage of senatorial rank belonged to the family of the consul who was put to death under Domitian. The name Pris- cilla suggests a relationship with the family of the senator Pudens from whose wife Priscilla the cemetery took its name. In this cemetery were likewise buried that Aquila and Pris- cilla (Prisca) who were companions of S. Paul, and the site of whose house upon the Aventine is marked by the church of S. Prisca (contraction for Priscilla). Their common use of the name Priscilla, together with the fact that both families were buried in the same cemetery, suggests some close tie between the family of the tent-maker upon the Aventine and the sena- torial family of the Esquiline. There is something to be learned from the very brevity of the early inscriptions ; there is argument to be drawn from their silence. During the first four centuries of the Church no single mention is made of a slave, and but rarely of a freedman, among the thousands of inscriptions of the catacombs — justifying the Christian boast that master and slave recognized their equality in the Church. In a later time the inscriptions occasionally record the manumission of slaves in suffrage of the departed. In contrast to the pagan custom, even the noblest of the Christians recounted none of the honors of their offices and 68 CEMETERIES rank, except that the initials V. C. (yir darissimns), C. F. (darissima femina), were not uncommonly inscribed to indicate membership in the senatorial order. The Christian attitude was that of looking forward beyond the tomb, rather than back over the course of earthly honor and success ; recessit a sceculo became a familiar formula in the fourth century. In the third and fourth centuries the profession of the defunct was often mentioned in the inscription or indicated by pic- turing the tools of his trade. We have in general in the catacombs a thorough vindication of Tertullian's boast 1 that Tombstone in the catacomb of DomitiUa. the Christians were to be found in every rank and in every profession. Nothing could be more simple than the epitaphs of the Roman bishops in the papal crypt at S. Callistus. The earliest which have been preserved in this crypt are those of Anteros (236) and Fabianus (250) : — ANTEPOC EIII (Anteros, bishop). 3ABIANOC • EIII • MP (Fabianus, bishop, martyr). The inscriptions of this crypt prove that Creek was still the official language of the Roman Church. The next pope, ( 'oinelius, was buried in a distant region of the same cemetery, the very region, in fact which seems in origin to have been the property of the Cornelii and the 1 Apol. 37. ixsenirrioxs on Cavilii. This probably explains tlio fact that the epitaph of this pope is not in the official language of the Church, but in Latin; — I'OHNELIVS- MARTYR Kl' The word "martyr " here is original ; on the epitaph of Fabianus, however, it was a subsequent addition. Most of the very early inscriptions in the Roman catacombs were in Greek, and the same language persisted here and there to a comparatively late period. (J reek inscriptions were sonic- times written in Latin characters, and Latin sometimes in Greek. The very general traits of Christian epigraphy which can here be noticed serve as well for the Greek as for the Latin, for the East as for the "West. It seems not unlikely, however, that early inscriptions in the Orient may have been more elaborate than those of the same period which we know in Rome. The earliest inscriptions of the East have not been pre- served ; but the epitaph of Abercius (Eig. '20), which is trans- lated in another place Q>. ~'X>), belongs to the middle of the second century, and is one of the most elaborate and interest- ing of all Christian inscriptions. It is inscribed upon a stele of the common pagan form. The use of the stele or eippits was not altogether rare in the Church, although the vast majority of inscriptions are upon plaques of stone. Lespite their pagan significance, the initials 1 > • M • (Vis manibirs) are sometimes found upon Christian tombs; partly because the plaques were thus inscribed as they were bought at the shops, and partly, perhaps, because they were so much the ordinary sign of a tomb that their more specific significance was forgot, B. M. (Imnw nwmoritv) was sometimes substituted in a later age. In point of orthography de Kossi distinguishes two classes of the primitive Roman inscriptions: those painted in red (in l'ompeian fashion), which are characteristic of S. l'riscilla ; and those cut in the stone, which are elsewhere almost univer- sal, 'the orthography is for the most part careless, and after the second century there begin to appear frequent mistakes which reflect the popular pronunciation and the popular idiom. Even in the concise terms of the early epitaphs there some- times lies a clear testimony to early dogma. In the third cen- 7(1 CEMETERIES mvmmutt i .-p.. X tury a greater fulness and variety appears. There are a number of prayers, particularly in Greek, which suggest a liturgical origin. Metrical inscriptions are ran' until the fourth cen- tury; the earliest examples of them are commonly brief, and show a dependence upon, if not an actual quotation from, the Classical poets. But there are also inscriptions in quasi versus, a variety of verso invented by Com- ~— - v niodiaii, a Christian poet of the third -\ century. It is not of much interest \ to record that the inscriptions, ; ^k eal 'ly an, l l Ichval kI cpi Ihi ixscRirrioxs 7i martyrs were regarded as advocates before God for the souls of the departed. For this period, however, the popularity of the view is proved, not only by inscriptions, but by some of the paintings of the catacombs which represent the soul introduced into heaven by the saints, and the same theme appears later in the mosaics of the basilicas. It is in this cult of the martyrs we find the roots of the later doctrine of the saints ; in the offi- cial recognition of martyrdom, and in the special efficacy which was attributed to the martyr's intercession, we have the essen- tial factors of the mediteval doctrine. It was this conception of the martyrs as advocates in the Judgment which made Fig. 21. — Loculus of a martyr, closed with tiles, catacomb of Domitilla. burial near them seem so desirable. The following inscrip- tions are of the fourth century : — CVIQVE PRO VITAE SVAE TESTIMONIO SANCTI MARTYRES APVD DEVM ET CRISTVM ERVNT ADVOCATI (Cemetery of Cyriaoa.) DOMINA BASILLA COM MANDAMVS TIBI CRES CENTINVS ET MICINA FILIA NOSTRA CRESCEN . . . QVE VIXIT MENS X . ET DES. . (Cemetery of Basilla.) Domina (dominus) was the title given to martyrs. The latter inscription reads : " lady Basilla, we commit to thee Crescen- tinus, and our tiny daughter Crescen(tia) who lived 10 months and . . . days." 72 CEMETERIES Another, from Aquileia, MARTYR ES SANTTI IN ■ MENTE IIAVITE JUARIA reads, " Holy martyrs, remember Mary." But to return to the earlier forms which regard all the faith- ful departed without distinction : I have spoken of them as a sympathetic subject of study, because they are so human, so naive, and spring so promptly from the heart. The prayer for a place of refreshment, of light and peace, of rest in God, in behalf of the departed soul, was impossible from the stand- point of the pagan, simply because the other world was not conceived in such terms. To the Christian, on the other hand, these were the ideas which were naturally associated with the death of the believer ; and if there was nothing in the Chris- tian teaching which positively required such prayers, there could at least be no more solid objection brought against them than the claim that they were superfluous. "What more natu- ral, however, than that the Christian hope for the dead should at the very tomb itself be expressed as a prayer ? What more natural than that such prayers should appear upon the tomb- stones before ever they were formulated in the liturgies, and before the doctrine of a purgatory of pain had turned their glad confidence into a tearful and doubtful supplication ? The simple exclamations we have here to record bear evidence of being the fruit, not of any clear doctrinal conception, but of a popular and natural fantasy. • ■ vIBAS IN PACE ET PETE PRO NOBIS "Live in peace! and pray for us," reads an ancient inscrip- tion in S. Domitilla. The following', of the fourth century, gives the theological ground which justifies such a prayer to the dead, " l'ray for us because we know that thou art in Christ": — ixsriurriONS 73 GENTIAXVS FIDELIS IX PACE QVI VIX IT ANN IS XXI MENSS \1U DIES XVI ET IX ORATIOXES TVIS ROGES PRO XOBIS QVIA SCI All :s TE IN CHRISTUM i (Lat. mas. VIII. 15.) That prayer for the dead was not associated with harrowing doubt about their fate we see, for example, in an early Greek inscription in S. Domitilla, which -at the same time demands the prayer of the departed in behalf of the surviving friends : — ZHCAIC ■ EN • KO • KAI . EPOTA • YIIEP ■ IIMQN. " Mayest thou live in the Lord ! and pray for us." This is simply the realization of the communion of saints. Of the third or fourth century is the following : — ANATOLIVS FILTO BENEAIERENTI FECIT QVI VIXIT ANNIS VII MEXSIS VII DIE BVS XX ISPIRITVS TVVS JSEXE REQVIES CAT IX DEO PET AS PRO SORORE TVA (Lat. mus. VIII. 10.) " Thy spirit rest in God : pray for thy sister." The demand for prayer in behalf of one's own soul seems to manifest a too anxious solicitude about one's fate; but it is found as early as the end of the second century in the epitaph of the Phrygian bishop, Abercius, written by himself (page 234). The following metrical inscription from S. Priscilla belongs probably to the fourth century : — EVCIIARIS • EST • MATER • PIVS • ET • PATER • EST ■ • • VOS • PRECOR • O FRATRES • ORARE • HVC • QVANDO • VENItis ET • PRECIBVS • TOTIS PATREA1 • NATVMQVE • ROGATIS SIT • VESTRAE • MEXTIS ■ AGAPES ■ CARAE • MEMINISSE VT • DEVS • OMNIPOTENT • AGAPEN . IN • SAECVLA . SERVET 1 The name of Christ is represented by the monogram. 74 CEMETERIES There is unfortunately but little space left to treat of the inscriptions with which Damasus adorned the tombs of the martyrs. They deserve more attention than can here be given them. They are interesting, not only as a type of Christian poetry which was admired by contemporaries and frequently copied in succeeding centuries, and because of the beautiful and characteristic letters in which they were cut; but for the fact that they reveal several pages of the history of the martyrs which but for them would lie absolutely unknown, that they testify clearly to the character of the cult which was rendered to the martyrs in the fourth century, and that they make it possible to identify in each cemetery the position of the most venerated tombs. There was no cemetery at Rome which had not at least one such inscription, and still others were placed in the cemeterial basilicas and chapels. Most of the original inscriptions have totally perished, many of them at the hands of the Goths; but the text of about forty of them has been preserved through the copies made by the pilgrims. In consequence of this lucky preser- vation a mere fragment of the original marble suffices for the restoration of the whole inscription and serves often to hx its original location. The accompanying illustration (Fig. 23) shows a marble which S. Damasus placed at the end of the crypt of the popes (Fig. 10); it is now restored to its original place and is almost complete, although recomposed from one hundred and twenty- five minute fragments into which it was broken. 1 give here the translation, wine], must suffice for an example of Damasus's poems: " \\cn\ if you inquire, lies crowded together a throng of the righteous, the venerable tombs hold the bodies of the THE POEMS OF DAMASUS 7a saints, their loft} spirits the palace of heaven took to itself. Here (he companions of Sixtus who bore trophies from the enemy ; here a number of the leaders who ministered at the altars of Christ; here is placed the priest who Lived in long peace; here the holy confessors whom Greece sent; here young men and hoys, old men and their pure descendants, who chose to keep their virgin modesty. Here, J confess, I Pamasus wished to deposit my body, hut- 1 feared to disturb the holy ashes of the righteous." It is not unreasonable to suppose that a "throng'' of martyrs were often buried in a single tomb, particularly such as suffered together in (he same J \ T 3 ARy ^keqa&$rE$l^ysfj P( ) tot a jhTQ v iibai o sxTjKUvea i *£lfjrCT(^TysLOw^ ] An loscriptlo Hi., irnpnl cr yI persecution. In the ease of such as were burned or thrown to the beasts, often only very small portions of their bodies could be recovered. Sixtus II. and his companions in martyr- dom are here mentioned, although Dama'sus set up m this same crypt a special inscription in his honor. Those "who ministered at the altars of Christ" are probably unnamed deacons and presbyters; and the " sacerdos " of the next verse may refer to the Roman bishops who were buried here (using the singular for the class), though de Rossi understands Mil- tiades, who was the first pope to enjoy the peace given by Constantino. The "confessors" from Greece are unknown, but they may have been Hippolytus and his companions. CEMETERIES The last lines seem intended as a rebuke to those who dis- turbed the bodies of the martyrs in their zeal to be buried near them. Damasus was in fact buried in a little basilica connected with the cemetery of Domitilla, in which he prepared also the tombs of his mother and sister. For himself ami for them he composed inscriptions. This chapel has not yet been dis- covered, but a small fragment of an inscription found near the church of SS. Cosma e Damiano was recognized by de llossi as belonging to Daniasus's inscription to his sister, the text of which was known. This piece was again lost, and has been rediscovered in Ihe course of the excavation of the Forum, ft awakens surprise that it is not inscribed in the customary Damasian letters ; but this is explained by the fact that the sister died before Damasus became pope and before he had adopted the type of letter which is associated with his name. De Rossi has traced the author of these beautiful letters, which though frequently imitated in a later age were never precisely copied. On the marble which contains the inscrip- tion to S. Eusebius, discovered in S. Callistus, there is at each end a line of smaller letters which read from top to bottom : Damasis Pajipa?. cultor atyne amatot Furius Dionysius Filocalus scribsit — ''Furius Dionysius Filocalus the reverer and lover of AFTER THE FOURTH CENTURY 77 Pope Damasus wrote it." This famous personage was the secretary of Damasus. In this inscription one is struck not only by the false spelling, but by the character of the letters, which in fact are only a distant imitation of the Damasian. This is explained by the fact that the original inscription had been broken, and was restored again about the end of the sixth century, perhaps by Pope Vigilius ; it was then cut on the back of an inscription of Caracalla. The interesting inscription which has been translated above is enough to show that Damasus was not a great poet; his verses are not always regular, and he shows a lack of invention in his frequent repetition of favorite words and phrases, many of them taken from Virgil. But his style was accounted ele- gant by Jerome (elegans in versibus scribendis), and he seems to have been a conscientious historian. The historical researches which he must have made about the martyrs were doubtless facilitated by the fact that he was archivist of the Roman church before he was made pope. In the composition of metrical inscriptions Damasus had imitators even among the popes. Many such inscriptions were in dedication of basilicas ; some of them we shall have occasion to notice in connection with the mosaics. Suffice it to say here that with the end of the sixth century poverty and ignorance had become so general that hardly any inscriptions were produced, except the rude epitaphs of popes or of other rulers. * HISTORY OF THE CATACOMBS AFTER THE FOURTH CENTURY In the foregoing description of the Roman catacombs, and in the treatment of the problems which they raise, enough of their history has been introduced to give a fair conception of the earliest period, which illustrates specially the spirit of Chris- tian brotherhood in the extension of the privilege of the family sepulchre to all who were of the same • family in the faith. Enough has been said of the second period to reveal the extent and character of Church organization in the third century; enough also of the third period to illustrate the growing cult of the martyrs. It only remains to give a brief account of the last period of their history, which, though a long one, has little 78 CEMETERIES to register but the various stages of their abandonment till their very sites became unknown. De Rossi has proved that the custom of burying in the sur- face cemeteries above the catacombs had made such progress by the last quarter of the fourth century that only a third of the burials were subterranean. The desire for burial near the martyrs found a natural obstacle in the limits of the space which could be utilized. A Roman epitaph of the year 381 speaks of a Christian who for her great merit obtained a sepul- chre in the abode of the saints, which many desired but few obtained : . . . quae pro tanta MERITA ACCEPIT sepulchrum intra LIMINA SANTORUM quod -multi cupiuN ET RARI ACCIPIVN. From the year 400 to 409 there are almost no inscriptions proving subterranean burial ; and with the capture of the city by Alaric in 410 burial in the catacombs ceased altogether. This date is memorable as well for ecclesiastical as for political history. " In one single city the whole world perished," exclaimed S. Jerome when the news of the fall of Rome reached him in far-off Palestine. As the beginning of the barbarian inroads which laid low the whole civilization of the West, it has a special importance in the study of Chris- tian art; in the history of the catacombs it marks the most definite crisis. This did not, however, mark the abandonment of the cemeteries above the catacombs, which must already have reached a considerable size ; it was probably not till near the end of the sixth century that the law of the Twelve Tables forbidding burial within the walls ceased to be generally regarded. By this time the population of Rome had so much diminished that there remained no practical objection to intra- mural cemeteries. Even to-day a considerable part of the area within the Aurelian walls is occupied only by ruins or by vine- yards ; in the sixth century the greater part of the city must have been quite deserted, or at least left to the farmer. One of the earliest of the cemeteries in the city was established on the Esquiline near the church of S. Bibiana ; another near S. Cosimato in Trastevere. Soon each of the parish churches became the centre of a parochial burying-ground. But to return to the beginning of the fifth century. The disuse of the catacombs for burial by no means signified their abandonment. The pious veneration of the martyrs' AFTER THE FOURTH CENTURY 79 tombs is witnessed by frescos as late as the seventh century. Lamps burned perpetually before the tombs of the most famous saints, and pillars surmounted by a broad bowl of pottery or glass for the floating wicks are still to be seen in the catacombs and serve to designate the crypts which were once the resort of pilgrims. The descriptions which Jerome 1 and Prudentius 2 give of visits to the catacombs in the fourth century picture them as they remained for three centuries more. Notwith- standing the successive incursions of barbarians from the fifth to the seventh century, the popes did their utmost to preserve " the crown of martyrs " which encircled Rome. Pope Vigilus (537-555) was especially active in restoring the inscriptions of Damasus and in repairing other ravages which had been wrought by the Goths. Of John III. (561-574), the Liber Pontificalis records, amavit et restauravit coemeteria sanctorum, martyrum. From the same source we learn that this pope established out of the Lateran revenues a weekly gift of oil for every cemetery, and laid again upon each of the urban titles the obligation to send a priest every Lord's Day to the cata- combs for the celebration of the Eucharist — a custom which was continued to the end of the seventh century. 3 The seventh century was the period of the great pilgrimages from the North to which we owe the Itineraries. Pilgrimages had been com- mon since the fourth century, but they were of a private char- acter ; and it was not till the eighth century that the cult of relics led to the violation of tombs and the desecration of the bodies of the saints. Any mention which is made of relics before that time refers to a more innocent custom, chiefly the custom of carrying away some of the oil from the lamps which burned before the tomb of a saint, though in general any object — a handkerchief for instance — which was brought into con- tact with the tomb acquired the properties of a relic, and such relies were often carried to a great distance for the consecration of a new church. But the relics most usually carried away were little vials (ampullae) containing some of the oil from the lamp which burned before the tomb of a saint. The traffic in the oil am- pullae, was by no means confined to Rome; it was a custom i In Ezech. 12: 40. 3 Lib. Pont, in Sergio, I. 1. 2 Peresteph. XI. 80 CEMETERIES almost universal. The ampullae were made of glass, clay, or metal, ornamented with Christian subjects in relief (see p. 355, Fig. 54). The ampullae which were carried from the tomb of S. Mennas in Egypt were spread by pilgrims all over the world. But the most interesting collection is that which was made in the time of Gregory the Great (590-604) by the " Abbot " John, who was sent to Rome by the Lombard Queen Theodelinda to ask relics for her cathedral at Monza. John executed his mission with care ; upon a parchment attached to each vial he wrote the name of the saint from before whose tomb the oil was taken, and made, besides, a list on parchment of the whole collection. Originally there were more than seventy ampullae in this collection; many of them are still preserved in the cathedral ; and so likewise is the original draft of the list, which, as it follows precisely the order in which the tombs were visited, proved to be of the highest topo- graphical value in the rediscovery of the catacombs. The Roman Church was more conservative than any of the churches of the East in the respect which was shown to the inviolability of the tomb and in the refusal to disturb the bodies of the saints. It is certain that up to the middle of the seventh century no bodies were transferred within the walls, for the Itineraries mention SS. John and Paul (see p. 166) as the only martyrs who were to be visited within the city : " In itrbe Roma beatorum martyrum corpora Johannis et Pauli tantitm quiescunt." Indeed the fifth and sixth centuries witnessed a notable development of the buildings above the catacombs. It was in this period that the greater basilicas (basilica? majo- res) were erected over many of the tombs, sometimes incorporat- ing the basilica ad corpus of Constantinian foundation. It was in connection with these basilicas that the first transla- tions occurred. Honorius I. (625-038) changed the position of several of the tombs ; but only from the crypt underground to the basilica above. With this period the Itineraries begin to speak of the martyrs deorsum in distinction from those in basilica sursum. The first bodies to be transferred within the city were taken from the suburban towns : those of SS. Primus and Pelicianus from Momentum in 64N, and those of Beatrice and Paustinus from Porto in 682. But with the eighth century the Campagna became so thor- AFTER THE FOURTH CENTURY 81 oughly deserted and so unsafe that even the basilicas could not be preserved, and then began the wholesale translation of the bodies of the saints to the basilicas of the city. In 757 Paul I. transported a great multitude of relics to the church of S. Sil- vestro in Capite, which he had just built on the site of his family mansion. His successors, however, did not imitate his example ; Hadrian I. endeavored on the contrary to restore the suburban cemeteries. But Pascal I. effected the greatest translation of martyrs' relics which was ever made, removing in 817 the bodies of no less than twenty-three hundred martyrs to the church of S. Praxede. An inscription on marble in which he commemorated this event is still preserved in the church ; it gives the name of each martyr, the cemetery from which each group was taken, and the part of the church in which they were buried. Leo IV. so thoroughly completed this work that, by the middle of the ninth century, the catacombs were totally despoiled of the treasures which attracted pilgrims and wor- shippers, and before long they had fallen not only into dilapi- dation, but into oblivion. Those of S. Lawrence, S. Pancras. S. Sebastian, and S. Valentine were the only ones which con- tinued longer to be known and venerated, and that only be- cause of the monasteries with which they were connected. The translation of the relics of the saints was accompanied by a ruthless destruction of the cemeteries which surpassed anything they had ever suffered at the hands of Lombards, Goths, or Vandals. But after all, the bodies of the saints were saved by this official translation from a far worse viola- tion to which they were already exposed at the hands of pri- vate though licensed venders of relics. The names of several of these relic-mongers have been preserved in history. One of the most celebrated was the deacon Deusdona, who seems to have had charge of the cemetery of SS. Pietro e Marcellino, and who took advantage of his position to sell the bodies of the martyrs in Germany. It was chiefly with the northern countries that this commerce was carried on ; those who sought relics were no longer satisfied as of old with simple memorials carried from the tombs of the saints, they must have parts, however minute, of their very bodies. With the oblivion into which all but a few of the cemeteries fell, there grew up the most hopeless confusion even in the 82 CEMETEB1ES reminiscence of them. The great cemeteries of the via Appia and the via Ardeatina were popularly identified with the single cemetery which remained accessible, that of S. Sebastian ; and in the same way those of the via Tiburtina were confused with the cemetery of S. Lawrence (S. Cyriaca). It was this confu- sion more than anything else which baffled all attempts at their rediscovery, till in our own time the puzzle was unrav- elled by the genius of Giovanni Battista de Rossi. Ill CHRISTIAN ARCHITECTURE What has been said in the Introduction (p. 1 seq.) about the dependence of Christian upon Classic art is nowhere more pertinent than in the case of Christian architecture. Early- Christian architecture has often been accounted an original and significant development of the nascent Christian civiliza- tion. The architectural innovations which have been ascribed to the early Church have been reckoned to its credit or to its discredit, according as one was minded to see in them a preg- nant seed of subsequent architectural development, or merely an unjustifiable and unintelligent employment of the constitu- ent elements of Greek architecture. As a matter of fact the Church deserves neither praise nor blame on this score ; from a technical point of view it was responsible for no architectural innovations ; it brought to the world a new conception of reli- gion and a new impulse toward morality, but it brought no new art teaching and no novel architecture. Christian archi- tecture, more than any other province of art, was dependent for its forms and for its methods upon Roman, Greek, and Oriental traditions ; and it was regulated in its development by two prime factors which were both of them extraneous to Christianity : by the tendency, early noticeable in Rome, and culminating about the end of the third century, to deal freely with the constituent elements of Greek architecture without reference to their original symbolism ; and by the poverty of means which in the decline of the Roman Empire made monu- mental architecture of the earlier Roman type (vaulted halls of great area) impossible, and prescribed economic methods of construction. We shall see subsequently to what extent the shape, the construction, and the decoration of the Christian house of worship were due to causes within the Church itself ; 83 84 ABCHITECTUBE but such innovation as there was hardly involved any technical development. To speak at all of development in relation to early church architecture is somewhat paradoxical, for — to judge by exist- ing monuments — it seems as if it had rather been created. This observation is true, at least, of the West and more par- ticularly of the basilica. In the East, or under Eastern influ- ence, there was one marked development, in the construction of the dome upon a polygonal — ■ more especially a quadrangu- lar — base, a development which culminated under the reign of Justinian in the construction of S. Sophia, combining the dome with the oblong quadrangular plan of the basilica. In the case of the so-called Christian basilica, there was no incon- siderable variety within well-defined limits ; but only such variety as existed from the beginning. It seems as if the Christian type of basilica sprang at once and fully developed into existence with the Peace of the Church, as a product of Constantine's zeal for church building. Such as it then was in the earliest examples which are either recorded or pre- served, such it remained without substantial change for a period much longer than we are here obliged to study in con- nection with any of the other arts. The same type prevailed throughout the West from the fourth to the eighth century, and in some regions for two centuries longer. So long a period of arrested development it would be hard to parallel except in the history of Egypt or China. This long permanence of the basilica type proves the impotence of artistic invention which prevailed after the decay of the Roman Empire ; but it proves also the thorough fitness of this type for the need which it was designed to satisfy. This observation is substantiated by the fact that the principal constituents of the basilica (the nave with aisles, the transept, the projecting sanctuary, and the clearstory) have been incorporated in every subsequent type of church architecture. The fact is, however, that the admirable appropriateness of the basilica as it first comes to light in the buildings of Con- stantine, and its substantial uniformity throughout the whole Empire, suggest that it did not come suddenly into existence, as the invention of an emperor or his architect, but that it must have had a long, though unrecorded, development during INTRODUCTORY 80 the centuries of persecution. What that history was, we can only conjecture, lacking the explicit testimony of early monu- ments ; the following section, however, will present the reasons for believing tin it the general plan of the basilica was deter- mined by the custom of worship in the private house. But, whatever was the course of development during the age of persecution, it is obvious that it must have been thoroughly within the traditions of Roman architecture, for the erection even of church buildings, previous to the fourth century, must have been commonly in the hands of pagan artisans. Techni- cally speaking, the preparatory development of the Christian basilica is not a part of the history of Christian architecture, but of Roman. We can, therefore, the more readily dispense with a technical study of the architectural elements which enter into the Christian basilica, and all the more because such a treatment would comport neither with the character of this handbook, nor with the space! here at command. The fact that the individual elements which enter into the architectural complex, known as the basilica, were for the most part of Creek origin, only serves to point the contrast between Christian architecture and the Greek, to reveal the gap which exists between, let us say, the Parthenon and S. Paul's. It is the history of Roman architecture which bridges this gap. It may be well to sketch here, in brief terms and with special reference to the points immediately at issue, the course of this development. The effort of Greek architecture was concentrated upon a single task, the construction of the temple. The scheme of the temple was almost invariable. In base, it was rectangular and generally oblong. The Greek cult demanded a single room of no very considerable dimen- sions for the housing of the image. This room was easily fur- nished through the roof with light and air ; it was adorned in front, or on all sides, by a colonnade. The worshippers and the altar both had their place without, and it was, therefore, on the outside that the architectural decoration was chiefly expended. The architectural scheme which was devised was perfectly suitable for the object in view, and it was suitable for hardly any other. Alexandria began the free application of Greek forms to public buildings of various sorts, and Rome carried on the same process with even greater freedom and 86 ARCHITECTURE with greater resultant variety. In the construction of their temples, the Romans adhered more or less strictly to the principles of Greek architecture, which was well known in Italy before Roman times. They added, indeed, the round temple with circular colonnade, but even this was in keeping with the Greek principles of construction. For all other uses (in the construction of private or imperial palaces and villas, of public halls of justice, and of public places of amusement, such as baths, theatres, and amphitheatres) the Romans dealt with the Greek forms exclusively in a decorative interest and with little or no heed to the constructive symbolism which the Greeks always rigidly observed. This tendency culminated with the decline of the Western Empire and particularly under the reigns of Diocletian and Constantine ; and freedom in the treatment of the elements of Greek architecture was therefore the natural heritage of the Church in the fourth century. Whatever regret the purist is disposed to feel at this de- fection from strict architectural principle, and whatever blame he is inclined to mete out to the architects of the late imperial and the early Christian period must be diminished by the reflection that with the Parthenon Greek architecture had already reached its perfection, so that further progress along the same lines was no longer possible ; that a more complex civilization demanded an architectural solution for buildings of a very different character; and that the combinations re- sorted to resulted in forms which were noble in themselves and were pregnant with all the developments which have marked the history of European architecture. Roman archi- tecture developed chiefly by the elaboration of one principle which was unknown to the Greeks, and which had greater architectural worth, or at least greater potentiality for variety of application and for colossal construction, than any principle employed by them, namely, the principle of the arch — includ- ing the vault and dome. The development of these elements was the original contribution of Roman architecture, and the further development of the dome — especially its adaptation to a polygonal (most characteristically a quadrilateral) base was the sole contribution of the early Christian period. The Roman solution of the problem of the dome was inade- quate, because it was not capable of universal application ; it INTRODUCTORY 87 demanded the unlimited resources and the colossal enginery of the undivided Empire, and it depended in part upon the unique quality of Roman cement. The Roman vault and dome was moulded upon a temporary wooden support in a concrete of cement and stone; the character of the volcanic sand and stone abundant about Rome (pozzolana and tufa) made great thickness of concrete consistent with lightness ; and the whole mass solidified so thoroughly that it practically constituted a monolith with almost no thrust. In the Byzantine Empire the preparatory support of timber was commonly unavailable, and the material of construction was brick, or in some regions hewn stone. Under the constraint of building both vaults and domes in free space and without support, brick was the material most naturally employed. Various ingenious devices of construction were resorted to, but they need not be enumerated here, inas- much as they affected but little the appearance of the finished work. It is very necessary to observe, however, that the whole disposition of the building was conditioned by the necessity of providing a nice system of counterbalances to meet the thrust of the dome. The vault had a subordinate importance in early Christian architecture, but it became the essential feature of the Romanesque, and was developed under a new inspiration into the Gothic. The dome, on the other hand, was the most characteristic feature of Byzantine architecture, and the domed basilica represents the crowning and unsurpassed achievement of the early Christian period. The use of the vault and dome had from the beginning a revolutionary effect upon Classic architecture, inasmuch as it tended necessarily to transfer the architectural emphasis from the outside of the building to the inside. In Christian architecture this tendency was carried to its extreme; the exterior form of the Byzantine domed churches was arbitrary, and it commonly masked the interior disposition of the struc- ture. The type of building which we are accustomed to asso- ciate expressly with the name "basilica," the type which was all but universal in the West, and predominant even in the East, made almost no use of the vault and none at all of the dome (apart from the half dome of the apse) ; the feature which has been accounted most characteristic of it is to be traced to 88 ARCHITECTURE the Roman use of the arch. It was the employment of the arch which first made it natural to support a wall upon pillars, or upon an arcade with niched recesses. Characteristic of Roman architecture was the use of the engaged column as a mere ornament upon the face of a pillar. The engaged column played no important part in early Christian architecture, but it became of the highest importance in the development of the Gothic and entered then for the first time into dynamic rela- tion with the building. From the support of the wall by an arcade it is a conceivable step — though it must be owned a long one — to its support by a colonnade. This was a procedure unknown to the Greek, and the step was not taken by the Roman till the third century a.d. The colonnade was origi- nally designed to support only the architrave, or beam, upon which rested the timbers of the roof. In its earliest use for the support of a wall the architrave was retained with arches above it, either visible or disguised, to relieve its centre of weight. In the final stage of the development the architrave disappeared, and the arches rested immediately upon the capi- tals, or upon broader blocks inserted between. But though the architrave disappeared, — so far as its practical purpose was concerned, — the entablature was retained, being curved to follow the course of the arch. This development has been accounted the chief innovation of early Christian architecture, but it had already been carried out on purely Roman monu- ments, the best known of which is the palace of Diocletian at Spoleto. This was anything but a solid method of construc- tion, and it is open to criticism from many points of view; but it must not be forgotten that we owe to it some of the most graceful creations of early Christian, Mahometan, and Renas- cence architecture. Although the elements which enter into early Christian architecture have thus to be traced back to Roman precedents, the triumph of Christianity did involve a revolution even in the sphere of architecture, for it again concentrated architectural effort (as in Greece) upon the solution of one chief problem, the church. Palaces and other secular buildings were of small im- portance in comparison ; it was upon church construction that the zeal of the secular powers as well as the religious was chiefly intent. The Christian temple, unlike the Greek, was designed THE BASILICA — Name 89 to house and inspire a congregation gathered within it, not to attract the gaze of worshippers without. It was, if so loose an expression may be allowed, the Greek temple turned outside in. Although we have to confess that early church architecture owed so little to Christian invention, its claim to be called Christian cannot reasonably be denied. It has often been claimed for Gothic architecture that it is the only type which can properly be called Christian ; and this designation is denied to early church architecture on the ground that it is but a re- flection, though faint and somewhat confused, of Classic forms. Gothic architecture was unquestionably a most original devel- opment of the Germanic spirit under the impulse of Chris- tianity ; but, in a strict use of the term, no type of architecture can lay claim to the name " Christian " ; Christianity has funda- mentally nothing to do with it. The invidious comparison ought not to be raised ; it derives its force from the mistaken, notion that the scheme of the Christian basilica was literally copied from the Koman judicial basilica, — if indeed the halls of justice were not bodily ceded to the Church. The compari- son loses much of its point when it is reflected that even Gothic architecture is derived from the Roman vault construc- tion enlivened by the introduction of the Mahometan pointed arch. The Christian basilica has been consecrated by centu- ries of Christian worship ; it has not been surpassed, perhaps hardly equalled, in point of suitability to the various require- ments of the Church's service ; and if the origin of its most characteristic features is to be traced back to the private house and to the simple cult of the disciples who gathered there dur- ing the age of persecution, no Christian monument is more venerable, none more inseparably associated with the history of the Church. A. THE BASILICA The word " basilica " denoted originally anything kingly or lordly. Applied to a building, it suggested especially grandeur of proportion. The civil basilicas which were built by the emperors served not only as halls of justice, but for all the uses — political, commercial, and social — for which we employ our assembly halls, bourses, and market places. The greater 90 ARCHITECTURE patricians of Borne had their private basilicas for social gatherings and business of state. The term came to have a significance almost as broad as our word " hall." Although many of the basilicas, and particularly the most sumptuous of them, had an oblong rectangular plan, and were divided longitudi- nally by rows of columns or pillars, this was not true of all, and even buildings which conformed to the general type ex- hibited a radical divergence in the fact that the middle might be either covered or uncovered. The name, in fact, did not characterize a distinct type of building, nor did it specify any particular employment of it. The choice of this name for the Christian house of worship was doubtless due to its very general and rather colorless use among the Romans; they could employ it in intercourse with the heathen without sus- picion, and yet for the faithful it must have had a deeper sig- nificance, reminding them of the great King (fiamXtvs) whose house it was. The word "basilica" was radically synony- mous with dominicurn (house of the Lord), largely used in the West for the church building, and with the corresponding Greek word kyriakon, from which is derived our word " church." The Christian use of the word denoted specifically a house of worship, but it did not specify, any more than did the heathen, a particular type of building. Domed and round churches, as well as the oblong wooden-roofed buildings which we are accustomed to associate exclusively with the name, were anciently called basilicas. For example, Sulpicius Serverus relates that Helena erected " basilicas '' on the sites of the Lord's Passion, of his Resurrection, and of his Ascen- sion ; and yet the Church of the Holy Sepulchre and that upon the Mount of Olives were round structures. That the name came to be expressly associated with one type of church building was altogether natural, inasmuch as this was the original type, a type which was all but universal in the West, which remained predominant even in the East, and became absolutely normal for the church architecture of all subsequent periods. Some modern writers would extend the name " basil- ica" even to the Romanesque churches, which are similar to the earlier type chiefly in ground plan. It is important to recog- nize the kinship of these two styles, but for the sake of clear- ness it is convenient to adhere to the narrower and more THE BASILICA— Characteristics 91 popular use of the name " basilica," limiting it to the early- Christian period and excluding all round churches and all domed and vaulted structures. Expressed in the briefest terms, the characteristics of the basilica are : an oblong rectangular ground plan divided longi- tudinally into three or five aisles by rows of columns (or pillars) which support a wooden roof, the roof above the middle aisle (the nave) being raised so much above the side aisles that the windows in its supporting walls admit air and light (constitut- ing a clearstory) ; a half circular presbytery or sanctuary sur- mounted by a half dome (the apse), projecting beyond the rectangular plan, facing the nave, with which it corresponds in width, and adjoining it either immediately, or with the in- tervention of a transverse aisle (the transept) of the same height as the nave. The minor features of the basilica were subject to consider- able variation according to the taste of the architect, or to local exigencies and traditions ; there were occasional ex- ceptions even to this general scheme, which will have to be noticed in due place ; but the fixed elements of the basilical scheme decidedly outweigh the variable, and the uniformity of the basilica throughout the whole Empire is one of the first and most significant facts which confront us in this study. The origin of this type of building has of late years been the subject of much study and of much controversy ; as a conse- quence, it is now possible to give, at least in general terms, an account which will do justice to the various considerations which have divided opinion. The minds of scholars were once satisfied with a very easy solution which referred it to the invention of one of the architects of the Emperor Constantine. This notion was based at least upon the just observation that the Christian basilica, as it first appears in the extensive building operations of Constantine, appears at once in per- fect completeness, neither needing nor receiving any architec- tural improvement in the whole course of the four centuries which followed. But this fact is rather a proof that the whole of its development lay before Constantine, so that when the Church emerged from persecution into the light. of imperial favor the tentative stages were already passed, the type of 92 ARCHITECTURE building suited to its needs was already worked out and already fixed. Creations so stable do not spring all at once into exist- ence ; this view, which must be emphatically rejected, is only less absurd than the opinion, current only fifty years ago, which ascribed the Cologne Cathedral, and the origin of Gothic architecture in general, to the invention of some unknown architect dwelling at Cologne. Another view, which till recently was the dominant one, ignores the peculiarity of the Christian building and considers it merely a copy of the civil basilica. The theory is that with the triumph of the Church a great number of the civil basilicas were turned over to Christian use and thus established a tradi- tional type of church architecture. In point of fact there is no record of any civil basilica being made over to the Church, and such an alienation of public property is on the face of it far from probable if one reflects that they were needed for the purposes of civil and social life after the triumph of the Church just as much as they were before. It need hardly be said that it is the identity of name which suggests this opin- ion, and the account which has just been given of the pagan and Christian use of the name " basilica " suffices to show how little force there is in the suggestion. The theory is especially weak in that it ignores the diversity of the civil basilicas, and takes too little account of 'the most striking characteristic of the Christian type, namely, its uniformity. Even if all the elements which enter into the Christian basilica are to be found here and there in the civil, it still remains to be ex- plained when and how the Church selected and combined the forms which peculiarly suited her use and fixed them in an unalterable tradition. This is not to say that there is no instruction to be got from the comparison of the Christian with the pagan basilica ; there was, as a matter of course, great similarity between buildings which originated under the same architectural influences and which were alike designed as halls of assembly. Whatever account one may give of the origin of the type of building which became fixed as the normal scheme of the Christian house of worship, it cannot be supposed that the Christians ignored the solutions which Roman architecture furnished for problems which were substantially the same as their own. THE BASILIC A— Origin 93 It is sufficient to acknowledge thus in a general way a certain dependence of the Christian basilica upon the pagan ; the attempt to carry out a detailed comparison is at any rate ren- dered futile by the defectiveness of our knowledge of the civil basilica. We know well only the greater basilicas of Rome, and we know even them hardly well enough to deter- mine with confidence the method of roofing and lighting, — a crucial point when it is a question of comparison with the Christian type. There is at least one feature of the Christian basilica which, so far as we know, is without example in the civil, that is, the transept. The transept, it is true, was only an occasional feature of the .basilica, but it has a characteristic significance because it is specially prominent in the earliest buildings and in the greatest ; as, for instance, S. Peter's, S. Paul's, and S. Maria Maggiore. It has further to be noted that, even in the case of features which were more or less frequent in the civil basilica but invariable in the Christian (as the aisles and the apsidal extension), we are prompted to seek elsewhere for the conditions of development. From whatever point we start, we come back to the necessity of seeking the conclusive stages of the development of the Christian basilica before the Peace of the Church. It is the traditional misconception of the condition of the Church dur- ing the first three centuries which explains the prevalent indis- position to seek the origin of the basilica within that period. The idea that throughout all of this period Christians were a prey to almost unceasing persecutions, that they were under the necessity of concealing the existence of their religion, or at least their adherence to it, effectually excludes the notion that they could then have been engaged in developing the main lines of the church building, or, indeed, that they could have had any church buildings in the strict sense of the word. The account which has already been given in connection with the catacombs (p. 53 seq.) of the corporate property of the Church during the ages of persecution, and in particular of the possession of numerous church edifices which were publicly known and officially recognized as such, puts another light upon the situation and not only leaves room for the supposi- tion of such development as is here in question, but positively demands it. 94 ARCHITECTURE It must be confessed that the fact of a long preparatory- history of church architecture is far more soundly assured than any of the particular theories which pretend to trace it in detail. It is probable that there are extant no church edifices earlier than the reign of Constantine ; the last and greatest of the persecutions sufficiently accounts for their dis- appearance. In the lack of monumental evidence we are relegated to conjecture. Of the early subterranean chapels of the catacombs the most we can say is that they seem to point to a gradual development of the church edifice along lines which lead directly to the form which appears in the fourth century ; but they give us no information about its origin. There may be mentioned by way of example — and chiefly because from a German source it has been popularized in Eng- lish — the theory which would derive the basilica from the school building (scJiola). This theory has hardly anything in its favor except the suggestion of S. Paul's example at •Ephesus : " disputing daily in the school of one Tyrannus." l It is not unlikely that this precedent was frequently repeated, but only as one of a variety of expedients to find a place for Christian assembly. This theory breaks down utterly in view of the fact that, little as we know of the ancient sehola, it seems to be probable that it did not conform to any settled type and in no case very greatly resembled the basilica. • A common, and obviously a very convenient, disposition for the schoolroom was an oblong rectangular plan with an apse or exedra at the end ; but it had no division into aisles by a colonnade, and had consequently no clearstory ; the basilica, on the other hand, was never without them. Certainly the most attractive theory of the development of the basilica which could be advanced is that which refers its origin to the private house, and the Apostolic custom of gather- ing there for worship. Whereas in the Holy Scriptures there is but once reference made to the use of a school as a hall for preaching, there is very frequent reference to worship — in particular to the celebration of the Lord's Supper — in the private house, and it is evident that even in the apostolic age certain houses were distinguished as the habitual resort of the Church. 2 We have every reason to believe that the Lord's l Acts xix. 9. 2 R om , xv j. j_5_ THE BASILICA— Origin 95 Supper was, during the Apostolic age, and indeed through the first quarter of the second century, celebrated invariably in a pri- vate house. This was altogether natural in the case of a sacra- ment which was in its institution — still more obviously in its prototype, the Passover — a household meal, and represented the Church in terms of the family. That the dwelling-house was the regular place for the celebra- tion of the Eucharist throughout the first stadium of the devel- opment of the Eucharistic cultus is especially important in this connection, because the practices and ritual which centred in the Eucharist have always been one of the chief factors regulative of church architecture, and nowhere more obviously than in the case of the basilica. We gather from Justin Martyr's account of the celebration of the Eucharist that, by the middle of the second century, it had been separated from the agape and united with the ordinary morning service of exhortation and prayer. There is no reason to suppose that it ceased to be celebrated in the private house, but Justin's account seems to presume that a house which was used every Sunday for an assembly of the Christians of the whole town, and of the outlying districts, — if not for daily service, — must have been set apart exclusively to the use of the Church. From this time on we meet with the expressions domus dei, domus columbce, or simply ecclesia, used by Christian authors in such a way as leaves no doubt that they refer to houses con- secrated exclusively to Christian worship. The extension to the Church of the Eoman institution of patronage (see p. 54 seq.), which was so highly important for the history of the catacombs, was presumably of no less importance for the history of the house of worship. It enables us to realize the relation of the wealthy disciple to the brethren whom he accommodated in his dwelling, and to understand the status of the property when it was relinquished to the exclusive use of the Church. A congregation which was accustomed to worship in the chief room of a dwelling-house would be prompted to make but few changes when the property was put exclusively at their disposition, and even in the construction of a new build- ing they would be likely to adhere to the familiar plan. We may suppose that the earliest churches were either actually 96 ARCSITECTUSE dwelling-houses which had been adapted and perhaps enlarged for Christian worship, or new buildings which preserved both without and within substantially the appearance of the private house. The forms of Christian worship — particularly in the celebration of the Eucharist — were developed in the private house, and they were not easily to be divorced from the archi- tectural arrangement to which they had there been accommo- dated. If we must recognize that the scheme of the basilica was prescribed by the necessities of the Christian cultus, we must recognize that the cultus was in turn determined in part by the arrangement of the private house. It is probable that the state of affairs just now sketched broadly characterized the Church as late as the beginning of the third century. So far we may reasonably claim to be on historic ground ; conjecture begins when we seek to bridge the gap between the third and the fourth centuries. To state the case briefly, at the end of the first century — probably at the end of the second — we leave the Christians worshipping in the private house ; early in the fourth we find them in the fully developed basilica. There is here a strong presumption that the peculiarities of the latter house of worship are to be traced to the former ; that is to say, that the basilica is the product of a continuous development out of the data furnished by the Roman dwelling-house, under the concurrent influence of vari- ous types of public architecture. The disparity in point of size between the great basilicas and the ordinary dwelling- house must not be allowed to weigh against this presumption, for the scheme of the basilica was singularly independent of size, and remained the same for very small buildings as for very great. It may seem at first sight as though the disparity were equally marked in point of architectural arrangement; but as a matter of fact, there are numerous and very striking analogies between the basilica and the atrium of the Roman house. It is important to notice that the points of comparison which may be instituted do not, either singly or altogether, constitute a proof of an actual relation between the basilica and the house; but they do constitute a highly interesting corroboration of the presumption which is here posited, and it is strictly in this light they must be estimated. The pre- sumption that the basilica is in some way or another a devel- THE BASILICA— Origin 97 opment from the dwelling-house is far stronger than the proof of any particular theory designed to trace this development along specific lines. This derivation of the basilica has been for some time the dominant theory in Germany, and it has rapidly come into favor elsewhere. The general proposition is not to be prejudiced by the discrepancy of the theories which refer now to this and now to that special part of the dwelling as the seed of the basilica, nor by the instability of certain overelaborate efforts to trace the derivation and development in detail. The first application of this theory was still fettered by the accident of the name, and sought the origin of the Christian basilica in the basilicas which, according to Vitruvius, 1 were attached to the palaces of the greater nobility, especially of those who held a public magistracy, "because both public councils and private judgments, and decisions had often to take place in their houses." With the possible exception of the house but just now brought to light under the church of S. Caecilia in Trastevere, the whole extent of Roman excava- tions fails to reveal a single instance of such private basilicas as an adjunct of any but imperial palaces. It does not appear from Vitruvius's account that it was an invariable feature even of the greater houses, still less that it was characterized by any uniform style of architecture. It was precisely in respect to palace and villa architecture that there was allowed the great- est liberty of invention ; the one point, however, which most broadly characterized it was the fact that it was vaulted archi- tecture, — it stood, therefore, in the greatest conceivable con- trast to the Christian basilica. Besides, the presumption which we have raised in favor of the derivation of the basilica from the private house of wealthy disciples by no means extends to the palaces of the greater aristocracy. It is not conceivable that any considerable number of such houses could have been put at the disposition of the Church during the centuries of persecution. It is the ordinary dwelling of the well-to-do citizen, the typi- cal Roman house, to which we must look as the customary place of assembly for the early Church ; it is only to such an house we can hope to trace the history of the basilica. One of 1 De Architectura, VI. 8. H 98 ABCHITECTUBE the principal factors to be explained is the uniformity of the church edifice throughout the Empire. This is not so much a reflection of the corporate unity of the Church — for schismati- cal bodies clung to the same type of church edifice — as it is an evidence of a common origin in the dwelling-house, and of the uniformity of the better class of dwelling-houses in all the chief centres of Boman influence. It is not necessary to speak of the Syrian or the Egyptian house, or of other national pecu- liarities ; it is enough for our purpose that either the Greek or the Roman type of dwelling, or a combination of the two, was common throughout the Empire. The large " upper room," which is frequently mentioned in the Xew Testament as the place of meeting of the Church, was a peculiarity of Syrian house architecture, and it has no relation to either the Greek or Roman house. In the Greek or Roman house there was respectively but one room — the peristyle or the atrium — which could have served for Christian assemblies. The Greek peristyle (Eig. 25, a) was an open court surrounded on three sides by a covered colonnade about which the various rooms for public or private use were irregularly disposed. Opposite the doorway, and opening in full width upon the court, was a large covered room — the prostas — which was marked by the presence of the altar and hearth as the place of chief dignity and as the centre of the family life. The atrium of the early Italian house (atrium testudinatum) was very different : it also was a large rectangu- lar room surrounded by small sleeping-chambers and work- rooms ; but it was covered, together with the rest of the house, by a single gable roof. It had its tablinum, precisely corre- sponding to the Greek prnKtan, but it had also a peculiarity in two similar rooms — the alee — which flanked the tablinum and opened at right angles to it, with their full width upon the atrium, and extended in depth to the outer walls of the house. This was originally a device for giving light to the women who worked about the hearth in the depth of the atrium, which was but poorly lighted by the door, and especially so wdien the front of the house was occupied by shops. This device was effective only in the case of an isolated house ; when the towns became crowded and the houses immediately adjoined one another, the windows could no longer be introduced in the akv, and light THE BASILICA— Origin 00 and air had to be introduced by giving up the gabled roof and sloping it downward from all sides toward the atrium (atrium tuscanicum, Fig. 25, &), till it ended with a rectangular opening above a great water basin (the compluviwri). The ate, how- ever, had become so incorporated with family customs — they were in particular the place for the portraits of the family ancestors — that they remained an established constituent of the Roman house. This is of especial interest because it alone seems to explain the transept of the Christian basilica, and to explain also why this feature appears only in Rome, or in lands directly under Roman influence. -Plans of Greek and Roman house a, Greek house : peristyle and prostaa. b, Roman C, Roman house : peristyle ouse : atrium, ala?, and tablinum. - atrium. The old type of Roman house was with the beginning of the Empire relegated exclusively to the country ; with it we have nothing to do, since the propagation of the Christian faith was confined almost exclusively to the great cities. We have here to do especially with the atrium as it was embellished through the influence of the Greek peristyle. The rich required a larger house, and this was effected in either of two ways : by extend- ing it in depth by the addition of a peristyle ; or by transform- ing the atrium in imitation of the peristyle. The simplest improvement of the atrium was effected by supporting the roof by four columns at the corners of the impluvium; a more ele- gant device was that of surrounding the impluvium with a 100 ARCHITECTURE colonnade on three or four sides. The latter arrangement per- mitted the enlargement of the central open space to any extent that might be desired, and the consequent multiplication of the surrounding rooms. It is especially this type of house to which we must refer the origin of the basilica. In the example here given of it (Fig. 25, c) the alee are transposed from their proper place, evidently on account of the exigencies of rebuilding in a restricted space. The syncretism of the Empire made this transformation very natural and very common in all large towns. There is therefore very little substantial difference between the theory of Victor Schultze, who derives the basilica from the peristyle, and that of Dehio, who derives it from the atrium ; the peristyle was in the East the only room available for Chris- tian gatherings ; in the West the atrium transformed into a peristyle was the most natural choice. It only remains to point out some of the more striking anal- ogies between the basilica and the chief room of the dwelling- house. The tablinum was, obviously, the place which would be occupied by the presiding officers of the Church; by very ancient tradition it was marked as the room not only of social honor, but of a certain religious reverence. It cannot appear strange that in the basilica this room assumed an apsidal form, since this was a motive so exceedingly common in all Roman architecture of a monumental character. Between the tabli- num and the open part of the atrium stood an ornamental stone table, the only reminder of the sacred hearth. It is a very striking fact that this is precisely the position of the holy table in the basilica ; when we take into account the similarity of many of these tables with the most ancient altars (see Fig. 54), we can hardly fail to admit a close relation between them. A peculiarity connected with early Christian worship — one which had a great effect upon the architecture, inasmuch as it exacted a strict adherence to tha three-aisle scheme — was the location of the faithful in the side aisles and transept, rather than in the nave, which was more commodious, better provided with light, and distinctly more convenient both for seeing and hearing. Strange as this custom is, it seems to have its expla- nation in the fact that the nave corresponds to the open court of the atrium, which was unprotected from the weather and, therefore, constituted the least desirable room. The choir was THE BASILICA— Fundamental Traits 101 naturally grouped in front of the altar ; when the whole room came to be roofed their place was extended well down the nave, the rest of which was occupied by the catechumens. We must suppose that in adapting the atrium to the use of a church one of the first cares of the Christians would be to roof it completely. Allowing for a provision of light and air — the house, it must be remembered, was enclosed on both sides by contiguous buildings — this could be accomplished substantially in but one way : by raising a roof above the cen- tral aisle (the open court) high enough to admit light from the sides, necessarily changing at the same time the slope of the side roofs. This, which is the well-known and invariable solu- tion of the basilicas for lighting and ventilation, may perhaps have been already foreshadowed by its occasional use for the protection of the private house against the cold ; at all events, it was a device sufficiently frequent in Roman architecture, and was probably resorted to in the case of some of the civil basilicas. The proportional lengthening of the plan in the development of the basilica was due to the obvious conditions of this type of structure : it might be extended to any desired length with- out change of scheme, while very great extension in breadth required an additional pair of colonnades. Purely artistic considerations — the desire for horizontal lines of perspective converging at the altar — worked as obviously in the same direction. The omission of the colonnades on the short sides of the basilica is one of the points which is not explained by the architecture of the private house. It is to be remarked, however, that in the peristyle there was usually no row of col- umns in front of the prostas, and that in the atrium of similar form the same rule was frequently followed. Finally, in the light brick construction of the Christian basilica, which was so sharply in contrast to Roman usage in the construction of pub- lic buildings, in the almost total absence of architectural deco- ration on the exterior, and in the frequent lack of windows in the side walls, it is not unreasonable to see a reminiscence of its origin in the private house. Hitherto our attention has been directed chiefly to the early preparatory history of the basilica ; we have now to study it in its complete development, as it appears immediately after 102 ARCHITECTURE the Peace of the Church in the buildings of Constantine, and as it remained till after the eighth century. Though we have traced the general scheme of the basilica to an origin far back in the age of persecution, it is not to be denied that the tri- umph of the Church and the reign of Constantine did consti- tute an epoch, and the most critical epoch, in its history. "We can hardly call it a turning-point, because the line of develop- ment was not substantially deflected ; but it is unquestionable that up to that time the church edifice must have been very modest both in point of size and of decoration, and that it a, Cathedral of Parenzo. b, S. Pietro in Vincoli, Rome. d , S. Apollinare in Classe. Fiq. 26. — Plans c, S. Clemente, Rome. assumed then, for the first time, monumental proportions and a lavish wealth of decoration which enabled it to vie with the temples of the Classical religion, which was still intrenched in popular custom, and to attract and hold minds which were unaffected by the deeper influences of Christianity. In this final stage of its development the basilica naturally drew upon all forms of Eoman architecture and art which were compatible with its general scheme. Christian dependence upon Classical forms of architecture was rendered more direct in certain details by the frequent employment of fragments of disused buildings, — particularly columns, capitals, architraves, and cornices. The adaptation to a church use of pagan temples, or even civil buildings, was THE BASILICA— Classical Forms 103 very seldom resorted to ; but the pillage of disused buildings — a piece of vandalism which was begun by the Romans themselves — was very common after the fourth century and was carried on with immeasurable and increasing destruction throughout the Middle Ages. Rome was, of course, the rich- est field for such pillage, and the object of it was chiefly the great store of marble which was there accumulated. The early Christian period was advantageously distinguished by the fact ■' w n ^X IP i of Basilicas. e, S. Lorenzo, Rome. /, Basilica, in Suweda, Syria. 7t, S. Paul's, Eome. ff. Basilica Ursiana, Ravenna. that it could appreciate the productions of Classic art and use them in their original form, while in a later age they were regarded simply as a quarry for rough materials. The popular notion which ascribes to the early Church the wanton destruction of ancient monuments out of a radical opposition to art and a hatred of everything connected with the pagan religion, is very far from being justified. As a mat- ter of fact the temples were most of them in use in Rome long after the recognition of Christianity as the religion of the Empire, and when they were finally closed they were long kept in repair at the expense of the Christian state as the chief glory of the city. Almost all the individual forms which entered into the basil- 104 ARCHITECTURE ica were Classical, and although taken as a whole the basilica violated the principles of classical construction by ignoring the proper constructive symbolism of each part, it is nevertheless true that one of its most important formative principles — the horizontal perspective formed by parallel colonnades — was thoroughly Classical in spirit. This principle was brought to more forceful expression in Christian architecture because it was brought into relation with a ruling religious idea. Church architecture is necessarily internal architecture ; the idea which has most fundamentally inspired it is the commun- ion of saints — the Christian temple is the house of the congre- gation. This communion was concretely represented by the altar, which by ancient tradition was fixed between the clergy ♦n Maria in Cosmedin, Rome. 6, S. Apollinare Nuovo, Ravenna. Ravenna, d, S. Spirito, Ravenna. Fig. 27. — Plans o, S. Agata, and the people, — -that is, close to one end of the basilica and on the middle axis, so that toward it the horizontal lines which were emphasized by the long parallel colonnades seemed to converge. In Gothic architecture this motive was modified but not lost. There is certainly a difference to be observed in this respect between the Gothic cathedral and the early basilica, but not so blunt a contrast as is commonly drawn ; for even in the Gothic the perpendicular lines, strongly as they are marked, remain subordinate to the horizontal. In the round churches, and in domed churches of every description, there was a very different architectural principle at work ; for there the princi- pal axis was central and vertical. It seems hardly just to say THE BASILICA — Plan 105 that the Christian cult could not accommodate itself to this type of church by transferring the liturgical centre to the principal axis of the building — that is, by placing the altar in the mid- dle of the room. But it certainly did not do so ; and as long as the altar was kept near the periphery of the building there was a discord between the perpendicular axis of the architec- ture and the horizontal liturgical axis. We see from this why it is that the scheme of the basilica has remained a normal and regulative factor in church architecture, and why it must remain so just so long as the altar retains its traditional position in the of Basilicas. e, Xenodochium of Pammachius, Porto. /, 9. Maria Maggiore, Rome. g, Basilica in Kalb-Luseh, Syria. building, and the Eucharist its same high dignity in the Church. The most successful solution of domed architecture as applied to the church adapted itself to this exigency by adopting sub- stantially the ground plan of the basilica. In plan, the basilica was marked by two principal divisions : the great rectangular hall, which was the room of the laity, and the apse, which was the place of the clergy. To this must be added a vestibule, which in the East appears as a definite architectural feature of the building (the narthex), while the same purpose was served in the West by the adjoining portico of the atrium (see p. 178). The vestibule was the place for 106 ARCHITECTURE the penitents, of whom, as is well known, there were many- grades. There were many grades and kinds also within the church, each to be distinguished from the other, if not actually divided. The catechumens, I suppose, were usually admitted to the rear of the nave ; the faithful occupied the side aisles, the men on the right side of the entrance and the women on the left ; those who were held in chief honor in the congrega- tion, as the widows and virgins, and those who on account of their age or social position were entitled to peculiar regard, had their place in the forward end of the aisles or in the tran- sept; the different orders of the clergy were in turn distin- guished among themselves, the bishop had his seat in the middle of the apsidal circle, the presbyters were seated on either side of him and at a lower level, the deacons stood near the altar, and the inferior clergy had their place with the choir in the nave. All of these minor divisions were not marked architecturally, but generally only by chancels (p. 168) or cur- tains (p. 377), yet even by this the effect of architectural unity must have been seriously marred. Architecturally, even the three principal divisions — between the clergy and the laity and the unbaptized or excommunicate — were not dis- tinctly enough marked, or rather the spaces which they marked off became in time inadequate for the several classes for which they were designed, so that both a part of the clergy and a part of the unbaptized encroached upon the nave. Fundamentally the basilica constituted a single room in which priests and people gathered in common worship about the common altar. According to a strict application of church symbolism the unbaptized were not to be admitted within this room ; a place was provided for them without the doors in the atrium or in the narthex, an exception being made, however, in favor of the catechumens. The atrium was a square open court surrounded by a portico (Figs. 26, a and c, 27 e, 29, 30). With the fifth century it became infrequent, particularly in the East ; but even in case it was removed, the portico adjoining the church was generally retained, or in rebuilding it was imitated, till late in the Middle Ages (Figs. 26, b and e, 27, a). In the East this nar- row vestibule was architecturally incorporated with the church (Figs. 26,/, 27, g, 31, 52), and was called the narthex. The nar- thex, when attached to round or polygonal buildings (churches, THE BASILICA — Nave 107 baptisteries, or mausoleums), terminated at each end in an apse (Figs. 44, d, e, and 49). In Syria, where the atrium had never been in use to hinder the effect of the facade, the narthex received an architectural solution (Fig. 32) which had a marked influence — through the Crusades — upon the mediaeval church facade in Europe. The vestibule opened into the church by at least as many doors as there were aisles. We must suppose that of the penitents only the " wailers " were kept here, for the " hearers " could have heard the lections and the sermon only in case they were admitted within the church. The great body of the church, the room of the congregation, was divided by colonnades into aisles, which were always odd in number (three or five), as the colonnades were of course rm* Si Fig. 28. — Plans of Basilicas. a, Basilica of S. Reparatus, Orleansville, Africa. b, S. Agnese, Rome. Dear Rome. c, S. Sinforosa even. The width of each of the side aisles was about half that of the middle. The five-aisle system was rare, and was only adapted to the greater basilicas ; even in this case the united width of the two side aisles remained below that of the middle. The width of the aisles was determined by technical considerations, depending especially upon the strength of the roof beams. The height of the basilica was insignificant in proportion to its breadth ; the predominance of the horizontal lines was thus thoroughly preserved. The total width of the basilica very much exceeded its height; but of more impor- tance for the interior effect was the fact that the height of each aisle (including the nave) surpassed its breadth only by a 108 ARCHITECTURE small fraction — i to f . The height also was conditioned by technical considerations, namely, by the strength of the colon- nades which supported the walls. Thin and light as the brick walls were, they had but an unstable support in the columns, which were the characteristic and almost invariable feature of the basilica. This departure from Roman tradition — the adoption of the column instead of the pillar as a wall support — - we have traced to its origin in the Roman house ; and it is not difficult to detect the influences which conduced to its perpetuation. In the first place it effected a sufficient division of the room without too much marring its unity or interfering with the transmission of sight and sound ; another consideration was the artistic delight which was felt in the form and material ; while by no means least was the ease with which columns and capitals could be procured from ruined buildings. For the Constantinian basil- icas the columns were expressly made ; but later it became common to use old columns, and those of different orders and of different lengths were sometimes united iu the same colon- nade. The columns — -except those which were taken from ancient monuments — were rarely channelled; the skill to make them was doubtless lacking, but there entered also into consideration the effect of the rich glass mosaics which covered the walls, for nothing harmonized so well with this as the tone of the smooth polished marble and its broad reflec- tion of the light. The mosaic decoration had its effect also upon the capitals : beautiful Classical capitals were sometimes employed in the basilicas and even made expressly for them ; but the rich golds and purples of the mosaics with their glints of reflected light rendered somewhat vague and ineffective their sharp outlines, and the tendency which culminated in the" Byzantine capital was toward a simple treatment in low relief, without detail, or an intricate surface design deeply perforated so as to present a broad effect of light and shade. A general change in the shape of the capital was due to the disuse of the architrave and the substitution of the arch (archivolt) resting directly upon the capital. The corners of the Greek capitals were ill- fitted to support this sort of pressure ; the Byzantine capital assumed generally a very simple and solid shape, it was square THE BASILICA — Capitals 109 at the top and tapered conically to meet the circular head of the column. It presented therefore four smooth surfaces for the low relief decoration which was the only type of stone carving then practised. "When the wall was constructed of stone the two arches which sprang from the capital demanded a broader base, and in this case the capital had to be oblong with bracket-like extensions on either side. One of the peculiarities of Byzantine architecture, and one which was strictly limited to it, grew out of this same necessity ; it consisted in interpos- ing a sort of second capital as the immediate support of the arch, a nearly cubical block of stone which was tapered slightly at the bottom. This received a decora- tion similar to that of the capital, though generally less elaborate (see Fig. 45). The architrave fell more and more into disuse after the fourth century ; it lingered longest in places like Rome, where the people were accustomed to the lines of Classic architecture, and where, it may be added, this member, might be taken ready made from disused buildings. In S. Peter's, which was a ba- silica of five aisles, the central colonnades were surmounted by the architrave, the second rows by the archivolt; but it is doubtful if this arrangement dates from the original construction. The architrave was used in S. Maria Maggiore; the archivolt, in S. Paul's. The archi- traves in S. Lorenzo outside of Rome are taken from an ancient monument. Even after its general disuse as the sup- port of a wall, the architrave continued in general use for a variety of purposes, and especially to bind together a row of columns (cf. Figs. 36, 38). The substitution of the archivolt in this connection did not comport thoroughly with the horizontal principle of this archi- Fig. 23. —Plan of old St. Peter's. 110 ARCHITECTURE tecture ; but as the columns were very close together the arches were insignificant, and at any rate the horizontal line was strongly maintained by a cornice, if not also by a frieze. The panels of mosaic, or fresco (Fig. 38), or still more signally the procession of saints such as decorates the wall of S. Apollinare Nuovo at Ravenna (Fig. 37), conduced to the same effect. Structurally there was not much to choose between these two types, for if the architrave proved a poor support for the wall, the thrust of the arches tended to dislocate the columns. The arch had this advantage, however, that it permitted a greater Via. SO. — The old basilica of S. Peter's, Eome. interval between the columns, and consequently a freer pros- pect from the aisles. This, as time went on, became of ever greater importance, for ancient columns became rare and new ones more difficult to make. As a rule the interval between the columns was small in the beginning and greater in the later period ; it is specially great in the churches of Ravenna. This was carried no farther, how- ever, than the limits prescribed by the dimensions of the Classi- cal shaft. Further progress toward the unification of the room of the congregation was possible only by the use of the pillar, which, affording a more solid base, permitted the construction of a broader arch. THE BASILICA— Columns and Pillars 111 Iii the early Christian period the pillar was seldom substi- tuted for the column except in Central Syria. Its use in Syria was due in part to the lack of ancient columns to draw upon, but still more to the fact that, no wood being available, the con- struction was exclusively of stone, and consequently too heavy to be supported upon columns of Classical proportions. In one type of Syrian church three broad arches on either side of the nave replaced the colonnade (Fig. 34). In one section of the country, where wood was so scarce that even the roof had to be made of stone, the pillars of the nave served as support for great transverse arches which spanned each of the three aisles at such close inter- vals that plates of stone could be laid across from one to another to form the roof. Even in Syria, how- ever, most of the churches conformed to the common type. Elsewhere pillars were sparingly used, and with evident dis- dain. The only basili- cas in the West which are known to have been built with pillars are the basilica at Orle'ansville (Fig. 28, a), S. Smforosa (Fig. 28, c), the basilica attached to the Xenodochium at Porto (Fig. 27, e) and the older basilica of S. Felix at Nola for which S. Paulinus substituted one with columns. 1 They sometimes appear interchangeably with the columns of the nave (Fig. 38), evidently to insure greater stability ; in some eases at least they are the result of rebuilding. In the case of a transept, pillars were always used to support the arch which divided it from the nave. In the later development of dome architecture pillars became the chief structural support, while the column retained hardly more than a decorative importance. In the basilicas the height of the Fig. 31. — Church in Tlabuda. Snia. 1 Poem 28 : v. 200 : nam steterant vasto deformibus agmine pilis. 112 ABCIIITECTUBE architrave (or archivolt) varied from about T \ to -fa of the total height of the clearstory wall. A peculiarity of the East was the construction of galleries over the side aisles and opening upon the nave ; the walls above them were in turn supported by lower arcades. This was ap- parently a common feature of the civil basilicas, but in the West at least it was not originally employed in the Christian. Even in the East it was probably not used in the beginning ; Fig. 32. — Basilica in Tarmanin, Syria (restoration of facade). at all events the Syrian and Palestinian churches almost with- out exception have but one story. Elsewhere, however, the two-story scheme became the rule in the course of the sixth century, — especially in domed structures, where it was more in keeping with the architectural lines of the building (Fig. 46). As employed in the basilica it had a tendency to raise the line of vision above the altar. The basilicas of Ravenna were in this respect uninfluenced by Byzantine custom ; only the polyg- onal church of S. Vitale was constructed with galleries. The THE BASILICA — Windows 113 galleries which appear in Roman basilicas (S. Lorenzo, S. Agnese,' S. Cecilia, and SS. Xereo ed Achilleo) belong to the period (sixth to ninth centuries) during which Rome was subject to Byzan- tine influence. The galleries were apportioned to the women, and it seems probable that the stricter separation of the sexes in the East was the chief reason for their employment. Even in case a basilica was provided with galleries, the clear- story walls rose still above them (with rare exceptions, which were probably due to reconstruction), for the direct admission of light to the nave. As a general rule, the windows corre- sponded to the intercolumniations of the colonuade. They became, therefore, less frequent as in the course of time the in- tercolumniations were made greater ; but in any case they were much more frequent than the taste of a later age demanded, for it became the custom in the Middle Ages to wall up every alternate window. The same intervals were observed in the clearstory walls of the transept. There were generally three windows in the facade, — or a double row of three, — and the same under the gables of the transept. In the gables there were commonly no windows, because of the flat ceiling of the basilica. In Syria the window aperture was often rectangu- lar; but even there it was more commonly closed above by an arch, and this was the almost invariable custom elsewhere. In case windows were introduced in the walls of the aisles, they were placed at much rarer intervals. It is remarked as a point of contrast between the basilicas of Rome and of Ravenna, that the former — at least the more notable of them — had no windows in the side walls, as had the latter. The Roman basil- icas were also without windows in the apse, while elsewhere, especially in later constructions, they were usual, generally to the number of three. This is probably to be connected with the fact that the earliest and greatest basilicas of Rome, which served as patterns for the rest, had the apse directed toward the west, while later custom prescribed an easterly direction (see p. 176). Small window openings were sometimes closed only by shut- ters, as a protection against cold or rain. Ancient stone shutters, moving upon hinges, are still preserved in Central Syria. But in general the windows were closed by a latticework of metal, or by thin plates of stone — often translucent marble or ala- 114 ABCHITECTUBE baster — which were closely perforated with small apertures composing a more or less ornamental pattern (Figs. 42, 153). A window at Grado, which is here illustrated, exemplifies a method of making such plates at small expense, for, though apparently of stone, it is actually of cement, and was formed in a mould. The small apertures in the window plates were sometimes filled with glass, clear or colored; or with some translucent stone, as in the case of the early windows of S. Lorenzo, which are incorporated with the mosaic of the trium- phal arch (Fig. 130) ; but they were usually left open, allowing ingress to air as well as light. The smallness of the aper- Basilica in Turmanin, Syria. tures sufficed to protect the interior from rain, and they divided the broad shafts of light into scattered beams, which were par- ticularly favorable for the effect of the mosaics. The basilicas were very clearly lighted in comparison to most of the Gothic churches ; the obstruction to the light which was offered by the stone window plates was compen- sated by the frequency of the windows, and still more by the reflection on all sides, from the polished surfaces of marble and from the bright glass mosaic, which at once increased and soft- ened the light of the building. The cathedral of Monreale near Palermo, though a late mediaeval building, gives a more perfect notion than does any other monument of the effect of early Christian decoration in marble and mosaic. In this THE BASILICA — Doors 115 church one cannot but recognize that the effect would be en- hanced if, for the broad windows of clear glass, there were sub- stituted perforated marble plates of the ancient type. It must be recognized, too, that colored glass, such as filled the windows of the late Gothic churches, would have interfered in some measure with the definite color scheme of the mosaics. In fact, the means resorted to for lighting the early churches was far from being imperfect ; and not the least of its advantages was that it insured purity or air and good ventilation without causing strong draughts in the lower part of the church. The introduction of windows in the lower wall of the apse did not have a good effect. They would hardly have been used except for the symbolical interest which attached to the rising sun shining upon the morning Eucharist. The mosaics them- selves are the proper illumination of the apse, and the play of reflected light upon the rounded or spherical surfaces of the apse is one of its singular attractions. This effect was enhanced by the multitude of little lamps which often adorned the altar room, but it could only be marred by direct shafts of light piercing the waif. The doors of the basilica were commonly limited to the fa- Qade. The middle door leading into the nave was greater than those which flanked it, and the leaves of this door especially were a favorite field for decoration. Notable are the carved wood doors of S. Sabina in Rome (Figs. 105, 106) and of the cloister of S. Catharine on Mt. Sinai. Bronze doors from Clas- sical buildings were employed, as in the Lateran Baptistery (from the Baths of Caracalla) and in the church of SS. Cosma e Damiano (formerly Temple of Romulus). Bronze doors, as they were made in the Christian period, commonly consisted of thin plates of metal supported upon wood, as, for instance, the doors of S. Sophia, which are still in part pre- served. Such doors were generally decorated with a purely architectural ornamentation. Honorius I. decorated the mid- dle door of S. Peter's with 975 pounds of silver, and inscribed a long Latin inscription upon each leaf. Richly ornamented stone doors are found in Central Syria. Entrances upon the sides of the church were not common except in Syria, since the front alone of the church was usually free. The roof of the basilica was of wood. The only exception 116 AECIIITBCTVBE is constituted by some of the Syrian churches above mentioned; but the churches of Central Syria stand so thoroughly apart and by themselves that they deserve a separate treatment and a fuller one than can be given here. For adequate informa- tion about this interesting architecture, one must refer to the work of de Vogue. 1 In the basilicas, therefore, East and West, — and in Syria wherever wood was to be had, — a low gable roof surmounted the nave and transept, and a shed roof of the same angle covered the aisles. The commonest covering for the roof was, of course, terra-cotta tiles. At Ravenna colored glazed tiles were sometimes employed, and patterns were formed by them. The roof of great churches was often covered with lead, but the noblest material for this purpose was bronze. Hono- if 3 Fig. 34. — Longitudinal section of basilica in Euweha, Syria. rius I. had the roof of S. Peter's covered with bronze tiles. 2 The bronze tiles of the Constantinian church of the Apostles at Constantinople were gilt, so that in the sunlight they nearly blinded the beholder. 3 It is probable that in the basilicas of Central Syria the roof and its supports were visible from below. That these basilicas were not ceiled, but were open clear to the roof, is proved es- pecially by the windows which appear in the gable, sometimes a round window divided into four compartments by a cross, a precursor of the rose window. It has been commonly sup- posed that the basilicas of the West were often, if not regu- larly, covered in the same way. This notion is due solely to the fact that in the restorations which were undertaken in the 1 See Bibliography. 3 Lib. Pont., Vita Honor. 8 Eusebius, Vila Const. IV. 58. THE BASILICA— Ceiling 117 Middle Ages the flat ceiling was not retained, because it was no longer in keeping with the lines to which people had be- come accustomed in Romanesque and Gothic churches. It cannot lie denied that this scheme was often rendered very attractive — as in S. Miniato and in the cathedral of Monreale — by the rich designs in color with which the rafters were painted. But there remains no doubt that the general rule prescribed a flat ceiling for the ancient basil- ica, as well above th nave ; and it is probal was subject to no than that mentioned In this respect the tecture of the basilic influenced, not only b a Classic tradition, but by two consid- erations, the one artistic and the other practical, which attached themselves di- rectly to the reli- gious use of the building. Artisti- cally it was of great importance that the eye should not be tempted to stray up among the raf- ters of the roof and away from the apse and the altar, toward which all the chief lines of the building led; in maintaining the horizontal lines of the basilica no feature had so decisive an influence as the ceiling. Practically, it was of no less importance that the building should be acoustically adapted to its use as an audi- torium. No other established type of church building has ever so thoroughly fulfilled this requirement, nor ever seri- ously aimed to do so. The character of church worship in the Middle Ages rendered this a question of minor consideration ; 118 ARCHITECTUME but in the early period it was, as it has come to be again in the Protestant churches, a matter of supreme importance. Even for the intoning of the prayers and the chanting of the Psalms a good acoustic quality was a desideratum, but besides that, it must be remembered that the long lections and the sermon constituted a very important part of the Eucharistic celebration in the early Church. During the earlier part of our period the bishop was accustomed to address the congregation from his seat in the apse ; but in the course of the fifth century, the great size of the basilicas, as well as the interposition of the ci- borium between the cathedra and the nave, naturally prompted the practice of preaching from one of the lectors' pulpits in the body of the church. The flat ceiling of the basilica was supported by heavy beams extending from wall to wall and crossed at right angles by longitudinal beams which formed with them square or oblong spaces for deeply recessed panelling. It appears from ancient texts that various styles of panelling were in use, but that the most usual was the style traditional to the Greek temple, known as lacunaria. The ceiling was richly decorated in color and gold. Some ceilings are spoken of as entirely covered with gold, and when gold was used it must of course have dominated over any other decoration ; but it is obvious that the effect of the gold would be enhanced by some contrast in color, and in this case it is natural to suppose that the beams would be principally decorated in gold and the reeessed panels in color — dark blue and red. Some of the descriptions of early ceilings are very enthusiastic : Eusebius, describing the Constantinian church of the Holy Sepulchre, says that the golden ceiling reflected the whole temple as in a sea of light. The numerous windows which were immediately under the ceiling must have added much to its effect, and the ceiling in turn must have contributed greatly to the lighting of the whole building. Owing to the perishable material, no ancient ceiling has been preserved. The Liber Pontificalis frequently recounts the restoration of the ceiling timbers. But the res- torations were often accomplished piecemeal, and they may be supposed to have followed often the ancient pattern. The decoration of the floor and walls is treated in the chap- ter on Mosaic (page 295) ; it need be described here only so far THE BASILICA — Decoration 119 as is necessary in furnishing once for all a general impression of the basilica. The floor, in accordance with Roman custom, was finished in stone mosaic, predominantly in light-colored marble. The side walls and the recess of the apse, up to or above a man's height, were lined with colored marbles com- posed in large geometrical patterns (opus sectile). The rest of the wall space was covered with mosaic pictures, which during the early part of the fourth century were executed with natu- ral stones, afterward with cubes of colored glass, in many '■y~-::' r i.'.',^aj>^ > h^j^w^-v ^ ,., : mi\ f Fig. 36. —Cathedral of Torcello. Seventh century. tints, but usually upon a background of purple or gold. Where means failed for the execution of pictures in mosaic, they were done in fresco, and in one way or another the whole surface of the brick wall was covered and concealed, as the traditions of Roman construction demanded. There was no effort to represent in the decoration the structural purpose of the wall ; above the cornice, or the frieze, the structural symbolism was entirely ignored; the wall was treated merely as a screen which enclosed the room, with complete abstraction of the fact that it actually supported the weight of the roof. In the Byzantine domed architecture there was the same freedom in the treat- ment of the wall merely as a field for pictorial decoration. 120 ARCHITECTURE This has often been accounted a serious defect in early Chris- tian architecture ; it stands in specially strong contrast to the strictly organic treatment of the Gothic. This practice and this conception of the wall were nevertheless thoroughly in accord with Greek traditions, indeed with the early traditions of the Aryan race. The wall, in fact, constitutes but a poor support ; if it happen to serve as such, nothing better can be done with it than to ignore the fact. The comparison with the Gothic is not to the point, for in the developed style the wall practically disappeared, or rather was represented by the col- ored glass windows. Most of the wall space was portioned out into rectangular fields convenient for the Biblical scenes or other religious themes which were to be represented. There was no fixed rule, such as obtained in later Byzantine times, for the location in different parts of the church of the several themes which were represented. But, in general, historical pictures (literal Biblical illustrations) adorned the nave, while the great sym- bolical subjects were in the neighborhood of the altar. The half dome termination of the apse was invariably the place for the principal theme ; its spherical surface was singularly favor- able for the best effects of the lustrous glass mosaic. On the curved but perpendicular wall below it there was often room for another subject. As the height of the apse was consider- ably less than that of the ceiling, and its breadth generally less than that of the nave, it intersected the wall of the church in a great arch (the apsidal arch) which provided a space hardly less eligible for a single grand theme. It was under this arch that the altar was located ; or, in case there was a transept, under a second arch (the triumphal arch, according to a later terminology), of the same height and similarly decorated, which separated the transept from the nave. These two arches with the apse formed one complex as seen from the nave ; the Apoc- alyptic visions which adorned them are the very triumph of the mosaic art. The apse and these neighboring arches con- tributed most essentially to the total effect of the basilica; they constituted the only variety which enlivened the monotonous lines of its architecture ; it was at the triumphal arch the hor- izontal lines were first broken, and the gaze which was irre- sistibly drawn to the altar was then first bid to soar. This THE BASILICA — The Presbyterium 121 complex of arch and apse, with the noble decoration which adorned it, constituted the crowning glory of the basilica; it deserves to be compared to the prospect which was enjoyed by the Greek worshippers, who stood without and gazed beyond the altar to the pediment of the temple. Having already described the vestibule of the church and the room of the congregation, it remains only to speak of the room Fig. 3T. — S. Apollinare Nuovo, Ravenna. Sixth century. of the clergy. This part of the church was known by various names according as one or another aspect of it was present to the mind. The name presbyterium was given it as the place of the higher clergy ; apsis, exedra, and concha refer to its form ; bema refers to the fact that it was raised by several steps above the floor of the church ; and tribunal (also tribuna), to its like- ness to the platform occupied by the magistrates in the civil 122 ARCHITECTURE basilica. A good deal that belongs to this topic has already- been anticipated incidentally in the course of the preceding paragraphs ; still other subjects which one might expect to find treated here must be postponed, so far as any detailed descrip- -S. Maria in Cosmedin, Rome. Eighth century (restoration). tion is concerned, to a later page, — all, that is, which belongs not so strictly to the architecture as to the furniture of the church, as the altar and its ciborium, the chancels, the pulpits, and the seats. In this connection, however, attention needs to be called to THE BASILICA— The Transept 123 the architectural importance of the ciborium, or canopy, which surmounted the altar. The nearest analogy of the ciborium is to be found in the temple-like structure which frequently sur- mounted a tomb or covered a sarcophagus. The immediate suggestion of its use was obviously the desire to do honor to the altar by covering it, and by separating it ideally from the rest of the building. But, besides this, with the construction of great basilicas there arose an architectural necessity for this or for some similar device. The altar, no matter what might be the size of the church, retained always the same very limited dimensions. Of itself, therefore, it was ill fitted to constitute the architectural centre of a huge basilica ; it needed then, as it has always needed, some architectural adjunct which might vary in size with the proportions of the building, and serve even in the greatest to mark the altar as the preeminent object of regard, as the centre and explanation of the liturgical order of the church. This was approximately the only solution which was possible so long as the bishop's throne remained in the middle of the apse and the celebrant faced the people from be- hind the altar. The reredos of the Gothic church was another solution under changed conditions. It can hardly be accounted so successful a one, since it afforded only a background and not a shrine for the altar, and was besides too closely incorporated with it. It is evident that in the original disposition of the basilica the apse was intended to furnish sufficient room for all of the higher clergy. How thoroughly traditional this feature was, is proved by the tenacity with which it was conserved after it had ceased to correspond to the growing importance of the clergy and the increasing elaboration of the ritual. The room of the clergy was clearly circumscribed, and therefore the more ex- pressly limited, by the fact that the half-circular floor of the apse was raised by two or three steps above that of the nave. The altar was commonly raised a few steps higher and stood in the middle of the chord of the apse ; on either side of it were chancels which separated the room of the clergy from the rest of the basilica. If this spaGe was not actually too small to accommodate the number of clergy who might be present, it was still not large enough to comport with their dignity, or to distinguish clearly 124 ARCHITECTURE the different orders. The want of a greater space must have been felt already before the end of the third century, that is, before the construction of any of the basilicas which we know. The fact that this want was not satisfied by any architectural innovation, — such, for example, as a deepening of the apsidal room after the manner of the Gothic choir, — but by the mere makeshift of extending the chancels into the nave and across the aisles, is attributed to the impotence of architectural inven- tion in the early Christian period. But it must be observed that, whereas the Gothic with its lofty pointed arches naturally invited the construction of deeply recessed spaces beyond the quadrangular plan, the lines of the basilica as distinctly for- bade it. The round or polygonal churches which were con- structed with dome and vault readily allowed of a deeply recessed apse (Figs. 49, 50), but in the basilica it must have destroyed the clear and noble effect of the half dome and arch, which we have but just now had occasion to praise as its chief architectural excellence. The transept furnished an adequate solution for this need, although it was not developed expressly to meet it, but was derived, as we have seen, from a traditional feature of the Roman house, and was therefore limited to Rome and to the lands which were most directly under the influence of the Roman Church. The transept was as a rule somewhat narrower than the nave ; the only exception is the transept of the Lateran, which is slightly wider. The lateral termination of the tran- sept generally coincided with the wall of the aisles, and there- fore did not interrupt the quadrangular plan of the basilica. The exceptional cases in which it extended slightly beyond this line belong, however, to the oldest and most important of the Roman basilicas, — S. Peter's, S. Paul's, and the Lateran (Figs. 26, h, 29), — and it was probably the great importance of these churches which contributed, together with purely architectural and symbolic considerations, to impress this scheme upon the Gothic cathedral. In a church which was provided with a transept, the altar was not located beneath the apsidal arch, but on the line which separated the transept from the nave, beneath the areas major, as it was called in the Liber Pontificalis. This left between the altar and the apse ample space for the clergy, and in its wings provided accommodation for the altars of the THE ISASILICA — Pruthcsis 125 prothesis, upon which tlie offerings of the people were bestowed and from which the bread and wine for the Eucharist were taken. The whole of the transept, or at least the part reserved \ ' - ",\ > "' - i i i Km;, SO. Cathedral i>f 1 to the clergy, was raised like the floor of the apse, and sur- rounded by chancels, sometimes also by rows of columns and curtains. 126 ABCHITECTUBE The altars of prothesis were obviously a necessary part of the furniture of the church so long as the offerings of the people were largely made in kind ; a mere reminiscence of them is preserved in the credence upon which the Eucharistic ele- ments are placed before they are carried to the altar. It appears that one table of prothesis often sufficed ; but the separation of the sexes to the one side and the other of the church must as a rule have required one table at the end of each aisle, or, in case there was a transept, in each of its wings. The Liber Pontificalis records the gift by Constantine to the Lateran of " seven altars of purest silver weighing two hundred pounds each." This has been relied upon as a proof of the plu- rality of altars in that early age of the Church ; but these silver altars were evidently nothing else than tables of prothesis, and they correspond, as Duchesne points out, to the number of dea- cons in the Roman Church. In case there was no transept, there was room found for these tables and for the services connected with them at each side of the apse at the extremities of the aisles, and this room was set apart merely by means of chan- cels and curtains. Remains of such chancels are still in place in the ancient basilica of S. Reparatus in Orleansville (Fig. 28, a). In churches which were without the transept — as were many of the Roman basilicas, and all those of Ravenna and of the East — there was no means of enlarging the apsidal room except by placing the altar some distance in front of the chord of the apse, and extending the platform and chancels corre- spondingly. But the Eastern Church found some compensa- tion for the transept in the construction of two rooms which projected beyond the quadrangular plan, flanking the apse on either side and opening upon the side aisles (Figs. 26, d and /, 27, d, 33). Except in Central Syria the original shape of these rooms cannot often be established with certainty, for few parts of the church were liable to suffer so much from restora- tion. One of the rooms, called the prothesis, was intended to receive the offerings of the people; the other, the diaconicon, served, like the later sacristy, for keeping the church utensils and the sacred and liturgical books. The diaconicon was used for the vesting of the clergy, and to it the people also might resort to study the Scriptures. The latter fact is indicated by a titulus which Paulinus of Nola devised for the door : — THE BASILICA — Prothesis 127 SI QUBM SANCTA TENET MEDITANDI IN LEGE VOLUNTAS, HIC POTEKIT RESIDENS SACEIS INTENDEEE LIBRIS. In plan the two rooms were identical, but the prothesis opened by a greater door into the church. There was sometimes a direct communication between the diaconicon and the presby- terium : but in the East there was never such a communica- Fig. 40. — Apse of the basilica Severiana, Naples. Fourth century. tion in the case of the prothesis, for one of the most impres- sive features of the Oriental ritual was the passage of the priests carrying the Eucliaristic elements through the body of the church and in through the chancel doors to the presby- terium. There was no rule to establish upon which side of the church the prothesis and the diaconicon ought respectively to be placed. In the course of time, as offerings in kind fell out of use and the other uses of the prothesis were conveniently served by the 128 ARCHITECTURE diaconicon, which was commonly absorbed in a large sacristy, both of the rooms were often transformed into chapels for the side altars which came into use as a consequence of the cult of relics. In external appearance the basilica stands in very disadvan- tageous contrast to all other types of Roman public architec- ture. The building material was almost invariably of brick, Fig. 41. —8. Paul's, Rome, interior as recently restored. and the brick construction after the second century of the Empire was of a very coarse character. Commonly as brick was used by the Romans, it was accounted too base a material to be exposed to view, and in all buildings of any pretension it was covered with marble, or hewn stone, or at least with plaster and paint. Although the Church followed this tradi- tion so far as concerned the interior of the basilica, on the out- side the bare brick wall was left exposed without any, or with the very least, attempt at architectural decoration. The justice THE BASILICA — Exterior 129 of the principle, however, was acknowledged by the fact that the faqade was sometimes decorated, like the interior, with mosaic. It has to be remembered that, while the facade was freely visible from the atrium, the rest of the basilica was likely to be so completely surrounded by chapels, mausoleums, hospices, and schools (Fig. 55) that the exterior finish of the lower wall was not of prime importance ; and the wall of the clearstory was to some extent enlivened by its close row of windows. It was by slight mouldings around the windows that the first step was made in Ravenna toward a decorative treat- ment of the brickwork. In Ravenna and in the East the apse was generally faced with a polygonal wall (Fig. 33) ; in Africa the apse and the adjoining rooms were often masked by a straight wall, — for a Syrian example see Fig. 61. The best that can be said for the external appearance of the basilica is that it gave a perfectly frank and lucid idea of the interior disposition of the room. For the rest it revealed only too plainly its derivation from a private style of architecture. The lavish, expenditure of costly decoration upon the interior stands in contrast not only to the bareness of the exterior, but to the exceeding cheapness of the construction. In the domed type of architecture the walls had of course to be much thicker, the construction was altogether much more costly, and even the exterior could not but have a greater architectural interest and variety. It was this type which continued the traditions of Roman public architecture ; but the exterior followed the example of the basilica, and the bare brick walls had in some respects a worse effect, because they were meaningless, and masked the interior disposition of the building. This rudeness was not altogether due to carelessness about the exterior effect, for there was a distinct effort to make at least the approach to the church — the f aqade, the atrium, and its porch — as attractive and imposing as possible. It was pre- cisely the faqade, however, which proved the most intractable feature of the exterior. From outside the atrium only the low gable and the clearstory wall were visible ; from within, the case was not much better, for the portico with its close line of slender columns could in no way be reconciled with the sheer mass of the upper wall, unbroken save by a few windows. The problem of the faqade was to be solved only by incorporating 130 ARCHITECTURE the vestibule or portico more closely with the church, as was done with such signal effect in Central Syria. But this brings us to another story: these churches of Cen- tral Syria — most of them of the fourth or fifth century — stand in the sharpest contrast to all other architecture of the early Christian period. Looking at such a facade as that which is represented in Fig. 32, it is hard to believe that it is not at least seven centuries later. In point of fact, we are compelled to believe that the early Norman churches in the neighborhood of the great crusading ports of Southern France and Apulia were directly inspired by these Syrian basilicas. Syria, it must be remembered, was one of the most flourishing centres of early Christian life, and it was not only in close contact with Greek Hi! m ma m si Fig. 42. — Stone windows and Roman civilization, but it inherited more fully than any other Christian land the traditions of Oriental art. It is altogether likely that Syrian architecture had a marked influ- ence upon all neighboring lands — except Palestine, where all the great churches were due to the munificence of the Roman or Byzantine emperors. At all events its free ami strong development of the theme which was presented by the basilica was not surpassed till the advent of the Gothic. From a decorative point of view the interior of the Syrian basilica was far less effective than the more usual type which we have been studying ; it lacked the rich pictorial mosaics, and relied upon a strictly architectural decoration of the stone surface. From an architectural point of view it was incomparably stronger and more consistent, an advantage which was dis- played even more conspicuously on the outside than within. It THE CENTRAL TYPE— In General 131 is quite beyond our scope to do justice to the admirable char- acter of its masonry and the detailed architectural decoration which did so much to ennoble its appearance ; it is better frankly to omit this theme than to give the impression that it can be exhausted in a few words. But one word must be said about the fagade, though it be only to direct attention to the characteristic scheme which is exemplified in Fig. 32. It will be seen that the chief features are : a free flight of steps lead- ing through a broad arch into a vestibule, which is surmounted by an open loggia and flanked by two gabled towers which symmetrically balance the gable of the nave. The fundamental characteristic of this scheme consists in the two towers which mask the ends of the aisles and balance the clearstory ; when this was given, all the rest was sufficiently obvious. The church tower introduces a problem which is not .alto- gether explained. It is not sure at what date it was first used for church bells ; it is probable that its original purpose was defence. But whatever its purpose, it was common from the fifth century both in the East and West ; and always, till late in the Middle Ages, it maintained its isolated position (Fig. 30). It was only in Syria that a pair of towers was employed, and only there that they were brought into organic relation with the facade. B. THE CENTRAL TYPE In the endeavor to give a total impression of the decorative as well as architectural effect of the basilica much has been said which applies equally well to the Byzantine domed churches. The ritual arrangements were the same in both cases, as were also the principles of decoration. There is left to consider merely the structural conditions and architectural principles which distinguished the round, polygonal, or domed structures, of various sorts, whether churches, baptisteries, or mausoleums. Even upon this express topic it has been neces- sary to anticipate incidentally a great deal which, having once been said, need not be repeated here. By the central type of architecture no definite scheme of con- struction is indicated, but only the general principle of symmet- rical — more strictly eurythmic — arrangement with reference to a central perpendicular axis. This principle is exemplified 132 ARCHITECT USE as well in the round buildings which consisted of concentric colonnades covered by a conical roof (Fig. 44 a, b, and c), as in those, whether polygonal or round, which were surmounted by a dome. The dome, however, is the typical example of this principle, and, wherever it is employed, it exercises a strong centralizing effect. In the case of the mausoleum and the baptistery the eurythmic symmetry could be strictly observed, but it has already been remarked that, as applied to the church, it encountered an opposing principle in the horizontal axis which was exacted by the ritual, and in S. Sophia at Constanti- nople we have what has been called a compromise, though it is more properly a reconciliation, with the liturgical plan. It is evident that, in turning from the study of the basilica to the central structure, we pass from a definite and invariable type to a congeries of types which were applied to divers uses and were even more various in their character ; from a building which had no history and only a modicum of architecture we pass to the study of an intricate and original development on purely architectural lines. It is quite out of the question to treat here the many and important technical problems which this study suggests ; for this the student must be referred especially to Choisy, Histoire de I' Architecture, Tome II. The earliest Christian examples of this class of architecture are those which were constructed at Rome, or by the Emperor Cbnstan- tine under purely Roman influence at Jerusalem. But the development which we have to trace belongs exclusively to the Byzantine Empire, where invention and progress were still possible, owing not only to favorable political conditions and immunity from barbarian attack, but also and chiefly to the influence of the architectural traditions of the Orient. The development of Byzantine architecture culminated before the middle of the sixth century in the reign of Justinian. The various stages of this development are marked by buildings of a very high interest ; its culmination in S. Sophia at Constan- tinople registers the highest limit attainable along these lines. This building' was never imitated, but alongside of it the archi- tects of Justinian developed another type — the church of the Apostles at Constantinople — which furnished the pattern for subsequent Byzantine architecture (example, S. Mark's at Venice). Justinian distinguished his reign not only by the de vel- THE CENTRAL TYPE — In General 133 opment of a distinct type of architecture, but by the great number of buildings which he erected, in Jerusalem, in Constan- tinople, and in many of the great cities of his Empire. In his zeal and munificence as a builder of churches he has not been equalled either before or since. It must suffice to state here in the briefest terms the chief innovations which characterized the development of Byzantine architecture. It has already been stated that in Roman con- struction the vault and dome consisted of a thick bed of concrete which hardened into a monolith, exerting no lateral thrust, but requiring a heavy wall for its support. The technic of Byzan- tine construction was borrowed from the East ; the use of brick made it possible to construct both vault and dome without the cost of a temporary support and with far less expenditure of con- structive material and of manual labor. This change of tech- nic had a far-reaching effect upon the architecture. On the one hand the comparative lightness of the dome permitted its support upon a lighter wall ; or, what amounts to the same thing, the organic solution of the wall into a series of pillars (Fig. 47) which were connected above by arches, and gave free prospect and passage into a surrounding room (Fig. 48). On the other hand the lateral thrust which was developed required a nice adjustment of buttresses and balances, which — according to the traditions of ancient architecture — must be provided within the building and contribute to the interior space ; that is to say, the thrust required the construction of a room, or rooms, surrounding the central chamber, and the use of free pillars made possible ready communication between the one and the other. The most thorough attainable continuity between the cen- tral and the surrounding rooms was provided by reducing the number of the supporting pillars to four (Fig. 52). This in- troduced a square plan with which the round base of the dome corresponded only at four points, and those the weakest for support, namely, the centres of the four arches which con- nected the pillars. This difficulty was solved by the penden- tive, which was the most characteristic invention of Byzantine architecture. The pendentive was a construction which filled out the angles between the square (or polygonal) plan and the circle of the dome : its surface formed a spherical triangle, one 134 ARCHITECTURE side corresponding with the first course of the dome, the others following the supporting arches and concentrating the thrust upon the pillars (Fig. 51). The final development in dome construction consisted in erecting above the pendentives a cylindrical wall (the drum), which allowed room for window apertures without opening them in the dome itself. It is the dome with a drum which became the common type in later Byzantine architecture, and which was universally employed in the Eenascence ; it contributed, however, more advantage- ously to the exterior than to the interior effect of the building. So much variety and invention entered into the develop- ment of the Byzantine church that it is difficult clearly to trace its course. It is impossible to reduce this course to a single line of development ; for it is evident that there contributed to its formation several distinct types which were contempora- neously in use at the very beginning of Christian architecture ; and it is evident, moreover, that even in the final development these lines did not converge upon one point, but continued to reveal their several origins by the constitution of three distinct schemes — the polygonal, the cross-shaped, and the quadran- gular. The influence which contributed to the formation of the quadrangular scheme was obviously the basilica. As to the equal-armed cross, that was an early and usual plan for a mausoleum or memorial. The round and polygonal buildings which date from the earliest period of Christian architecture were likewise intended for mausoleums and memorial shrines, and also for baptisteries. Throughout the fourth and fifth centuries the employment of the central type for the church proper — that is, the congregational house of worship — was exceedingly rare; it is reduced almost to the vanishing point when we exclude the buildings which were originally intended for mausoleums or memorials, and such as were directly imi- tated from them. Of the round and polygonal buildings we have to distinguish two chief types according as the plan is simple or composite. Among those of composite plan we must distinguish again, with reference to the mode of roofing, between the buildings which consisted merely of concentric colonnades surmounted by a wooden roof, and those which were surmounted by a dome (supported by pillars or columns) and a ring-vault. THE CENTRAL TYPE— Simple Plan 135 BOUND OR POLYGONAL BUILDINGS SIMPLE PLAN The simplest case is a cylindrical wall surmounted by a dome in the form of a half-sphere. The heavy dome construc- tion of the Romans required a very thick supporting wall, which for economy of building material, as well as for purely artistic considerations, was broken by a series of niches. The Pantheon at Rome is the most colossal and altogether the noblest example of this type ; the wall surface is diversified by eight great niches (including the door), alternately square and semicircular, with another row of niches above; but it becomes continuous again before it meets the dome. Light is admitted by a great opaon in the centre of the dome. The great magnitude of the Pantheon made possi- ble a special decorative feature which consisted in rows of columns in front of the niches, but the application of the general scheme was obviously quite independent of size. Rooms of this type were constructed by the Romans as ad- juncts to palaces, but more commonly as bath rooms and mauso- leums. The Christians employed them for the same purposes, that is, as mausoleums and baptisteries. This type was mani- festly appropriate to the mausoleum, not only on account of its shape, but on account of its monumental solidity. While in the construction of their churches the Christians showed a strange indifference about the solidity and endurance of the buildings, they seem to have shared the solicitude of the pagans for the eternal perpetuation of their tombs. The two round buildings which formerly adjoined S. Peter's (Fig. 30) were of this type, adorned with eight square niches, and they were probably constructed as mausoleums. One of them was in the eighth century dedicated to S. Petronilla ; but it Fig. 43. — Plans of buildings of the central type. Simple plan. a, Orthodox baptistery, Ravenna. 6, S. George, Thessalonica. c, Arian baptistery, Ravenna. 136 ARCHITECTURE was originally the mausoleum of the family of Theodosius, and it is probable that the other, dedicated to S. Andrew, had a like origin. Quite similar to these is S. George in Thessa- lonica (Fig. 43, b). Doubtless this too was originally a mauso- leum, either Christian or pagan; for the presbyterium seems evidently to have been a later addition, connected with the building by breaking through one of the original niches. Though such buildings were never constructed for use as churches, they were obviously very appropriate as baptisteries. The baptistery was essentially a bath room, and it naturally assumed the shape which was common for public and private baths among the Romans. The pool or basin occupied the centre, and there was sufficient room about it to accommodate the candidates, with their sponsors and the clergy ; baptism, being essentially a personal and private sacrament, did not require the presence of the congregation. The altar had properly nothing to do with the baptistery, though in the East it was occasionally accommodated in one of the niches, or an exedra was even expressly constructed for it. Among the Romans the pool was commonly octagonal, and this became the almost invariable tradition in the Church. S. Ambrose attempted to attach a symbolical significance to the octagonal form ; 1 and the fact that this form has been perpetuated in the mediaeval (and modern) fonts is perhaps to be traced back ultimately to the Roman bath. In the employment of deep niches in the thick wall (a con- stantly recurring feature of Roman construction) there lay the suggestion of a further development ; the niches had only to be opened clear through the wall, and the central room was at once put in communication with surrounding rooms or with a corridor. This step was taken in the construction of the so-called temple of Minerva Medica at Rome ; in this case ten niches open upon a circular corridor. From the imperfection of its construction the building is judged to be not earlier than the end of the third century. Close as it seems to stand to the Byzantine round church (Fig. 47), the relation was in fact not direct ; the line of Byzantine development passed through other forms and was influenced by other examples. Equally outside the line of development is the great dome of the Baths of Cara- 1 Gruter, Thesaur. inscript., p. 1166. THE CENTRAL TYPE— Simple Plan 137 calla, which was supported upon eight mighty piers ; this ex- ample merely shows how natural was the deduction of the Byzantine type. The mausoleum of Theodoric at Ravenna (Fig. 53) is hardly to be classed with any other Christian monument. _ It seems, however, to be a product of Syrian architecture. It contains two rooms, the lower of which is in the shape of a cross, while the upper is round. Externally it is ten-sided, surrounded by a platform which was originally covered, and it is surmounted by a colossal monolith. It must now be noticed that there were, strictly speaking, no baptisteries with cylindrical walls such as we have been considering. This type was expressly imitated in early Chris- tian baptisteries, but under changed structural conditions which modified the form of the room. One of the consequences of the brick construction which was employed in Christian architecture was a comparatively thin wall which did not admit of deep niches between its two surfaces. If the tradi- tional niche decoration was to be retained in use, it could be accomplished only by the construction of apsidal additions projecting beyond the wall. That the niches were to be re- tained was practically assured by the fact that they were not only the essential decorative feature of this style, but served at the same time as buttresses to counteract the thrust which was developed by the dome under the new type of construc- tion. Illustration of this is furnished by the two baptisteries of Ravenna. The plans (Fig. 43, a and c) show how the con- struction of four semicircular niches, with intervening wall spaces of equal width, transformed the interior into an octagon, while the exterior in the one case closely approached the form of a square, in the other suggests the form of a cross. It is evi- dent that the construction of niches on the remaining four sides would have given an unsightly form to the exterior. The harmony of the interior was preserved by eight false arches which covered not only the niches, but the intervening walls, and rested upon engaged columns. The octagonal wall was carried up considerably above the summit of the niches in order to admit light beneath the dome. The diagonal section of such a building gives substantially the profile of the basilica, the high wall supporting the dome corresponding to 138 ARCHITECTURE the clearstory ; while, on the other hand, it also presages the drum of the developed Byzantine style. This already leads us a step in the direction of the Byzan- tine round church. It would only be necessary to develop the structural symbolism which was indicated by the false arches and the columns which decorated the interior — that is to say, to support the dome upon real pillars, do away with the wall, and merely mark its place by rows of columns — and we should have the central hall of SS. Sergius and Bacchus. A similar plan, but with eight niches, would give the central hall of S. Vitale (Fig. 47). What is lacking here, however, is the sur- rounding room or corridor, which alone could furnish any sug- Fiq. 44. — Plans of buildings of a, Mosque of Omar, Jerusalem. 6, S. Stefano Eotondo. Home. c, Church of the Ascension, Jerusalem. gestion or any reason for thus opening the niches. We have therefore to consider the influence of two other classes of buildings which were contemporaneous with the very begin- nings of Christian architecture, and which served to accustom the eye to the perspective effects of concentric rows of pillars or columns. In this connection, however, we are already brought a step nearer S. Yitale by the octagonal church which Constantine constructed at Antioch. The account of Eusebius seems to represent that it was surrounded by deep niches, alternately square and semicircular, which were divided by galleries into upper and lower stories. THE CENTRAL TYPE— Composite Plan 1.39 COMPOSITE PLAN This topic comprises two distinct early types of building, besides the Byzantine polygonal type. The fundamental dis- tinction between the two early types is that the first (Fig. 44, a, b and c) was roofed with wood, the second (Fig. 44, d, e and/) surmounted by dome and vault. It cannot be claimed that the first type had any influence upon the development of the Byzantine style beyond the merely formal suggestion which was given by its concentric colonnades. In all other respects it stands at the farthest remove from the Byzantine style of construction, and in the closest relation to d e f the central type. Composite plan. d, S. Costanza, Rome, e, Lateran Baptistery. /, S. George, Ezra. the basilica. Its fundamental features were: a circular (or polygonal) wall enclosing two or more concentric colonnades, each of which — like the colonnades of the basilica — supported a wall upon which rested the roof. The room within the cen- tral colonnade was sometimes completely or partially uncov- ered ; and in this case the roof of the surrounding aisles had the shape of a truncated cone. In case the central room was cov- ered, its walls were raised high above the aisles (like a clearstory) for the admission of windows, and surmounted by a conical roof. What has been said in reference to the basilica, about technic of construction, about wall decoration, ceiling, floor, columns, and arcades, applies equally here. Certain general similarities of scheme between this type and the basilica will also readily be noticed. But the attempt which has been made 1(0 ARCHITECTURE to deduce it directly from the basilica rests upon a too fanciful comparison. On the other hand, it has not proved possible to trace this type to any example of classical architecture, unless it be the Marnion at Gaza. It emerges for the first time in the reign of Constantine, and may therefore be ascribed, if one will, to the Christian period, though hardly to any dis- tinctive Christian influence. It seems more just to say that its origin is explained by the free inventiveness, the almost wan- ton liberty, which characterized Roman architecture at the beginning of the fourth century. But whatever its origin, this type of building has a peculiar interest for Christian history, since it was employed by Con- stantine in the erection of the first great shrines which marked the holy places of Jerusalem. These buildings have all of rf Fig. 45.— Early them perished; we know them only from the descriptions of Eusebius and of mediaeval pilgrims, and from the imitations to which churches of such fame naturally gave rise. The only churches of this sort which exist are S, Stefano Rotondo at Rome (Fig. 44, &), of the middle of the fifth century, and S. Angelo in Perugia, a reduced copy of it made about a cen- tury later and now half destroyed. S, Stefano has long stood as a riddle among the monuments of Roman architecture. Dehio with some plausibility attributes its erection to Placidia, and thus connects it with the church which her grand- mother Eudoxia built at Jerusalem in honor of the proto- martyr Stephen and as her own mausoleum. This historv brings it at once into relation with the memorial build- ings of Constantinian foundation; it also explains how this type, which was expressly and appropriately designed for a THE CENTRAL TYPE— Composite Plan 141 memorial, came, through the celebrity of Constantine's exam- ple, to be copied for this use not onl}', but also for use as a house of worship, for which it was quite unsuitable. That it was not originally intended as a substitute for the basilica, nor for the ordinary uses of church worship, is clearly shown by the fact that Constantiue was not content with the erection of round shrines about the Holy Sepulchre and upon the spot whence our Lord ascended, but adjoined to each of them a great basilica. The round temples were doubtless used for private devotions and for extraordinary religious celebrations, but their most definite purpose was that of doing honor to a sacred spot and protecting it from profanation. They had therefore practi- cally the same intention as the Greek temenos, and the central type was manifestly prescribed by this purpose. f n the year 326 Constantine erected the church of the Holy Sepulchre at Jerusalem. The sepulchre itself was enclosed Christian Capitals. within the inner colonnade of a round church of the general type just described. Eusebius gives a detailed but far from clear description of it, though the chief traits are tolerably well fixed. Its chief peculiarity was that it opened into a huge five-aisled basilica. It is supposed that the outer wall of the round building, after it had completed something more than half a circle, met the extended walls of the basilica and left therefore free communication between the two, equal to the whole width of the basilica. The round building was cov- ered with a wooden roof which left part of the inner circle open to the sky. This shrine was most commonly called, not the church of the Sepulchre, but the Ana&tasis — the Resur- rection. It was the round church which properly constituted the shrine ; the basilica was added to it to accommodate the congregation. The whole edifice was destroyed in the year 142 ARCHITECTURE 614, by Chosroes II. of Persia, and only the round building was afterward restored. Together with other churches of Palestine, it was described, early in the eighth century, by the Scotch abbot, Adamnanus of Iona, who merely recounts what he had heard from the French pilgrim and bishop, Arcul- phus. This building was a number of times destroyed and restored, and finally the crusaders added again to it a rectan- gular church. The building was conserved in substantially the same form till a fire in the year 1808. The round church, altered as it was by Byzantine reconstructions, exercised a great influence upon late mediaeval architecture. The Constantinian church of the Ascension on the Mount of Olives has likewise suffered destruction, though Arculphus, at the end of the seventh century, still saw it in its original state. It likewise had a basilica annexed to it, but not in direct com- munication. Its central room was likewise unroofed, — and with manifest propriety, since it marked the spot whence our Lord had ascended into the heavens. Dehio reconstructs it as a round building like S. Stefano Rotondo (Fig. 44, b aud c) ; but he admits that it may originally have been octagonal, as was the crusaders' restoration of it. This receives further support from the fact that the mosque of Omar (Fig. 44, a) is octagonal, and was probably imitated from it (cf. p. 310). The mosque of Omar is not only a notable instance of this style, but one of the most remarkable monuments of all time. It was erected in the year 638, upon Mt. Moriah, on the site of the Jewish Temple. Although it was founded by an Arabian calif, to enshrine the rock from which Mahomet started for his flight through the heavens, it has many claims to consideration in a study of Christian architecture ; it would deserve it. if merely for the fact that it was directly inspired by the Con- stantinian temples, and is the only existing monument which gives an adequate impression of the architectural possibilities of this style. Much has been disputed about its origin. De Vogiie's solution is in all probability the correct one, that it is a Byzantine building constructed by Byzantine architects, though for a Mahometan ruler. It is not unlikely that the roof was originally open at the centre — like the church of the Ascension, which in its purpose it so closely resembled, and which it may have followed as closely in pattern. About the THE CENTRAL TYPE — Composite Plan 143 year 690, however, it was provided with a wooden dome of very remarkable construction. In decoration it is only in the dome that Arabic traits appear, the rest is purely Byzantine. In 1099 it came into the hands of the Christians, who founded here the order of Knights Templars. Standing as it did upon the platform of the Temple, it was popularly believed to be actually a part of the ancient Temple of the Jews, and it fig- ures as such in the Sposalizio of Raphael. Having been copied in the first place from the churches of Constantine, it became in turn the pattern for the churches built by the Templars in Europe. We have record of but one other round church in the neigh- borhood of Jerusalem ; it was located in the Valley of Jehosha- phat, and commemorated the spot where the Virgin Mary remained between her death and her assumption. It is men- tioned as early as the beginning of the fifth century. Of the imitations which were made in early times of this type of memorial building, few have been preserved to us, or even recorded. Beyond what has already been said in regard to S. Stefano Rotondo (Fig. 44, V), and S. Angelo of Perugia, it need only be added that as these buildings were intended to be used as houses of congregational worship they were constructed with a clearstory and completely roofed. We learn from the Life of S. Willebald 1 that the church in which the Council of Niceea was held was similar to the church of the Ascension on the Mount of Olives. With the second type of round building, which we come now to consider, we return again to the cupola and vault construc- tion, and we approach, therefore, more closely the line of de- velopment of Byzantine architecture. This type also emerges with the very beginnings of Christian architecture under Con- stantine. It seems, however, that it is to be referred directly to more ancient models ; for though no classic buildings of the sort have been preserved (except probably the baptistery at Aix), the antiquarians of the Renascence saw and described similar temples in the neighborhood of Rome. Constantine erected two such monuments at Rome (Fig. 44, d and e) : one (S. Costanza), the mausoleum of his sister Constantia; the other, the baptistery of the Lateran. The scheme of both is essen- lActa SS. 0. S. B., saec. 3, P. II., p. 379. 144 ARCHITECTURE tially the same, a central room formed by a circular or polyg- onal colonnade which supports a dome, and a surrounding corridor covered by a vault which serves as a buttress. The cupola of S. Costanza is supported by a circular colonnade composed of twelve pairs of columns; within this circle stood the huge sarcophagus of porphyry which is now in the Vatican Gallery. Both the cupola and the surrounding ring-vault were originally covered with mosaics (Figs. 125, 1 L'li), but only those of the vault have been preserved. The thick outer wall is enlivened within by niches, and without it was originally sur- rounded by an open colonnade. The baptistery is smaller and its construction lighter; both the colonnade and the external wall are octagonal, — the character of the cupola was altered in later times. This early building lie- came the normal type fur Christian bap- tisteries, and it remained so as long as a separate building was required for this It was obviously more appropriate to its purpose than was the round build- ■- ' i. ing of simple plan, for je, Ezra, it not only provided a larger room, but archi- tecturally separated the font from the assembly room. These early buildings are obviously the direct prototype of a considerable class of round churches; for example, S. George at Ezra in Syria (Figs. 44/, 40), built in 512, and the similar and contemporary church at Bosra. In the ease of these Syrian churches the exigencies of stone construction account for tin' almost pointed profile of the dome. It can hardly be doubted that the typical Byzantine scheme (S. Vitale) was also strongly influenced by this example, although one of its features — the semicircular recesses and storied galleries — is to be referred more ex- pressly to the development of the niche decoration which we have already studied as a characteristic of the round build- ing of sini pie plan. It is certain at all events that we have now before us all of the antique monuments which could have contributed to the THE CENTRAL TYPE— Composite Plan 145 development of S. Vitale, — under the influence, it is always necessary to remember, of distinctive Byzantine principles of construction. S. Vitale in Ravenna was erected during the reign of Jus- tinian (begun in 526 and completed in 547). It would lead us too far to give a description of the building, beyond the few words which are necessary to render intelligible the plan and the interior view which are given in Figs. 47 and 48. The eight great pillars of the central room support an octagonal drum which is surmounted by the dome. The construction of the dome is peculiar : its core consists of terra-cotta pipes, fitted a b Fig. 47. — a, S. Vitale, Ravenna, b, S3. Sergius and Bacchus, Constantinople. the one into the other, and extended spirally from the bottom to the top, — a construction which diminishes not only the weight but the thrust. The arched window openings in the drum serve to concentrate the weight upon the pillars, which are supported by buttresses extending beyond the surrounding corridor. The corridor is divided into an upper and a lower gallery (both of them vaulted), and the buttresses are pierced above and below by arches to permit free passage around the galleries. Seven niches surmounted by a half-dome intersect both the upper and the lower galleries ; but as the walls of the niches are replaced by columns they permit free prospect from the galleries into the central room. Finally, the presbyterium and apse breaks the symmetry and completely interrupts the 146 ARCHITECTURE galleries, extending with its adjoining rooms beyond the oc- tagonal perimeter. The scheme, which is carefully and conse- quently thought out, is elaborate without being complicated ; the constructive purpose of the several parts is obvious, and the effect is therefore restful and harmonious. Essentially the same constructive scheme is repeated in SS. Sergius and Bacchus at Constantinople (Fig. 47), which was built about the same time. A very different effect is pro- duced, however, by a change in a feature which is almost purely decorative, name- ly, in the niches. The niche oppo- site the apse and the central niche on either side of this line were suppressed by simply carrying the row of col- umns straight across from pillar to pillar, with the consequence of giving the room a distinct longi- tudinal direction, and of permitting the inclusion of the whole room within a quadrangular wall. This shows the strong tendency in the Byzantine architecture of that time to accommodate the dome to the longitudinal room which was required by the ritual. S. Vitale, however, represents the consummation of the round type of church ; and before we follow the Byzantine develop- ment upon other lines it is necessary to return again to the earliest period of Christian architecture and trace the influence of the cross-shaped plan. Fig. 48. — Interior of S. Vitale, Ravenna. Sixth century. THE CENTRAL TYPE — Cross-shaped Plan 147 THE DOME ON A SQUARE BASE CROSS-SHAPED PLAN Like the so-called transept of the basilica, the earliest exam- ples in Christian architecture of a cross-shaped plan are to be referred to pagan prototypes, though in both cases the Chris- tian symbolism furnished the suggestion which led to further development. Cross-shaped chambers were sufficiently com- mon in Roman architecture ; in the early Christian period they were commonly used as mausoleums or memorial chapels. There were two such rooms connected with the Lateran bap- tistery (one visible in Fig. 44 e) and a considerable number among the chapels which lined the sides of S. Peter's (Fig. 29). The type, however, which is most characteristic of Christian architecture consists in an oblong room terminating in a group of three apses united by a small dome. This was the usual type for the memorial chapels which were erected above the cemeteries. Two examples are preserved above the catacombs of Callistus, and one is attached to the small basilica of S. Sinforosa (Fig. 28, c). This so-called apsis trichora was employed by Paulinus for the presbyterium of the basilica which he built at Nola, in honor of S. Felix. This use, how- ever, was exceptional, and it is not possible to say what influ- ence this traditional form may have had in stimulating the development of a more distinctly marked cross plan. There is at all events no doubt that this development was due chiefly to the symbolical interest of the form, and that it received a marked impulse from the discovery and veneration of the wood of the cross. In the case of large buildings, the application of a distinctly cross-shaped plan was practically so inconvenient that it could be due only to a symbolic motive. This scheme — two long and narrow halls intersecting one another at right angles — was of all central plans the most unsuitable to the Christian ritual ; and such churches were, as a matter of fact, designed expressly as mausoleums or me- morial. One of the greatest examples, and the best preserved, is the shrine (Kalat-Seman), which was erected around the col- umn of S. Simon Stylites in Central Syria. It must have been built soon after 450. As may be seen from' the plan 148 ARCHITECTURE (Fig. 63), it consists of four great basilicas grouped in the shape of a cross around a central court, in the midst of which stands the column. The central court was uncovered, and there was, therefore, no attempt at bringing the four separate halls into a relation of close unity. Most, if not all, the other churches of this type must, how- ever, have been covered with a roof or dome at the intersec- tion. Beyond the fact that it had the form of the Greek Fig. 49. — Exterior. Tbe mausoleum of Gallfl cross, nothing very clear is known about the church of the Apostles, which Constantine built at Constantinople as a mau- soleum for himself and his family. It is probable that S. Ambrose took this as the model for the church which he built in Milan in 382 (originally dedicated to the Apostles, after- ward to S. Nazavius) ; and if so, it is only another instance of the influence which famous memorial churches had upon the congregational church, in spite of their unsuitableness for this use. S. Nazario Grande at Milan was entirely rebuilt in the eleventh century, though it retains, with modifications, the cross form. S. Nazario e Celso at Ravenna, the mausoleum THE CENTRAL TYPE — Cross-shaped Plan 149 of Galla Placidia (Figs. 49, 50), though of much smaller di- mensions, is of the same type as the great church of S. Ambrose. The four vaulted arms are connected at the centre by a dome. The construction is poor, but the mosaic decora- tion is of the richest. This chapel originally stood in connec- tion with a greater church — likewise cross-shaped — which was dedicated to the Holy Cross. S. Arculphus gives a plan of a cross-shaped church surrounding the well of Jacob in Fig. 50. — Interior. Placidia, Ravenna. Fifth century. Samaria. Finally, it may be mentioned that there are several baptisteries of this shape. The Greek cross as it appears in Byzantine architecture takes a very different form : it is much less accentuated — that is, the four arms have less longitudinal extension in propor- tion to their breadth — and the angles which remain are par- tially or altogether absorbed by massively constructed rooms which serve as buttresses to the four central pillars. Within and without, the building was thus compacted into the unity of a single hall. Different as this scheme is, it was not only potently influ- 150 ARCHITECTURE enced, but essentially conditioned, by the less organic type which, has been described above. An ostensible proof of this influence lies in the fact that Justinian, in rebuilding the Con- stantinian church of the Apostles, in Byzantine style, adhered to the form of the cross plan. 1 But before considering the forms which the Byzantine development actually took, it is necessary to note the elements which the cross form of itself essentially contributed to the solution of the problem presented by the use of the dome as the prominent feature of church architecture, — that is, the problem of harmonizing the vertical axis emphasized by the dome, with the horizontal axis required by the ritual. To show how strongly this necessity was felt, one need only refer to the church of Irene at Constantinople, which reveals an exclusive preoccupation with the longitudinal direction : the side rooms are reduced to the width which was absolutely required by the buttressing system ; the middle room is roofed by two domes, two broad vaults, and the half-dome of the apse, — with a result which is highly unsatisfactory, since the cen- tral disposition is completely lost, the vertical axes of the two domes are entirely ignored, and yet the longitudinal direction which is strongly marked in the plan encounters an obstacle in the ceiling. The use of the cross form immediately introduced a factor which was of great advantage from the point of view of the ritual, namely, a square central room. The square room was practically attained in SS. Sergius and Bacchus as a develop- ment from S. Vitale. But the cross form, by reducing the supports of the dome to four pillars, opened the central room to the freest communication on four sides ; and the four arms of the cross which were thus brought into the strictest unity with the dome balanced the importance of its vertical axis by two strongly marked horizontal axes which were in perfect harmony with it. That is to say, the eurythmic symmetry de- manded by the central disposition was maintained, and at the same time the horizontal direction required by the ritual was clearly indicated. It may be said that this solution provided too much, since the transverse axis was not only not required by the ritual, but was in some measure discordant with it. 1 Procopius : de xdilic. Just. I. 4. THE CENTRAL TYPE — Cross-shaped Plan 151 But the evil had to be accepted with the good, and although a compromise was possible by way of accentuating the impor- tance of the longitudinal nave, the Byzantine architects kept this well within bounds, and rarely gave to the nave a length which was in marked disproportion to the transept. The church of the Apostles which Justinian built to replace the earlier edifice of Constantine has disappeared like its prede- cessor. Procopius informs us that the longitudinal nave was greater than the two arms, " as the pattern of the cross required." He also informs us that the altar room and presbyterium were located at the intersection of the cross ; and this is the only case in which we have any hint of such an adaptation of the ritual to the obvious requirement of the centralized plan. The long description which Procopius gives fails to furnish all the data we need for the reconstruction of the church ; but Choisy is confident that we have close copies of it in two famous churches of the eleventh and twelfth centuries respectively, — S. Mark's at Venice and S. Front at Perigueux. Whether or not it may be referred to the age of Justinian, this type deserves notice as one of the most successful creations of Byzantine architec- ture. Its main characteristics can be briefly described. It is a system of five equal domes. The cross plan is strictly adhered to, but not accentuated. The building is distributed by heavy piers into five square rooms of equal dimensions, one in the centre and four constituting the arms of the cross. Each is surmounted by a dome, and there are thus three domes on each axis, preserving the symmetry and marking clearly the cen- tral disposition. The longitudinal nave is accentuated chiefly by the addition of the apse and the narthex. It may be pre- sumed that in the church of Justinian there were galleries ; and though galleries were not used in the West, S. Mark's has retained in their place, as a purely traditional and decorative feature, a narrow passageway supported by colonnades. The galleries were entirely dropped in S. Front, which was copied directly, not from the church of Justinian, but from S. Mark's. The system of five domes was applicable only to very great churches. As most of the Byzantine churches of the Middle Ages were of exceedingly minute dimensions, they adopted a plan which was evidently derived from this, though it has only one dome. It is the type which is represented by the great 152 ABCH1TEGTURE majority of Oriental churches throughout the Middle Ages. It was constituted by merely dispensing with so much of the cross arms as was covered by the four domes. This reduced all four arms to about the dimensions of the vaults which had separated the domes one from the other. By this change the breadth of the arms was made to exceed their length, and they were distinctly subordinated to the greater central room. The four piers of the dome, though they had a diminished weight to support, retained their previous dimensions, and completely filled out the small angular spaces of the cross plan, so that the plan externally became rectangular. Even in the original construction, the pillars had each been separated into four shafts, and now, with a lighter weight to support, the shaft facing the central room was often replaced by a column, by which means four small corner rooms were practically added to the hall, — or, what was of more importance, the short cross arms were put into closer relation with the nave and altar room. These four rooms were covered with spherical vaults, which were later (in the tenth century) developed into as many cupolas, one at each angle of the building. It became the fashion also to crown the narthex with three cupolas. It is this feature, namely the great number of cupolas, which gives to late Byzantine churches the external effect with which all are familiar. It may be remarked that the longitudinal direction was generally emphasized by the extension of the presbyterium ; this, however, was flanked by the two sacristies, so that all but the apse was included in the rectangular plan. The nave also was often emphasized, at the expense of the cross plan and of fundamental principles of symmetry, by rows of columns which separated it from the arms. This was the only motive borrowed from S. Sophia, and it was adopted with but little propriety. But this has already led us far be- yond the limits of our period. It is a digression which proved especially tempting, because, upon the basis of the sixth-cen- tury development, it was possible in so few words to describe a type of architecture which has characterized so long a period of history, and characterizes to-day most of the churches of the East. THE CENTBAL TYPE— Oblong Plan 153 OBLONG PLAN Hagia Sophia, " the Holy Wisdom," was built by Justinian in 537, upon the site of an ancient church of the same name which probably dated back to the time of Constantine. The names of the architects, Anthemius of Tralles and Isidor of Miletus, deserve to be remembered. SS. Sergius and Bacchus (Fig. 47) is usually taken to be the immediate prototype, or rather the next preceding step in the development. One can certainly note here a resemblance which is far closer than can be accounted for merely by the common principles of Byzan- tine construction. The resemblance is specially marked in the main room by the use of semicircular niches in combination with the straight colonnades of the side walls (Figs. 51, 52). The room under the dome may be said to be merely an elongation of the central room of SS. Sergius and Bacchus : that is approx- imately a square, this an oblong rectangle twice as long as it is broad. It becomes clear how essential this difference is when we observe that in S. Sophia the main room has ceased to be a central room surrounded by a corridor, and become in- stead a middle nave flanked by two aisles. In both cases the room was surrounded by eight piers ; but in S. Sophia the dome itself was supported by only four of them. This tetra- style arrangement permitted a resort to the cross plan ; and in fact the buttressing system provided on either side of the dome the rooms which might serve as the arms. This solution, however, was entirely ignored, and the side arms were archi- tecturally shut off from the nave by a two-storied colonnade on each side. With this the central disposition and all effort after eurythmic symmetry were completely given up, notwith- standing that the plan of the building, exclusive of the narthex, was almost exactly square. Dehio compares S. Sophia with the basilica of Maxentius, which represents the type of vaulted construction commonly used in the great Roman baths and in this instance applied to the civil basilica. The comparison of the plans (Fig. 52) is indeed sufficient to prove that, however important were the elements derived from contemporary Byzantine construction, the express aim of the architects was to adapt the essential fea- ture of Byzantine architecture — the dome — to the funda- 154 ARCHITECTURE mental scheme of the basilica. That the model which the architects had in mind was expressly the Christian basilica is shown also by the square atrium in front of the church. It has already been said that the atrium was rarely used in the East, and never in connection with the central type. This combination of the two current types of church architecture resulted in a type which differed equally from both ; and it was accomplished by a device so thoroughly original that it Fig. 51. — Interior of S. Sophia, Constantinople. Sixth century. becomes superfluous to trace more minutely its relations with other types and earlier monuments. With S. Sophia we are, strictly speaking, dealing no longer with the central type of architecture, though it is obvious why the subject must be treated under this head. In thus employ- ing the dome to roof a hall of distinctly oblong proportions, its centralizing influence was not ignored, — it was modified. It was modified by the one device which is conceivably applicable to the case : by supplementing the dome with two half-domes of equal aperture at each end of the room. The two half-domes THE CENTRAL TYPE—Oblonq Plan 155 were in a measure merged with the central dome by the spheri- cal surfaces of the pendentives, and the effect of the whole ceiling was practically that of a long elliptical dome : that is to say, it no longer demanded an eurythmie symmetry; it de- manded, on the contrary, a predominant emphasis upon the horizontal axis which corresponded with the dimensions of the room and comported with the ritual. The combination of these various spherical surfaces in the ceiling required, however, a corresponding variety in the plan of the room, and this was furnished by the four niches and the apse. It is evident that Fig. 5'J. — Plans of S. Sophia and of the basilica of Maxrntius. a perfectly regular plan, either rectangular or elliptical, would not have harmonized so well with the ceiling. The consistent execution of the programme required that the dome should rest immediately upon the pendentives without the intervention of a drum, and that it should lie constructed as flat as possible ; it proved in fact too flat for stability. The buttressing system was very elaborately thought out, and with one exception it was thoroughly adequate. The half- domes were each supported by four piers, two of which they shared in common with the dome. They were amply buttressed at one end by two niches and the apse ; at the other by the corresponding pair of niches and a double narthex. The 156 ARCHITECTURE half-domes in their turn constituted a most effectual support upon two sides of the dome. The pendentives transmitted the weight and part of the thrust of the dome to four mighty piers which received the support of lateral buttresses of the full width of the aisle. The one weakness lay in the fact that a part of the lateral thrust of the dome had to be sustained by the great arches which supported it on either side, and they could oppose to it no resistance except such as was due to the comparatively slight thickness (five yards) which comported with their character. Twenty-two years after its building, a part of the dome fell during an earthquake. The dome was then immediately repaired, and from that time on the church has undergone frequent restorations, which, while leaving the nave substantially unchanged, have nearly transformed the exterior. The necessities of the case required a breach of one of the principles of Byzantine architecture, the erection of four huge external contreforts which rose high above the aisles, and were applied directly to the lateral support of the great arches. The construction of S. Sophia was> not only original, but bold. It was the first time so great a dome had been erected upon four piers ; and the chief structural peculiarities of the building were not only without example in the past, but have remained without imitation since. Notwithstanding the resto- rations which it required, the fact that the building has en- dured for more than thirteen centuries is proof that the work was feasibly and strongly conceived. It was undoubtedly the culminating triumph of the Classical age of Byzantine archi- tecture, though other types, the by-products of the same age, were found more available as patterns for a later time and for a decaying empire. Originally, the decoration and furnishings of S. Sophia were the most splendid that have ever been lavished upon a Chris- tian church, and the effect of the building doubtless justified the enthusiasm of its panegyrists. The golden altar, the pres- bytery, and the choir are now replaced by the inferior glories of the Moslem mosque ; even the mosaics, so far as they repre- sented distinctively Christian subjects, have been for the most part either destroyed or covered up. What remains, however, suffices for the imagination ; and this is fortunately a type of THE CENTRAL TYPE— Oblong Plan 157 building which does not, like the basilica, depend for its effect chiefly upon its decoration. Upon entering the building, the sombre light of the double vestibule serves to enhance the luminous splendors of the vast nave. Thanks to the disposition of the half-dome which spreads upward from above the entrance, the eye discovers at once the great cupola in its whole extent. At the same time the entire system of the vaults of the nave — the various Fig. 53. — Mausoleum of Theodoric, Kavenna. . Sixth century. spherical surfaces so apt for mosaic decoration — -is frankly exposed to view. In the prospect of the nave the double colonnades on either side serve to give something of the appearance and effect of the basilica. The numerous details fall naturally into three great divisions which are framed by the arches above. For all the elaboration, the governing lines produce an impression of simplicity, and the details are re- quired to enhance one's appreciation of the vast size. The great dome has an aperture of nearly one hundred feet, the dimension of the half-domes and arches is the same; yet without the lateral colonnades nothing would indicate the 158 ARCHITECTURE extraordinary sweep of the arches above them; they are needed to give the scale and to spare S. Sophia the question- able praise which has been given S. Peter's, — that there is nothing to intimate that it is huge. The illumination is unusually generous ; windows are pierced freely, not only in the high side walls, but in the half-domes, in the apse, and in the niches, while the crown of windows which illuminates the whole circumference of the dome seems to isolate it, as though it were suspended in the air. The four pillars which support the pendentives are hid behind the lateral galleries ; one sees no more of them than the angle, but that suffices to reveal the presence of the mass of the buttress and to assure the eye. The work astonishes, yet at the first glance it explains itself. Never has stability and daring, the eclat of color and purity of lines, never has the genius of Rome and that of the Orient, been associated in a more astonishing and a more harmonious whole. The preceding paragraph follows the language of Choisy almost closely enough to deserve inverted commas. I will here quote from him still more closely a sentence which applies to Byzantine architecture in general. After speaking of the simple geometrical relations which are to be traced in the monuments of this art, he adds 1 : "It is not only the feel- ing of unity which one experiences at the view of a Byzantine interior, but also a sort of tranquillity and calm which is nothing else than the full satisfaction of the mind before a work where all the combinations of equilibrium are clearly apparent. The buildings of our Gothic architecture awaken a sort of inquietude and uneasiness which is due above all to the fact that the buttressing organs are projected outside ; from within one does not at first view take account of the equilib- rium. Quite different is the effect of the Byzantine construc- tions : the eye embraces at one glance the vault which covers the building and the contreforts which support it ; it sees nothing which does not explain itself; it is the clearness of the art of Greece itself." i Hist, del' Arch. II., p. 34. FURNITURE — The Altar 150 C. FURNITURE OF THE CHURCH ALTAR AND CONFESSIO The preceding sections have several times furnished occa- sion to remark upon the central importance of the altar for the architecture as well as for the ritual of the church. In itself nothing could be more simple than this important piece of furniture. It was, as it was also commonly called, a table. In shape it was copied after patterns which were familiar in secular use ; it consisted of a rectangular and only slightly oblong top supported by one, four, or occasionally five, legs. The Holy Table of the Church had as little resemblance to a pagan altar as the basilica had to a temple, or the statue of the Good Shepherd to an idol. We can well understand how the Romans could make and the Christians admit the reproach that they had ''no altars, no temples, no images." This simple and natural shape the altar retained universally till well into the fifth century. The early altar was not so high and by no means so long as has been common in the West since the Middle Ages. From the beginning, stone as well as wooden altars were employed ; the one stood in relation wdth the tables of the triclinium, the other with the ornamental stone table of the atrium. The use of wooden altars is proved by early texts, and it is due only to the perishableness of the material that none have 160 ARCHITECTURE survived. A wooden chest, decorated merely with a single cross upon the front, is venerated in the Lateran as the altar of S. Peter ; but it can date no earlier than the age when the altar began to be used as a holder of relics. It is not men- tioned before the eleventh century. Gradually the Church came to discriminate in favor of altars of stone; this was very obviously on account of their more monumental character and greater durability. An edict pur- porting to be of Sylvester I. which forbids the use of wooden altars can hardly be regarded as genuine, since it rests solely upon the testimony of the Breviarium Romanum (eleventh cen- tury). Such a prohibition is, however, proved for Gaul at least by the twenty-sixth canon of the Council of Epaon (517) ; but the rule must have suffered occasional exceptions even after this period. Thanks to the durableness of the material, several stone altars of the fifth century have been preserved, at least in part. The altar illustrated in Fig. 54 is from South- ern France (S. Que"nin, now at Vaison) ; it is ornamented with the eucharistic vine, while in the middle are two doves which approach the monogram of Christ. A similar altar plate from S. Marcel (Ardeche), now in the museum of S. Germain, rep- resents six sheep on each side departing from the towns of Bethlehem and Jerusalem and approaching the monogram. An altar from Auriol in the south of France has on the front twelve doves and the monogram, and on the sides tendrils of the vine. This table is noteworthy for the fact that it is sup- ported by a single stout pedestal. An altar which Galla Placidia found in S. Giovanni Evangelista in Ravenna was supported upon five legs, which were later supplanted by plates of stone when the altar assumed the form of a chest; the original legs have been in part preserved. Likewise of the fifth century, and perhaps earlier than any of these examples, are the mosaics of the cupola of the Orthodox Baptistery in Eavenna (S. Giovanni in fonte). They represent a stone altar of the simple table form supported upon four round legs or columns. Four times the design is repeated around the base of the dome, and on each altar rests an open book bearing the name of one of the Gospels. The custom of leaving the Gos- pels upon the altar finds witness elsewhere. The change of form which the altar underwent in the sixth FUBNITUBE— The Altar 161 century was due exclusively to the cult of relics, or rather to the new form that cult then took of enclosing the relics within the altar. In treating of the catacombs it has been shown that the Eucharist was early brought into close relation with the tombs of the martyrs ; that it was celebrated either in the martyr's crypt and upon his tomb, or in a chapel directly above it. The great basilicas which Constantine and later emperors built above the tombs of the Apostles and martyrs were in thorough keeping with early traditions, and they were repeated everywhere throughout the Christian world. It was accounted a matter of prime importance that the altar be brought into Fig. 55. — Altar with confessio, S. Giorgio in Velabro, Rome. the closest possible relation with the tomb, or be located at least directly above it. To accomplish this end, hillsides were excavated and whole galleries of the catacombs were cut away. Many tombs had to be destroyed in reaching the grave of the martyr ; but his resting-place was scrupulously respected, and except in the rarest cases the body was not moved even the few feet which might be required to bring it into the desired relation to the church, all considerations of convenience were subordinated to the aim of bringing the church to the body. In the most favorable case the floor of the presbyterium was brought to a level with the top of the tomb (sarcophagus, loculus, cubiculum, or whatever it might be) ; and since the altar stood usually upon the edge of a plat- 102 ARCHITECTURE form which was raised by a few steps above the floor of the church, an opening in the wall of the platform beneath the altar afforded a view of the relics. The chamber surrounding the tomb, and the shaft or gallery through which it was put in relation with the altar, is called the confessio. No part of the ancient churches has suffered more than the confessio from the pious zeal of reconstructors. Though no early example has remained unaltered, there are several medi- eval constructions, like S. Giorgio in Velabro (Fig. 55), which Fio. 56. — Altar with fenest exhibit substantially the early scheme. The opening into the confessio was closed by a grating of stone or metal — the fenes- tella coi\fe.ssio}tis — which allowed the worshippers a glimpse of the tomb. It is evident that this arrangement required no change in the table form of the altar. In the case, however, that tin' presbyterium was not raised above the level of the church, the fenestella had sometimes to be inserted between the legs of the altar, as in S. Adessandro at Rome (Fig. 56). This example furnishes an excellent illustration of a typical fenestella of an early and Classical pattern. The fenestella was provided with an orifice large enough to admit an arm ; FURNITURE— Ttie Altar 163 small objects, particularly handkerchiefs, introduced here and brought in contact with the sacred tomb acquired the charac- ter of relics (brandea, palliola). Any mention of the exporta- tion of relics during the first five centuries must be understood to refer exclusively to such as these. This orifice was closed by costly doors — often of silver or gold — which were secured by a lock. The key itself was sometimes given as a relic ; and de Rossi calls attention to the fact that in Italy keys are still given to children as amulets. The bodies of both of the Apostles S. Peter and S. Paul lay deep beneath the floors of their respective basilicas, and were brought into relation with the altar by a vertical shaft similar to the luminaria of the catacombs. In the case of S. Peter's the shaft was twice interrupted by perforated plates, which were extended across it, — here called cataractoz. Objects de- posited upon the lower plate gained extraordinary value as relics. This is proved by the request of Justinian. He had first begged portions of the bodies of the two Apostles for a church which he was building in their honor at Constantinople. When this request had been denied him by Pope Hormisdas, on the ground that the custom of separating the bodies of the saints was one unknown to the Romans, he asked that his relics might if it were possible be placed upon the lower plate (ad cataractam, secundum). Constantine had surrounded the tombs of both the Apostles with cubical cases of wood covered with gold crosses in the form of the monogram of Christ, and since that time they had been rendered entirely inaccessible. Constantine himself, however, followed a different plan in the construction of the confessio of S. Lawrence. In this case likewise the tomb and its cubiculum lay deep below the floor of the basilica, but it was rendered accessible by two stairways (gradus ascensionis et descensionis), which were doubtless entered at the corners of the apse. The grave itself Constantine sur- rounded with an apse, ornamented below with porphyry and covered above, in the half-dome, with silver; silver chancels enclosed it, and before it burned a gold lamp of ten wicks, a silver corona with fifty dolphins (lamps), and two bronze candela- bra. In front of the tomb of the martyr was the instrument of his passion, a gridiron plated with silver. This confessio is already in this text called a crypt, but it was provided with 164 ARCHITECTURE no altar besides that of the basilica above, nor did it serve as a model for later crypt constructions. On the contrary, the arrangement accorded so little with the early ideals that in the reconstruction of the basilica (probably by Leopardus about the end of the fourth century) the floor was sunk in order to bring the altar into closer relation with the grave. It has been explained (p. 50) how the body of S. Valentine was made accessible by a gallery constructed in imitation of those of the catacombs. The type of gallery or crypt which was typical of later constructions is represented in the. church of S. Pancrazio near Rome, and was probably due to Pope Honorius I. (625-638). It is a semicircular passage, entered at the corners of the apse and following its foundation wall ; from the middle of this half-circle a straight horizontal gallery leads to the tomb. There are several examples of this type at Ravenna, and elsewhere in early churches, but it is doubtful whether they antedate the instance mentioned above. From this it is evidently but a short step to the excavation of a single large chamber under the presbyterium — the type of crypt which became characteristic of the Romanesque churches. Such a crypt often appears in mediaeval reconstructions or imi- tations of ancient basilicas, with the consequence that the presbyterium is raised high above the nave. In this connection may be mentioned a custom which occa- sionally introduced a novelty in the construction of the apse. It sometimes happened that a memorial basilica could not be built immediately above the tomb of the martyr, because the ground was already occupied by a venerable chapel which the builder scrupled to destroy. In this case the apse of the basilica was constructed upon a tangent to the apse of the earlier shrine,, and a window was opened at the point where they met. A perforated stone window plate (transenna) here took the place of the fenestella, and enabled the worshippers in the greater church to glance beyond the altar and enjoy a glimpse of the martyr's chapel. An instance of this arrange- ment is the basilica of S. Sinforosa near Rome (Fig. 28, c). Paulinus of Nola elaborated this idea in the construction of the new basilica of S. Felix. He opened arcades of three arches in the apses of both basilicas, leaving between them a covered court which served as entrance hall to the buildings. FURNITURE— The Altar 165 The arcades were closed by transennae, but Paulinus claimed that through them the worshippers in the new church could participate in the mass which was celebrated over the tomb of the martyr. This example seems to have led to numerous imitations. We have a contemporary instance of this con- struction in the recently discovered apse of the Basilica Severiana in Naples (Fig. 40). It seems to have been employed also with a merely decorative purpose in cases where there was no question of a martyr's tomb. S. Maria Maggiore, as reconstructed by Sixtus III. (432-440), seems to have had an apse which was put in communication with a surrounding gallery by means of such an arcade. At least this seems necessary to explain a curious reference which the Liber Pontificalis makes to this church in the life of Paschal I., where it is said that at the celebration of the mass the women stood behind the seat of the pontiff, so that he could say nothing to his assistants without being overheard by them. The room behind the apse seems in this case to have been used as a matroneum. An arcaded apse of this sort is illus- trated by a bronze lamp of the fifth century (Fig. 153.) It was naturally about the confessio and the altar that the costliest decoration was expended. There is frequent mention in the Liber Pontificalis of gifts of silver for the decoration of a confessio. Sixtus III. (432-440) persuaded the Emperor Valentinian to place as a votive offering above the confessio of S. Peter "a golden image with twelve doors and twelve Apostles and the Saviour adorned with precious gems." We have to understand that this was designed in relief with the Apostles and the Saviour framed in as many arcaded niches. The design was common upon the sarcophagi. Symmachus (498-514) placed the same design in silver above the confessio of S. Paul. Precisely where such a relief might be placed is not clear from the mere notice that it was above the confessio ; it may have been attached to the perpendicular wall below the altar ; but it is equally likely that it was in front of the altar and between its legs, constituting the earliest example of an antipendium. Less is said than one might expect about costly materials being employed for the altar. Nothing is said about the great altar of the Lateran, though seven silver altars of prothesis are mentioned. It may be that reverence 166 ARCHITECTURE for the traditional form of the altar served to perpetuate the use of stone, and the use of altar cloths had doubtless some- thing to do with it. However, altars of precious metal are occasionally mentioned : for example, the altar which Con- stantine gave to S. Peter's was of silver plated with gold and adorned with four hundred gems, weighing altogether 350 pounds. Justinian gave a gold altar to the church of S. Sophia. Whatever influence the early confessio may have had upon the shape of the altar, it was confined to the memorial churches without the city. It will be remembered that within the city limits burial was forbidden, and the congregational churches, the titles, were therefore without the confessio. The preserva- tion of relics of a specially sacred character, such as the wood of the Holy Cross for which Constantine built the basilica of S. Croce, may have required the construction of an imitation confessio. Secondary relics such as have been described above needed no such elaborate care ; at the most they were deposited in a cavity made in the plate of the altar, and they led to no change in its form. The next stage in the development of the altar was due to the custom, which grew rapidly during the sixth century, of translating the bodies of the martyrs and depositing them within the churches of the city ; — except in Rome, where no translations were made until after the middle of the eighth cen- tury. This often led to the construction of a true confessio or crypt ; but more commonly the relics were deposited immedi- ately beneath the plate of the altar. The reconstruction of the altar which this required was at first not very radical. The relics were enclosed within a stone cippus which closely imi- tated the heathen pattern. That is to say, it was a block of stone, roughly cubical in dimension, hollow within, and orna- mented on the front by a doorway, like a miniature tomb. This was placed beneath the altar and constituted its chief support, though as its lateral dimension was smaller than that of the altar it did not entirely do away with the table legs. There are a number of examples of sixth-century altars of this char- acter at Ravenna and elsewhere. In S. Giovanni Evangelista at Ravenna such a confessio was placed under an earlier table altar, and the space which intervened between it and the legs FUBNITUUB—The Ciborium 167 of the table was enclosed with plates of marble. This leads directly to the final step, in which the altar was enclosed on all four sides by plates of stone and became a mere chest for the preservation of relics. It was not till the Middle Ages that it became customary to enclose a sarcophagus, or a fully extended body, within the altar, with a consequent lateral extension which corresponded to the length of the human figure. The custom of erecting more than one altar in the church grew out of the cult of relics. Such altars came into use about the beginning of the fifth century, at first in side chapels, later in the church itself and even in the nave. THE CIBORIUM To protect and dignify the altar (see p. 104) a roof of square plan supported by four columns was often erected over it. The name ciborium — to be derived from ki/Jm/kov a cup, — is explained by the form of roof which was commonly given it in the Orient: a cupola, like an inverted cup (Fig. 144). In the West, however, a steep conical or pyramidal roof seems to have been more common. The four supporting columns were placed at a sufficient distance from the altar to permit free passage around it. They were surmounted either by architraves upon which the roof directly rested, or by arches which were filled out at the corners so as to form a horizontal base for the roof. The character of the arch is illustrated by Fig. 39, but the roof generally rose very much higher than in this example. No ciboriums of the early period have been preserved entire, and they are but scantily illustrated in Christian art. We have evidence enough, however, chiefly in the way of descriptions, to establish securely the general traits as they have been given above. The earliest ciboriums we know date from the ninth century, and they still adhere to the ancient type. It was not till about the twelfth century that any novelty was introduced ; at that time it became the fashion in Italy to support the roof above the architraves upon rows of colonnettes, giving it a much lighter and more graceful appearance than the earlier form could boast. In the early period the material was of wood, stone, bronze, or silver. It must suffice to mention a few of the most notable examples. 168 ARCHITECTURE The Liber Pontificalis describes the eiborium which Constan- tine presented to the Lateran. The roof was of silver and weighed 2025 lbs. It is to be presumed that it was supported upon marble columns, and it is probably to be restored like the eiborium in Fig. 39 ; but the account is engrossed with the decoration and gives no hint of the fundamental form. In front was an image of the Saviour seated, five feet in height and weighing 120 pounds. There were then the twelve Apostles, likewise five feet in height, and weighing 90 pounds each. As symmetry requires, there were doubtless two Apostles at each side of the Saviour, and four upon each of the adjoining sides of the eiborium. On the opposite side, facing the apse, the Saviour was again represented, seated upon a throne, and at each side of him two angels holding rods in their hands. The roof was ceiled with gold. For the lamps which adorned it see p. 351. This work of art was carried away by the hordes of Alaric in 410 ; but Sixtus III. (432-440) persuaded the Emperor Valen- tinian to erect a new eiborium, of which we know only that it was likewise of silver and weighed nearly as much as the old. In the Basilica Ursiana in Ravenna the Bishop Victor (539- 546) replaced an old wooden eiborium by one of silver weighing two thousand pounds, — presumably through the generosity of Justinian. In S. Sophia the columns as well as the roof of the eiborium were of silver; it is described at length and with enthusiasm by Paulus Silentiarius. This work had the fortune to endure till the thirteenth century. THE CHANCELS Cancelhis was the Latin word which was commonly used for the low screens which marked the separation of the presbyterium and choir from the rest of the church. In a later time the name came to be applied to the presbyterium itself — the chan- cel. Very few chancels of the early period have been preserved in place ; we get, however, a clear idea of the normal arrange- ment in S. Clemente at Rome, where the sixth-century screens (Fig. 58) of the choir and presbyterium were simply removed from the lower church and set up in the twelfth-century church above. We get an equally good idea of the arrangement of FURNITURE— Chancels 169 choir and presbyterium in S. Maria in Cosmedin at Rome (eighth century), where the marble chancels which had been employed for the pavement, or for other uses equally foreign to their purpose, have been judiciously restored to their original position (Fig. 38). A glance at the illustration will prove more satisfactory than a long description. In this case it will be noticed that the presbyterium extends some distance into the nave, and a few steps lead from the middle of it into the choir. In other eases, as in S. Olemente, the chancels of the presby- terium coincided with the chord of the apse, and the altar stood upon a line with them ; the approaches had therefore to be -Chancels of the fourth and fifth centuries, reconstructed from fragments in the oratory of Eqllizio and in S. Lorenzo, Rome. constructed on either side of the altar. In the illustration it will be noticed that the chancels of the presbyterium are sur- mounted by a light colonnade for the support of curtains. The choir too was sometimes provided with such a colonnade about its entire circumference. Here, as in S. Clemente, there are two ambons, or pulpits, at the sides of the choir. They were used for the singers and lectors, and later for the preacher ; they differed in dignity and use, as they differed in form. A very great number of ancient chancels, or fragments of them, have lately been brought to light in Rome ; they had been contemptuously employed in the Middle Ages for wall and floor decoration. These fiuds inform us as completely as 170 ARCHITECTURE we could wish about the form of the early Christian chancels, whether they were of stone, wood, or metal. During the fourth and fifth centuries they adhered closely to the patterns which had been long current in Rome, — that is to say, they were usually stiff geometrical designs in openwork, following the technic of metal. Fig. ."i7 shows a reconstruction of two chancels of this period from fragments found at Rome in tin- oratory of Equizio and in the church of S. Lorenzo. From the sixth to the ninth century the chancels received a new Fig. 58. — Chancel In S. Clemente, Kome. Sixth century. type of decoration, and became the chief field for such stone carving as was still practised, — that is, for the conventional designs in low relief which were copied from contemporary textile patterns (see p. 377). No other stone chancels have been preserved ; but wooden ones are also recorded, and so are also chancels of bronze and silver. The most elaborate of all were doubtless the silver chancels which decorated the church of S. Sophia. Constantine employed chancels of gilded bronze in the church of the Apostles at Constantinople. THE COLONNADE IN FRONT OF THE PRESBYTERIUM A row of four or six great columns sometimes adorned the presbyterium. Their object seems to have been similar U that of the ciborium, to add dignity to the chief room of the church. They were not originally intended to mark the limits of the presbyterium, for they did not always coincide with it ; still less were the)- designed for the support of curtains, or in any way to hide the altar from the people. It is probable that the earliest examples were simply decorative, and were not even connected by an architrave. It is from this ornamental FURNITURE— Tlie Iconostasis 171 feature, however, that we have to trace the rood-screen of the Gothic churches on the one hand, and the iconostasis 1 of the Greek on the other. In the round church of the Anastasis at Jerusalem Constan- tine erected twelve columns, corresponding to the number of the Apostles. From the notice that their capitals were sur- mounted by huge silver craters we are led to suppose that they were not connected by an architrave. Where they were placed or how they were arranged — whether in a circle or semicircle — does not appear. We are, on the other hand, very clearly informed about both the character and the arrangement of the six white marble columns which (according to the Liber Pontificalis) Constantine brought from Greece and erected in the church of S. Peter. They are described as columnas vitineas, which alludes to their adornment with traceries of the vine, with birds and flowers — an example of the decadent taste of the age. Notwithstand- ing this account of their origin, there grew up in the Middle Ages a tradition that they were taken from the Temple at Jerusalem. It is mainly due to this tradition that they were preserved in the rebuilding of S. Peter's, some of them are employed to decorate the niches of the piers high up under the dome, and one is venerated in a side chapel where it can con- veniently be studied. It is due to this tradition also that we have the familiar representation of them in Raphael's tapestry which pictures the healing of the impotent man at the door of the Temple. The position which these columns originally occupied in S. Peter's is exhibited in the well-known fresco (probably by Giulio Romano) in the Sala di Constantino in the Vatican. They stood in a row across the front of the presbyterium, at some distance behind the triumphal arch. This was probably their original position ; but it is likely that originally they were not connected by an architrave ; it seems at least as if the free position would comport better with Classical traditions. But at all events they were provided with an architrave as early as the time of Sergius III. (687- 701) ; and Gregory III. (731-741) added another row of six i The iconostasis is a solid screen of considerable height which separates the presbyterium from the nave. It is adorned with sacred pictures and pierced by a central door. 172 AECHITECTUBE columns, fashioned in imitation of them and placed behind them. A single colonnade of six columns is preserved in the cathe- dral of Torcello (Fig. 36) ; in this case, however, the choir as well as the presbyterium is included behind it. The horizontal beam which connected the columns offered a surface which was sometimes decorated like a frieze ; it served as a support for lamps and chandeliers, and, after the seventh century, for large images, — images of silver happen to be mentioned most fre- quently. It is easy to see how the iconostasis was developed from this form. In the illustration above alluded to, it will be noticed that, between the chancels and the architrave, there is but little open space left ; that too, it must be remembered, was often covered by curtains (see p. 378). As the Oriental Church permitted no images carved in the round, it was natu- rally disposed to make the most of the field which the screen offered for pictorial representations. THE CATHEDRA About the seating arrangements of the churches there is very little to be said. There are a few examples of a simple stone bench within the choir; more frequently the seat of the pres- byters has been preserved, and that too was a stone bench con- structed in a semicircle around the wall of the apse. Whatever dignity or adornment it had, it must have received from rich cushions and wall hangings. The presbyters' bench was interrupted in the middle by the cathedra or throne of the bishop. To give the greater distinc- tion to this seat, a niche was sometimes constructed for it in the apsidal wall (as in S. Balbina at Rome), it was sometimes surmounted by a baldachino of stone (as at Grado), and it was generally raised several steps higher than the benches which flanked it on either side. The cathedra itself was a chair of dignity, such as the Roman senators used on public occasions. Many of the earliest examples which we know were actually taken from pagan buildings, or at least bought in the public shops; it is especially the finest and the most graceful of them which are most obviously not of Christian origin. During the first five centuries one type of cathedra seems to have been all FURNITURE— The Cathedra 173 but universal : a solid arin-chair with high rounded back carved out of a single block of marble. The famous statue of S. Hip- polytus furnishes an early example of such a cathedra ; in the illustration (Fig. 118) it is seen, unfortunately, only in face. Later the cathedra was often constructed upon the spot, built up of large blocks of stone. The decoration was generally of a simple character, and rarely of any great interest from the point of view of Christian art. For distinction, as well as for comfort, the cathedra must have relied largely upon its tex- tile coverings, and in particular its cushion and stool. The singular importance of the cathedra lay in the fact, that it was not only the bishop's seat, but also his pulpit. For it was the invariable custom in the Church, as it had been among the Jews, for the preacher or teacher to address the congregation from his seat. It will readily be understood that the traditional position of the cathedra, at the back of the apse, was not well chosen as a point whence the congregation of a great basilica might be addressed. The interposition of the ciborium, or of a colonnade in front of the presbyterium, ren- dered the position still more unsuitable. It is related that S. Chrysostom preached from the lectors' ambon in the nave, in order that he might be heard. It is probable that the bishop often used a portable cathedra which might be placed — at least during the sermon — near the chancels of the presby- terium. Such seems to have been the use of. the ivory cathe- dra (Fig. 114) of Bishop Maximianus of Ravenna (546-556). The fact that it is sculptured behind, as well as before, proves that it was not intended to be set against the wall. The chair which is venerated in Rome as the cathedra of S. Peter is a portable chair of wood (sedia gestatoria), of pagan origin, decorated with ivory reliefs representing mythological subjects. The history of the chair is unknown. It is in a ruinous condition, and was long ago enclosed in a heavy sup- porting frame. The position of the congregation during the sermon was less uniform than that of the Bishop. In prayer they commonly stood with outstretched hands ; kneeling was a more unusual, and distinctly a penitential, attitude. In certain regions, par- ticularly in Africa and in Gaul, it remained the custom, at least as late as the fifth century, to stand during both the sermon 174 ARCHITECTURE and the lections. Tn such cases it is evident that no seats were required for the people. On the other hand, we have proof, as early as the middle of the second century, that it was the more general custom of the Church to sit during the sermon. Seats, therefore, must have been provided for the people. The char- acter of the mosaic floor proves that they were not a stationary part of the church furniture ; but nothing more precise is to be learned about their character. THE AMBON While the bishop preached from the cathedra within the limits of the presbyterium, the lector read the Scriptures from a pulpit erected outside the prcsbyterium ami nearer the people. From the fact that it was ascended by a flight of steps, the pulpit received the name ambon (dra/Wveiv) ; or from its hape and elevation it was , 7" ' :'"■>■., , called "the tower," -rrvp- yos. The ambon stood sometimes on the mid- dle axis of the nave, sometimes to one side of it. Two ainbons were frequently used to dis- tinguish the Gospel from the lections which were taken from the Old Testament and the Epistles. In this case, they flanked the choir or the presbyte- rium on either side, and were distinguished from one another by difference in size and ornamentation. The lectern of the Epistle ambon commonly faced the altar; that of the Gospel faced directly across the nave. The position of the two mil- lions, relative to the altar, was not fixed; the Gospel was read now from the right side, now from the left (see Fig. 3S). The paschal candlestick, which seems to have come into use about the sixth century, stood beside the Gospel ambon. Tn shape the ambon was commonly round or octagonal (Fig. 59) with a solid balustrade, and with an approach of some half-dozen steps upon one or both sides. The double flight of stairs was Fifth century. FURNITURE— of the Baptistery 175 probably required for the singers, some of whom stood upon the ambon or its steps during the chants. At S. Sophia there was a single ambon in the middle of the nave. It was ap- proached from the presbyterium by a raised passage, c™Xeas, across which the clergy could move without hindrance from the crowd. This ambon was so great that the singers found room beneath it ; above, it was crowned by a row of columns bearing lamps. FURNITURE OF THE BAPTISTERY The arrangement of the baptistery requires but brief notice. A flight of steps descended into the round or polygonal font (piscina or fans), which was sunk beneath the level of the floor, and sometimes raised also somewhat above it by a breastwork of stone. The font was surrounded by a row of columns which supported curtains to insure the most perfect privacy and decency during the immersion. The columns were united occasionally by archivolts, more frequently by architraves. The architraves were sometimes adorned by metrical inscrip- tions ; the eight distiches in the Lateran baptistery are as- cribed to Sixtus III. In the Lateran baptistery Constantine constructed a font of porphyry, covering with silver the out- side of the breastwork, and the inside as far as the water line. In the middle of the font he erected a porphyry column sup- porting a golden viol (" where the candle is ") weighing 52 pounds ; there also incense was burnt at Easter time. Upon the rim of the font and pouring water into it was a golden lamb weighing 30 pounds. On the right of this a silver image of the Saviour five feet high and weighing 170 pounds. On the left a similar statue of John the Baptist, bearing the in- scription : BEHOLD THE LAMB OF GOD, BEHOLD HIM WHO TAK- eth away the sins of thb world. There were besides this seven silver harts of 80 pounds each, likewise pouring water into the font. There is mentioned also a censor of gold, weighing 15 pounds, and adorned with 49 emeralds. 1 1 Lib. Pont., — in vit. Sylvestri. 176 ARCHITECTURE D. POSITION AND SURROUNDINGS OP THE CHURCH ORIENTATION The Christians adopted for the orientation of their churches the direction precisely opposite to that of the pagan temples ; that is, it was the apsidal end and not the front which was directed toward the east. From the earliest period of church building there was a disposition to give to the apse an easterly direction ; this direction may even be said to have been pre- scribed, and a rule to this effect actually occurs as early as the middle of the third century in the Apostolic Constitutions. 1 Athanasius regarded it as an Apostolic ordinance that the church must face the east. Until the beginning of the Middle Ages, however, the rule was so far from being scrupu- lously observed that it suffered almost as njany exceptions as it could boast instances. The direction of the street or the nature of the ground often interfered with a strict compliance with this rule, or even completely reversed the direction. The variety in the direction of the churches is specially great in Rome ; nearly every point of the compass is there represented. S. Peter's has the apse on the west end, and it was evidently obliged to take this direction on account of the positiou of the Apostle's tomb with reference to the Vatican Hill. The first basilica built over the tomb of S. Paul had likewise its apse at the west end, and without any apparent reason. The greater basilica which supplanted it in 389 reversed the direction and complied with the rule. The direction of the street may ac- count for the easterly position of the entrance in the case of S. Sebastiano and of SS. Nereo and Achilleo on the via Appia, and of S. Clemente and S. Maria in Dominica within the city. The Lateran basilica has the same direction, though the reason is not obvious. The westerly position of the apse seems to have been deliberately chosen in the case of S. Lorenzo fuori le mura. Some of the most important churches in Rome were directed toward various odd points of the compass, probably on account of the direction of the ancient streets. S. Maria Maggiore, S. Prassede, and S. Prudentiana are directed toward the northwest; S. Marco toward the north; S. Sabina, north i Book II., c. 57. POSITION AND SURROUNDINGS— Orientation 177 east; and S. Agnese and S. Saba, southeast. The great ba- silica at Tyre, built in the beginning of the fourth century, was intentionally directed with its entrance toward the east, and its builder, Paulinus, constructed a high door expressly to admit the rays of the rising sun. The great basilica attached to the Anastasis in Jerusalem had likewise its entrance toward the east. The basilica of S. Felix at Nola had the apse at the west end ; but Paulinus acknowledges that this was contrary to " the more usual custom," and he explains it by the neces- sity of bringing it into relation with the apse of the ancient chapel of the saint. Socrates relates that the chief church at Antioch had " an inverted direction " inasmuch as the altar was at the west end. Turning from the exceptions to the rule, it is to be noted that in Ravenna, and in general wherever Byzantine influence was felt, the easterly direction of the apse was almost as strictly observed as it came to be in the Middle Ages. It was, however, very rarely that the church lay exactly east and west ; there was a declination greater or less to one side or the other. It is a question whether this variety was due simply to accidental causes, the direction of the street, etc., or is to be referred to a symbolical purpose. The attempt has been made to prove the latter opinion by showing that the direction was chosen with reference to the point where the sun rose upon the date of the chief feast of the church in question, or upon one of the turning points of the year. The proof is too ingenious and too uncertain ; but it is hardly open to doubt that it was not the true east but the position of the rising sun which imposed the direction. It has been remarked that in Ravenna, where the apse was at the east end, it was pierced by windows, whereas in Rome, where the chief churches had the apse at the west end, it was without them. The numerous exceptions to the rule of the eastern apse have been taken to prove that the earlier rule in the West required the entrance, not the apse, to face the east, and that the custom which later became universal had its origin in the Orient. But when it is remarked that the most notable in- stances of this supposed Western rule are confined to churches which were built by Constantine or presumably under his in- fluence — the Lateran, S. Peter's, S. Paul's, S. Lorenzo, the 178 ABOHITECTUBE Anastasis, the basilicas in Tyre and Antioch — it seems likely that this easterly direction of the entrance was a particular fancy of that emperor's. That Constantine may have had a strong prejudice upon a subject like this is not improbable, for it is known that even after he was converted to Christianity he retained many traces of his earlier sun worship. ATRIUM AND PERIBOLOS With exception of the memorial shrines which were erected over the tombs of the martyrs, the early Christian churches were almost invariably within the cities. As Christianity was, for a long time, confined almost exclusively to the greater cen- tres of civilization, the churches had, for the most part, to be content with such scanty room as could be provided in densely populated towns. Like the private houses from which they traced their origin, the basilicas were commonly crowded in between surrounding buildings ; or, even if more room was at their disposal, it was filled up with chapels, hospices, schools, and the whole complex of buildings which were required as adjuncts to the church. The effort was made, however, to se- cure a free space in front of the church, a quadrangular court paved with marble and surrounded on all four sides by porti- cos (Figs. 27, e, 29, 30). This was known by various names, in the West most commonly as the atrium or quadriporticus. The atrium was an almost invariable feature of the early basil- icas in the West ; in the East it was less commonly employed, and apparently as an imitation of the Western custom ; after the fifth century it became everywhere less common, and only in rare cases was it imitated in the Middle Ages. The atrium provided a dignified entrance to the basilica, and at the same time protected it from the noise of the public streets. It was used for the instruction of the catechumens, for the feeding of the poor, and, doubtless, for many other of the purposes for which a cloister and court might serve. When burial within the city became customary it was the atrium which first pro- vided room for it. The atrium itself was sometimes com- pletely surrounded by subordinate buildings which the more effectually secluded the church. One may raise the query whether an atrium surrounded by buildings of either secular POSITION AND SURROUNDINGS— Atr 179 or ecclesiastical use would not be an apt provision for the requirements of our modern life in great cities, where the old- fashioned churchyard is out of the question, and the noise of the streets is more than ever discomforting. In the middle of this court there was as a rule a fountain of running water (the cantharus) for the symbolical purification of those who were about to enter the church. The washing was generally confined to the hands. This custom had its suggestion not only in the Old Testament and in the Jewish use, but m a practice which was invariably associated with the Classical temple. The symbolism was in itself so obvious that it hardly depended upon example or suggestion. These early foun- tains with their pure running water were vastly superior, from an aes- thetic as well as an hygienic point of view, to the holy water basins of the Middle Ages. The fountain was generally provided with spring water by an aqueduct. No spring water being available for his basilica at Nola, Paulinus depended upon rain water, which he collected from the roof of the church. He de- scribes this arrangement m one of his poems, and he glories in it as an example of humble reliance upon God's provision. Even in view of a dearth of water, he could console himself with the handsome appearance of the fountain and its ornaments. The fountain must have taken very various forms. The basin was commonly surrounded by a balustrade of sculptured marble ; it was very frequently surmounted by an ornamental roof supported upon columns. The most interesting ancient cantharus of which we have any notice was that which adorned the atrium of S. Peter's. It was erected by Symmachus (498-514) in substitution of an earlier and simpler one. The cantharus of Symmachus endured throughout the Middle Ages, but with the rebuilding of S. Peter's was brutally destroyed. The huge bronze pine cone which formed the centre of the fountain, and two of the bronze Fifi. 60. — Bronze cantharus formerly in the atrium ol'S. Peter's. 180 ABCH1TECTUBE peacocks which adorned the roof, are preserved in one of the courtyards of the Vatican Palace. It is supposed that the pea- cocks were taken from the mausoleum of Hadrian (Castel S. Angelo), and it is not unlikely that the cone too adorned the summit of that monument. Figure 60 gives a rough represen- tation of this cantharus as it was shortly before its destruction. It is from a sketch of one of the earliest of Christian archaeolo- gists, the Dutchman, Philip de Winghe ; valuable as it is, it does not accurately correspond with the descriptions we have of this interesting monument. According to the descriptions, the roof was of gilded bronze ; the peacocks, the four dolphins at the corners, and the great cone were all likewise gilded. Eight porphyry columns supported the roof, and between them were marble chancels, each with two griffins carved upon it. A great fountain of water issued from the cone and played over its surface, falling into a square basin. Where the atrium was lacking, as was generally the case in the Orient, the fountain — and in this case it must have been a smaller one — was placed in the vestibule. Figure 13~> repre- sents close by the church door a small basin upon a high pedestal, which was not unlike the holy water basin of the Middle Ages, except that it was supplied with running water. Very rarely in the West, but more frequently in the Eastern provinces of the Empire, the church stood in the midst of a great court surrounded by a wall — the peribolos. The direct suggestion of the peribolos was the temenos which surrounded the Greek temple. A number of churches which were built upon the sites of ancient temples retained as a decoration the temenos which had surrounded the shrine. In the simplest ease it was a mere wall (Fig. 61), but it was often adorned within by porticos. In any case it characterized only such churches as had considerable pretension to magnificence. A whole complex of church buildings was often included within its circuit ; in many cases the wall may have been needed for defence. One of the most important symbolical motives in church architecture is the broad and inviting doorway. This motive was first adequately developed in Gothic architecture ; in the early period it was only the Central Syrian churches which expressed it in the design of the facade. The motive, however, POSITION AND SURROUNDINGS— -.Prophylaon 181 was by no means ignored by the early Church ; but the pres- ence of an atrium or peribolos required that the symbolical importance of the door be given its chief expression, not at the entrance of the church itself, but of the church precincts. The entrance of the atrium was in fact given very great importance, and like the propylos of the ancient temple, it constituted a building by itself, projecting beyond the walls of the precinct. The high doorway which Paulinus built as an entrance to the peribolos of his church in Tyre was designed not only to admit the rays of the rising sun, but to allow those who passed before it a glimpse of the interior. The pious builder hoped by the striking ele- gance of the door- way to attract the adherents of the old religions and induce them to enter the church. A number of notable doorways are mentioned in early Christian literature, but except for mere vestiges of the foundations none of them have been preserved. The propylos of S. Peter's, as represented in Fig. 30, was a two-storied build- ing with three doors, but it is not possible to decide how much of that structure may have been due to mediaeval reconstruc- tion. It appears, however, that the doorway usually comprised three entrances, as in this case. Fig. 61. — Basilica with peribolos in Kuwena, Syria. 182 AliCIIITECTUBE ADJOINING BUILDINGS The baptistery belonged only to the church of the bishop. It was never placed directly in connection with the church building, though it often adjoined the atrium. It had there- fore a separate entrance and vestibule, and like the church it was frequently surrounded by chapels or oratories. A frequent adjunct of the baptistery was a hall where the candidates Fig. 02. — Complex of church buildings at El Barah, Syria. received confirmation (consignatio) immediately after baptism and before entering the church. It has already been remarked a number of times that the early churches were frequently surrounded by chapels. This is illustrated by the plan and the prospect of S. Peter's (Figs. 29, 30). The chapels on the south of this church opened directly into the side aisle ; those on the north were separated from it by a narrow passage. The chapels or cubicula served two purposes; they were either used as mausoleums before burial in the churches was allowed, or else as places of retire- ment for pious meditation and the reading of the Scriptures. POSITION AXD SUREOl'XDIXGS—TlH Tower 183 An interesting adjunct of the church was the tower. Various were the uses to which the tower was put, and it is uncertain which use may have constituted the original or chief sugges- tion for its employment. It was only in Syria that towers were employed in pairs and brought into architectural relation with the church ( Fig. 32). They were used there for the stair- ways which ascended to the upper galleries of the church ; and they served as places of refuge in regions where there was i 4 |E±5.Jj LJhfc: Flrt. 63. — Memorial church of S. Simon Stvlites, Kalat-Seinan, Syria, danger of sudden attack. In the West only a solitary tower was used, and it was never a part of the architectural scheme. The towers were either square or round in plan, and were di- vided into about a half-dozen low stories. The tower, however, was by no means a general necessity, and the majority of the churches remained without it. It is probable that in the West they were from the beginning employed for hanging hells: but the date of the introduction of church bells is as undeter- mined as that of the construction of towers. The ancient towers which remain in Rome and Ravenna are of mediaeval 184 ARCHITECTURE construction in the upper part, and even the bases give no cer- tain clew to their date. Several of the Syrian sites afford good examples of the com- plex of ecclesiastical buildings which often surrounded the church (Fig. 62). Monasteria seems to have been the name given to small chapels built in the vicinity of a greater church, particularly if they took the form of little basilicas. Next to the chapels the most important adjuncts of the church were the hospices for pilgrims or for the poor, hospitals, schools, baths, and the dwellings of the clergy. The hospice, or xeno- dochiurn, of Pammaehius at Porto (Fig. 27, e) shows an orderly disposition of the buildings which was quite uncommon. It was chiefly in the case of the later monastic settlements in un- cultivated and pagan lands that the whole complex of church buildings could readily be conformed to an ideal plan. The early buildings were, however, often very magnificently con- structed, and they included rooms for a very great variety of purposes. Particularly impressive is the description of Con- stantine's church of the Apostles at Constantinople. 1 1 Eusebius, Vit. Const. IV. 59. IV PICTORIAL ART Early Christian pictorial art is here treated under three chief heads : Painting, Sculpture, and Mosaic. The early Chris- tian painting which we have to consider is limited almost exclu- sively to the frescos of the Roman catacombs (early Bible miniatures are treated apart) ; the sculpture includes numerous sarcophagi, a few rare statues, and some interesting reliefs in wood and ivory; the mosaics par excellence are those which adorn the walls of the churches. One would hardly propose any other division than this, and yet it must be confessed that a classification of pictorial art which is based on the method of execution ceases to have any great importance when, as in the early Christian period, the content or thought of a picture predominates over the interests which attach to mere technic, to the greater or less success of execution, or to the medium even in which the work is wrought. This division, therefore, obvious as it is, cannot be adhered to strictly and exclusively ; it is crossed here and there by a classification according to sub- jects, which requires sometimes the tracing of a single theme through all three branches of art. This familiar and obvious division can, however, be the more reasonably retained, because it corresponds in a measure with three epochs of early Christian art, which were marked by the development of their own peculiar themes. Christian art of the second and third centuries is represented almost solely by the frescos of the catacombs, and the study of early Christian painting is chiefly concerned with this period, notwithstanding the fact that there are numerous catacomb frescos of the fourth century, and occasional examples of the art of every century down to the Middle Ages. The beginnings of Chris- tian sculpture are to be traced back to the third century ; but it was in the fourth century that it received its characteristic 185 186 PICTORIAL ART development, and within the fourth and fifth are included almost all of the monuments which we have to study. Chris- tian mosaic art was in its chief lines already developed in the fourth century ; but it is to the fifth and sixth that the great majority of the monuments belong. The themes which were represented in the mosaics differed substantially from those which were current even in contemporary painting and sculp- ture. For the decoration of the church there was developed a new range of subjects, which were in part symbolical, and yet different from those which originated in sepulchral decoration ; in part historical, and reflecting directly the examples of Scrip- tural illustration which were furnished by the Bible minia- tures. It will be seen, therefore, that in following freely the division which is here chosen, it will be possible to take account of the divisions which refer to time and to those also which are based upon a classification of subjects. There is another difficulty of method, which is due solely to the limited space at our disposal for the treatment of so large a subject. It is not possible here to allude to all even of the important monuments of early Christian art ; it is possible to present only a representative selection of them. There is per- haps sufficient space for a general critique of early Christian art; but with this there would not be space enough for a liberal illustration of it by concrete examples. Inasmuch as the chief claim of archaeology upon popular attention lies in the fact that it furnishes the most concrete material of history, it seems well in this instance to rest satisfied with what has been said in the Introduction in the way of a general account of the conditions and development of early Christian art, and to add here only so much of a theoretical nature as may be necessary for the appreciation of the point of view, and for the intelligent apprehension of the individual monuments. This chapter will therefore be mainly occupied with the task of describing a few of the most notable and representative monu- ments of early Christian art in the spheres of painting, sculp- ture, mosaic, and miniature. It is necessary, however, to explain somewhat at length the character of early Christian symbolism, and to treat with some degree of completeness a few of the principal subjects of early symbolic art — as, for example, the Good Shepherd and the fish. PAINTING — Introductory 187 So far as concerns the frescos of the Roman catacombs, the treatment is of necessity incomplete. The reproductions which have hitherto been made of the paintings of the cata- combs are far from being exact or faithful. The opinions and theories which have been founded upon them reflect the same defects ; much of the instruction which the art of the cata- combs is capable of furnishing has not yet been brought to light, and some of the pictures which are best known are the theme of conflicting theories. Now, however, Monsignor Wilpert has ready for immediate publication a great work, the fruit of long labor, which is at once a complete study of all the paintings of the Roman catacombs in the spirit of the most candid and painstaking research, and a perfect illustration of them by means of the most exact processes of reproduction. This work will itself reveal much that is new ; but, what is of still greater importance, it will put in the hands of all students the material for independent study. While expecting the publication of this book one would indeed be rash to venture beyond the commonplaces of this subject. For the illustra- tions of this section I have relied exclusively upon the pictures which Monsignor Wilpert has already published in several preliminary works. A. EARLY CHRISTIAN PAINTING BEING THE FRESCOS OF THE CATACOMBS, AND EABLT CHRISTIAN SYMBOLISM IN GENERAL Almost all that need be said about the technic of early Chris- tian painting — and indeed of sculpture too — is told in one word, when it is said that it passed by no single step beyond the formulas of Classical art, and that it followed that art in its rapid decline after the second century. Inasmuch, however, as the frescos of the Roman catacombs are almost the only examples preserved to us of the painting of the second and third centuries, it is in them chiefly that we have to trace the successive steps of the decline of this art. It may be that we derive from the catacombs an exaggerated idea of the impo- tence of late Classic art, for it must be observed that the execution of the catacomb frescos was from first to last exceed- 188 PICTORIAL ART ingly sketchy, as was natural in the case of a decoration in- tended for small sepulchral chambers which were rarely visited and were visible only by artificial light. At the same time, rapid execution reveals perhaps as well as more studied work the training of the artist. The mere decoration of the cata- combs demanded no more than a rough conventional design; the religious interest was satisfied with the bare indication of a well-known theme. In the later period the artist was con- tent to fulfil the bare requirements ; the artist of the earlier period rejoiced in picturesqueness of detail, in grace and natu- ralness of execution, the spontaneous expression of his skill. The decline in skill was so uniform and so uninterrupted that it enables one to determine approximately the date of most of the pictures. Hand in hand with this there went a growing carelessness even in the preparation of the plaster upon which the picture was to be painted, — as indeed in all the mechanical arts of building. The first impression of the paintings of the catacombs is rather one of uniformity, because in the general devastation the best are often reduced to the same state as the worst, and because, moreover, the color scheme is in all sub- stantially the same — yellow, red, and green upon a white ground. But the farther one goes back the more careful were the color gradations ; in the third and fourth centuries there was less and less attempt at gradation from one color or shade to another, and in later pictures which approach the Middle Ages the different parts of the figure were marked by the rude device of tracing with a line of black. But all of this is of more interest to one who would follow the decline of Classic art than to us who are pursuing the rise and progress of the Christian. Of so little interest from our point of view are the technical formalities of execution, that some of the best-preserved pictures are of less importance than others which are now to be traced only by the help of the artist's preliminary outlines which were scratched in the fresh plaster. Many of the subjects which were represented in fresco were reproduced in the rude graffiti which were scratched upon the tombstones or upon the walls, and the manifest imperfec- tion of this type of art hardly diminishes its interest. It will appear more plainly as we proceed that in early Christian art the substance was accounted of more importance than the PAINT ma — Introductory 1 89 form ; it is still more obviously true that our interest in these early monuments of art centres in the artist's purpose and thought rather than in the comparative success of his execution. The history of Christian art, if we regard art solely with reference to its form, is divided into three periods. The first and second centuries represent the best period of Classical art under the Koman Empire. This, however, was but a faint reflection of Greek art of the best period ; it will be remem- bered that the artists in Rome were for the most part Greeks, working under Roman influence and catering expressly to Roman taste. What was true of Rome in general was espe- cially true of the Christian community, which was thoroughly cosmopolitan in its constitution and actually used the Greek tongue as its official language. The third and fourth centuries were a period of steady decline. The following centuries wit- ness no check to this decadence, albeit they are characterized by a new feature, a certain stiffness and conventionality which was obviously due to a reaction from the side of the mosaic art, which was then fully developed in the decoration of the basilicas. But if we turn rather to its material than to its formal aspects, we get another division of the history of early Chris- tian art which is at once more definitely marked and more fundamental. To speak only of the first four centuries, we have first to consider a period, extending to some time in the early part of the second century, during which there was no Christian art properly so called, and the decoration employed by Christians was merely a selection of the more innocent themes of pagan art, to some of which a Christian symbolical sense might be attached. By the middle of the second century, however, we find an art as truly religious, as truly Christian, as any which was ever developed, and none the less Christian because it was couched in Classic form. It dealt already with a wide range of Biblical subjects, which, without any consider- able extension, remained the staple of Christian art throughout the third century. There was from the first a typical fixity in the mode of representing these subjects, which did not inter- fere, however, with a rather astonishing liberty in certain other respects. The order of time in which these subjects emerged, and their relative popularity, can rarely be determined with 190 PICTORIAL ART precision, and at all events it constitutes a study far too minute to be pursued here. It is sufficient for us here to regard them as a whole, and to note that they do actually display an unity of type which is due to the fact that they are stamped, expressed, and permeated by a certain mystic symbolism which was in many respects peculiar to the age of persecution, and seems to have lost its interest for the Church with the very beginning of triumph and prosperity. As a consequence of this, many of the subjects which were most familiar in the third century disappeared in the fourth, or were reproduced, especially in the sculpture of the sarcophagi, merely through the force of formal tradition. With the fourth century, therefore, begins the third period of Christian art. It was characterized by a new type of sym- bolism which comported with the new trend of thought which was initiated by the triumph of the Church, as well as by a new use of art as it emerged from the catacombs and was employed for the decoration of the churches. It was a sym- bolism at once more obvious and more formally dogmatic than that of the older type ; it may be said to be less profoundly religious and more ecclesiastical. During this period, however, symbolism no longer constituted the predominant, nor even the principal, interest of Christian art ; it was supplanted by a purely didactic and historical treatment which dealt chiefly in the description of Biblical scenes. But what is peculiar to the development of the fourth and fifth centuries belongs primarily to the arts of mosaic and sculpture and will be treated under those heads. We have here to deal chiefly with the art which is represented by the early frescos of the catacombs, or more generally with the whole range of Christian symbolism as it was expressed during the age of persecution. Inasmuch as early Christian painting is represented almost solely by the frescos of the catacombs, and more particularly the catacombs of Rome, we must take into consideration the query, whether, and to what extent, it is the expression of merely local traditions, and how far it was expressly designed for and limited to the use of sepulchral decoration. Even as between one catacomb and another one may mark differences which are expressed both in formal treatment and in a preference for one or another group of subjects. They PAINTING — Introductory 191 were evidently due to the individuality of the artists who were employed. But these differences do not go very deep, the range of subjects was after all so limited, each theme was repeated over and over again so many times and with so much similarity of treatment, even among the small minority of Roman frescos which have been preserved, that we have rea- son probably to mourn no absolute loss of a new and singular theme among that great majority in Rome and elsewhere which have been destroyed. We learn from literary sources that the same symbolism was current throughout the Church, and it is not only the uniformity of art throughout the Empire, but still more the wonderful unity which bound together the Christian communities during the second and third centuries, which leads us to believe that the art of the Roman catacombs was the art of the Church. We shall see that the themes which are represented in the art of the catacombs were in the main chosen with express reference to the purpose which they actually served, that is, as decoration ' appropriate to the grave. There is no reason to suppose that these same subjects were employed in the decoration of private houses, nor that beside this another cycle of Christian art was developed for the decoration of the earliest basilicas. We may, on the contrary, be quite sure that in point of decoration the house of a Christian dif- fered from that of a pagan in no other respect than in the exclusion of the indecencies of Classic art. Even in the fourth century no new domestic decoration was developed; at the most — as in the house of SS. Giovanni e Paolo in Rome — the conventional Classic decoration was eked out by a few of the traditional pictures of the catacombs. As for the decoration of the basilicas, it appears very evident that it was first developed in the fourth century. Even the mosaic decora- tion of S. Costanza was hardly more than a selection from among the gracious and innocent themes of Classic art. It is more closely related to certain of the catacomb frescos of about the end of the first century — as in the earliest cubicula in S. Gennaro at Naples and the vestibule of S. Domitilla at Rome — than it is to any of later date. The earliest frescos of the catacombs, those which antedate the creation of a distinc- tive Christian art, have therefore a peculiar interest, inasmuch 192 PICTOBIAL ART as they may be taken to represent the character of decoration which the Christians preferred for their houses and for the early basilicas which were developed from them. These early frescos of the catacombs, like the mosaics of S. Costanza (Figs. 125, 126), reproduce with a merely decorative interest birds, fishes, and beasts, trees, garlands, and flowers, ornamental fig- ures and busts, cupids and victories, and many mythological figures which were of too common a use to suggest expressly the pagan religion to which they traced their origin. Of these subjects there were some which gained a special popularity, because it was possible to attach to them a Christian symbol- ism ; such, for example, were the vine, the dove, the peacock, the fish, and the fisherman, and even such mythological figures as Psyche and Orpheus. But we must pass over this theme, we must confine our- selves here to that class of pictures which is beyond all doubt the most interesting, those, namely, which are distinctly sym- bolical, or at least religious, in character. There was a time when all the pictures of the catacombs were symbolically in- terpreted ; there has lately, on the other hand, been much controversial zeal displayed in the attempt to reduce the symbolical to the fewest possible number of instances. If we dwell here almost exclusively upon religious and symbolical pictures, it is simply on account of their superior interest, and it is not to be taken as a sign that we would ignore the fact that there are others, beside the purely decorative themes just noticed, which have no religious reference at all. It is beyond a doubt that there are many purely genre pictures, representing especially the professional or domestic life of the deceased ; it is true, too, that such pictures often throw an interesting light upon early manners and customs. It is only for lack of room that they must be dismissed with this bare notice, along with many subjects of an expressly religious character which cannot well be included under a few general heads which we can here study with some degree of com- pleteness. Before we pass to the study of any special themes it is necessary to consider the character of early Christian symbol- ism in general. PAINTING — Symbolism 193 EARLY CHRISTIAN SYMBOLISM It would hardly be necessary to discuss early Christian sym- bolism as a topic by itself and apart from the monuments which expressly illustrate its character, if it had not become a point of controversy more zealously contended than any other in the whole field of early Christian archaeology, if the symbolical element in the art of the catacombs had not been fantastically exaggerated by some and unduly minimized by others. The point at issue, however, is not whether early Christian art was symbolical or not ; but how far the symbol- ism extended, and what precisely was its character. From Bosio down, the archaeologists have commonly carried to so fanciful an extreme the symbolical interpretation of the pictures of the catacombs as to justify a sharp reaction. The reaction proves hardly less extreme, but it may be hoped that the controversy has now expired, leaving a substantial unanim- ity of opinion upon the principal points at issue. It may be remarked that de Rossi himself was not altogether a safe guide in matters pertaining to art, and his own immediate disciples have been prompt to make the needed retrenchments upon his teaching on this subject. A group of German Prot- estant writers who have lately attacked the Roman archaeolo- gists with some violence for their exaggerated interpretations of Christian art in the interest of Roman dogma, have perhaps had a sobering influence upon recent studies, though they were not justified in confusing all Roman archaeologists with the extremists, and they were unfortunate in transforming the question into a doctrinal and denominational issue. It is of no interest to us to follow the history of the contro- versy, it is enough to state its results as affording a sound basis for our own interpretation of early art. It must not be supposed that unanimous agreement is pos- sible upon all the subjects of early Christian art, for in many instances the interpretation must be acknowledged to be at the best uncertain. It is not to be hoped that all will confine themselves to a fair interpretation of symbols which may be so easily wrested to the pious purpose of supporting the faith. But neither uncertainties of interpretation nor differences of opinion must be allowed to prejudice the fact that some sym- 194 PICTORIAL AMT bolisin was actually intended by the artists and understood by those for whom they worked. It is of the nature of the pictorial symbol to be somewhat vague ; therein lies its chief advantage over the more definite forms of speech, its suitable- ness for the expression of themes — particularly religious notions — which evade the grasp of logical definition. It expresses, therefore, nothing with the definiteness of dogma, but in the freer terms of popular thought. It is equally of the nature of the symbol that it is capable of more than one interpretation, that it attracts to itself an aggregation of kindred suggestions whereby it is progressively enriched. It has been sought to throw contempt upon any symbolical interpretation of Christian art by showing that the early writers attached divers significations to the same symbol. What this does prove is not that the art of the catacombs was not symbolical, but that it was subject, even in early times, to various interpretations. It does serve, however, as a caution against an arbitrary attribution of a fanciful and purely sub- jective symbolism to the early age of the Church. One is at perfect liberty for one's own delectation to attach whatever meaning one chooses to any symbol, ancient or modern; but it is altogether fallacious upon this ground to attribute to the early age of the Church doctrines and religious conceptions which are not indubitably expressed by the symbol, or in some other way clearly proven to have been current at that time. It must be acknowledged that some of the Roman archaeolo- gists have extravagantly erred against this principle. From the pictures of the catacombs there is not so much to be learned about the dogmas of the early Church as has been supposed. They are chiefly interesting for the side lights which they throw upon the popular religious point of view and upon re- ligious or ecclesiastical practices. The fundamental error of the interpreters lay in supposing that the art of the catacombs had a didactic purpose, that it was intended, like the historical pictures of the basilicas, for the instruction of the ignorant and unlettered. This idea is refuted by the very form of these pictures, which, if they are symbolical, are on that very ac- count capable of expressing only that which was already familiar. It is equally inconsistent with the place where they were executed; for the cubicula of the catacombs were not PAINTING — Symbolism 195 places of frequent, far less of public, resort. We must recog- nize the probability that the pictures of the catacombs were appropriate to the place, and that their symbolism had refer- ence predominantly to the hopes which illumined the prospect of death. It has been claimed that the decoration of the catacombs was strictly superintended by the clergy, that the hand of the artist -was guided by the theologian ; and it is sought thus to justify, not only the discovery of a definite dogmatic import in the single symbol, but the concatenation of the separate pictures of a single cubiculum or arcosolium into a complex system of theological ideas. But there is no reason for attrib- uting to the early age of the Church the pedantic direction of art which is first proved in the case of the Byzantine Church in the eighth century. It is true that in the fifth century the bishops often personally directed the pictorial decoration of their churches, but the absence of any literary reference to Christian art during the first three centuries seems to prove that the ecclesiastical leaders had but little interest in it, that it grew up, therefore, in response to a popular demand and for the expression of popular ideas. Early Christian art is all the more important on this account, for we already know from their writings what the ecclesiastical leaders thought, and we desire very much to learn what notions were current among the people. We shall find that early Christian symbolism had a deep and mystical import, but that it by no means consti- tuted a learned and technical system. One must be on one's guard against weaving into a system any series of pictures which happen to be painted upon the same wall or in the same chamber ; they must be presumed to be separate unless they are connected by some range of ideas at once obvious and simple. A deeply thought out connection may, however, un- doubtedly be traced in a number of instances, and especially in the so-called Sacrament Chapels of the catacomb of Callistus. The proof that early Christian art was in fact symbolical is at once clearly displayed when we ask ourselves the ques- tion, Why do we find a monotonous repetition of an exceedingly narrow choice of Biblical subjects, to the total neglect of hosts of others which were no less important, no less familiar in 196 PICTOBIAL ABT Christian instruction, no less dramatic and picturesque ? There is no possible answer, except that their choice was de- termined by symbolical considerations. The symbolical em- ployment of art by the early Church is rendered antecedently probable by two considerations : first, by the fact that Roman art had, by the beginning of the second century and under the influence of Oriental cults, assumed largely a symbolical char- acter, especially in the decoration of sarcophagi ; second, by the common use in the Church of allegorical interpretations of Scripture. Alexandria, whose influence was paramount in the field of allegory, is likely to have had a leading part in fixing the character of the symbolism of Christian art. The character of this symbolism is best shown by the study of particular instances. What it is necessary to remark here is the effect which the exclusive symbolic interest had upon the execution of art. It is to this we have to attribute the absence of picturesque traits, indifference to such realistic effects as even late Roman art was able to compass, and carelessness about literal agreement with the terms of the Scriptural account. The subjects dealt with were familiar to all ; the artist could afford to abbreviate description, and in depicting the chief dramatic moment he relied upon the im- agination of the beholder to supply the whole scene. A man floating in a chest served to recall the story of Noah as an instance of divine deliverance. Christ touching with a rod a basket of bread recalled the miraculous multiplication of the loaves and symbolized the Eucharist. The story of Jonah was commonly represented in three scenes ; but even when it was abbreviated to one, that quite sufficed for the purpose of the symbolism. The artists could not have been ignorant that according to S. John's Gospel the face of Lazarus was covered with a napkin and his tomb was closed by a stone door ; yet he is invariably represented with uncovered face standing under an open tegurium ; — for some reason or another this mode of representation was adopted, and since everybody understood its meaning, the artist had no need to seek a more accurate expression. No one could suppose that the ark of Noah was a cubical chest too small to contain a single person ; but that was the traditional form, it demanded very little room, and the symbolism which was involved was as well PAINTING — Symbolism 197 expressed by that as it would have been had the artist stretched his invention to depict a boat which was capable of containing all the beasts of the field and all the fowls of the air. All of this is precisely what might be expected if the situa- tion be fully taken into view. The beginnings of pictorial art have always been symbolical, at least in the sense that more is suggested than a rude art is able to depict. All art is apt to strive toward the attainment of realistic representation, or, in other words, toward illusion. But with this attainment it at once loses a certain power which lay in its early suggestion. The crude wooden idols of archaic Greek art and the black Byzantine Madonnas of the seventh century were worshipped with a religious ardor which no marble of Praxiteles, nor any canvas of Raphael, ever evoked. The archaic smile of the early Greek image might suggest a divine benignity ; the perfect art of Greece could represent all that is gracious and noble in the human countenance, but by its very perfection it closed the way to all ulterior suggestion. Religious art is fundamentally sym- bolical, at least in the sense that it relies upon suggestion, because it deals with things which cannot be depicted. Chris- tian art was predominantly symbolical until the Renascence ; it cannot be denied that with the skill it then attained it lost much of its religious character. In this case the value of early crude suggestion was in a measure offset by the power of depict- ing the human face as it was moulded and inspired by Chris- tian motives and by the Christian life. At the commencement of the Christian period Classical art was, indeed, unable to express this lofty moral ideal — life had only begun to furnish the models for it. But it was very well able to depict with striking realism the common objects of na- ture and the common scenes of life. Only a little earlier the artists had boasted of their skill to delude the eye. If Chris- tian art did not take full advantage of the skill which was actually at its command, this was due to the fact that it was intent upon other things, upon the expression of conceptions for which mere skill of hand availed nothing. It was bound to ignore the possibilities of realism, of the descriptive and picturesque in art, and to rely upon suggestion and symbol. When early Christian painting and sculpture displays a care 198 PICTORIAL ART for artistic symmetry, for picturesque detail, for realistic pre- cision, it is the spontaneous expression of the artistic instinct which demanded its rights ; all this was not required of the artist, and it is not to be wondered that it was often lacking. From this it will be apparent why it was that Christianity in- terposed no obstacle to the decline of Classic art : it put no premium upon artistic skill; it subordinated technical excel- lence to the interest of the religious symbol. SYMBOLS OF DIVINE DELIVERANCE In the introduction to his Etude sur les sarcophages chre'tiens de la ville d' Aries, Le Blant had the credit of suggesting a theory which has been very promptly and very widely accepted by students of Christian archaeology. The theory was suggested expressly to account for the subjects which were sculptured upon the sarcophagi, but it applies quite as well to the frescos of the catacombs. It has the advantage of explaining many of the themes of early Christian art from a single point of view, and of referring them to a distinctly funereal symbolism, which expressed the Christian trust in the divine might which was able to deliver the soul even from death and the grave. Starting with the observation that the phraseology of many early epitaphs seemed to be derived from the liturgy of the Church, the suggestion lay very near that the pictorial themes also might have been inspired from the same source. It proved, in fact, very easy to show a relation between the themes of Christian sepulchral art and the funeral liturgies, which is alto- gether too complete and too precise to be fortuitous. Le Blant referred especially to the prayers in the Eoman Breviary which commend the soul to God in the hour of death. 1 After a long litany, come the following supplications : — Receive, Lord, thy servant into the place of salvation which he may hope of thy mercy. Deliver, Lord, the soul of thy servant from the pains of hell, etc. Deliver, Lord, his soul as thou didst deliver Enoch and Elijah from the common death of the world. Deliver, Lord, his soul as thou didst deliver Noah from the deluge. 1 Ordo commend adonis animse, quando infirmus est in extremis. PAINTING — Symbols of Deliverance 199 Deliver, Lord, his soul as thou didst deliver Isaac from sacrifice and from the hand of his father Abraham. And so the prayer continues with the same formula, men- tioning the deliverance of Daniel from the den of lions, of the Three Children from the burning fiery furnace and from the hand of the wicked king, Abraham from Ur of the Chaldees, Job from his sufferings, Lot from Sodom and from the flame of fire, Moses from the hand of Pharaoh, king of the Egyptians, Susanna from false accusation, David from the hand of King Saul and from the hand of Goliath, Peter and Paul from prison, and Thecla from horrible torture. It is remarkable that among these examples of signal divine deliverance there are very few subjects which are not repre- sented in early Christian art, and they are such as did not lend themselves to pictorial treatment — as the deliverance of Enoch and the departure of Abraham from Ur. On the other hand, this list includes almost all of the Old Testament subjects which were employed in sepulchral art. Many of these sub- jects are repeated in other prayers which are connected with the Roman funeral liturgies, and it is to be remarked that Lazarus and Jonah, omitted here, are elsewhere added to the list. They are here omitted, one may suppose, because they are most expressly types of the resurrection, whereas this is a prayer for deliverance from bodily death. It has to be ac- knowledged that these prayers are not to be traced to any texts which are earlier than the ninth century, and although in sub- stance they doubtless are very much earlier, it would be rash to refer them to a date as early as the fourth century, and still more so to suppose that they served as the text for the earliest examples of Christian art in the second century. They are, in fact, to be regarded rather as a concurrent witness with the art of the catacombs to the strength of a type of thought which was exceedingly familiar in early Christian literature and is to be traced back to the second century. This theory was put upon a much broader and a sounder basis by Victor Schultze, who sought out analogies even more complete in early Christian literature, and proved the broad currency of the argument which was drawn from signal ex- amples of divine deliverance and from the miracles of Christ, in proof of the power of God to deliver the soul from sin and 200 PICTOBIAL ABT from death. He shows also that the Scriptures themselves often furnish the hint which marks some of these subjects as typical of the deliverance of the soul from death, and he notes that the language of the Psalter obviously encouraged this mode of thought. Le Blant himself referred to an engraved glass cup of the fifth century found at Podgoritza (see p. 357), which groups together a considerable number of these subjects and accompanies some of them by inscriptions which seem like an echo of these ancient prayers. Schultze quotes as the Classical example a passage from the Apostolic Constitutions : 1 " He who raised Lazarus on the fourth day and the daughter of Jairus and the son of the widow, and rose also himself; who after three days brought forth Jonah living and unharmed from the belly of the whale, and the three children from the furnace of Babylon, and Daniel from the mouth of lions, shall not lack power to raise us also. He who raised the paralytic, and healed him who had the withered hand, and restored the lacking faculty to him who was born blind, the same shall raise us also. He who with five loaves and two fishes fed five thousand and had twelve baskets over, and who changed the water into wine, and who sent the stater which he took out of the mouth of the fish to those who demanded tribute by the hand of me Peter, the same shall also raise the dead." 2 In these texts there are included the majority of the subjects which were commonly employed in the adornment of the cata- combs and of the sarcophagi, and there can be no doubt that it was this argument of deliverance from death by the almighty power of God which accounts for the selection of these signal instances of deliverance and their monotonous repetition in early Christian art. The relation of these scriptural instances to the personal hope for departed friends is sometimes made very obvious by the substitution of the figure of the departed in the place of the P>iblical character. As an epitome of this whole series of representations we must regard the orans, the female figure with arms outstretched in prayer, which symbolized the soul of the deceased. One is left to wonder why the capital instance of this whole argument — the resurrection of Christ, i Book V. 7. 2 For other texts quoted by Schultze the student is referred to his Archaolo- gische studien, p. 15 seq., and Archaologie der altchristlichen Kunst, p. 181. PAINTING— The Orans 201 which the Scriptures treat as the most immediate proof of the resurrection of his followers — appeared so rarely in early- Christian art, and not at all until the fourth century. It is to be noted that some of these subjects appear first on the sarcophagi. This theory must not, however, be held to account for all of the symbolical art of the catacombs, nor must we suppose that this art had all of it a funereal reference. The subjects even of the cycle which we have just considered had some of them, in addition, a very different symbolical import. It is not to be doubted, for example, that the multiplication of the loaves was also a symbol of the Eucharist. On the other hand, there were many subjects quite outside this cycle which from different points of view brought their own contributions to the symbolism of the tomb. Such was the figure of the Good Shepherd, a many-sided symbol in which the funereal reference was at least incidental. Such, too, were most of the symbols which Christian art inherited from the pagan. Orpheus, the peacock, the phcenix, the dolphin, etc., were already, in Classic art, employed as sepulchral symbols and connected by mythology with the fate of the soul after death. We have now to consider in detail the artistic expression of the more important subjects included in the cycle of divine deliverance and miracle. It will be convenient to study here, once for all, the expression which was given them in sculpture as well as in painting. THE OKAKS The name orans, or orant, designates a figure in the attitude of prayer, with arms outstretched in the manner which was common to both Jews and Gentiles, and was accounted by the Church particularly significant, because it recalled the position of the Saviour upon the cross. Portraits of the departed were, in the art of the catacombs, generally represented in this atti- tude (Figs. 85, 124). We can readily understand why the Old Testament characters which are taken as examples of divine deliverance were also commonly represented in the posture of prayer. The Scripture itself furnished the suggestion, record- ing the prayer of Jonah and the song of the Three Children. 202 PICTORIAL AST Noah, upon his deliverance, immediately offered a sacrifice of thanksgiving; and the attitude of prayer was no less appropri- ate to Isaac delivered from sacrifice, to Daniel among the lions, and to Susanna accused by the elders. But though we may apply the name "orans" in a general sense to portraits of the deceased and to representations of Biblical characters, it is used more specifically to denote the artistic symbol of the souls of the faithful departed. It is not always possible to draw a sharp line of distinction between the portrait orans and this symbolical abstraction. Whole families, men and women and children, were sometimes portrayed in this pos- ture upon the wall of the tomb where their bodies reposed. Even where there was no attempt at a portrait, the distinction of sex was sometimes observed. But the great majority of orants were quite impersonal in their traits, and equally so, whether they consti- tuted part of the adornment of the ceiling, and represented the dis- embodied soul in general, or were brought into immediate relation with a particular tomb. The proof of the symbolical character of the orans lies in the fact that it was generally represented by a female figure, without regard to the sex of the deceased. So, for example, we find often a female figure painted above a tomb when the inscription denotes a man. An early lead medal in the Vatican Museum represents the martyrdom of S. Lawrence and depicts a female figure ris- ing from the body of the saint to receive the martyr's crown. In the Acts of SS. Petrtts and Mitreeflintts : "The executioner testified that he saw their souls issue from their bodies in the form of young virgins, who were adorned with gold and gems and clad in shining garments, and were carried to heaven by the hand of angels." It is impossible with any candor to Fig. 64. — Orans in palla, fresco in the crypt of Lucina. Second century. PAINTING— The Orans 203 deny the symbolical significance of the orans. The attitude of prayer comported with the situation of the soul at the moment of death: it represented at once thanksgiving for deliverance from death, and supplication of God's merciful judgment. The purely abstract symbol was not a development out of the portrait orans, for it appears among the earliest types of Christian art. Upon the ceiling of the crypt of Fig. 65. — Orans with dalmatic and veil, catacomb of Callistus. Third century. Lucina, which was decorated early in the second century, an orans of a thoroughly hieratic type (Fig. 64) is twice repre- sented, alternating with the figure of the Good Shepherd. In this case, the figure is clothed in the palla. But it is the posture and sex, not the costume, which denotes the orans. The costume, in fact, varied, and it registers the changes of female fashion in dress, from century to century (Figs. 64, 65, 66). It will readily be seen that 'the Old Testament heroes of 204 PICTORIAL ART deliverance who were represented in this posture, were thereby brought into relation with this symbolism, and their signifi- cance as types of the resurrec- tion was made more abundantly plain. The orans of the catacomb frescos was reproduced in the sculptures of the sarcophagi. This symbol had no doubt its influence upon the mediaeval representation of the soul as a diminutive body issuing from .the mouth of the dying to be re- ceived by angels, or by demons, as the case demanded. OLD TESTAMENT CHARACTERS It must be borne in mind that we have here to do solely with the Old Testament characters who were regarded as types of deliverance from death. But though we have this symbolism primarily in view, we shall not neglect to consider other sym- bolical notions which were asso- ciated with these same subjects. It is doubtful whether Adam and Eve properly belong to this cycle, or what precisely their significance may be in Chris- tian art. They appeared very early in the catacombs of S. Gennaro at Naples, where they were associated with the myth- ological decoration of an early crypt. But at Borne the sub- ject did not become popular till the fourth century, after which time it was employed very frequently upon the sar- cophagi and upon the gold-glasses. The creation of man is represented but three times ; the best-known example is upon a sarcophagus in the Lateran (Fig. 95). The theme commonly chosen is the temptation : in the midst is the tree with the serpent coiled about it, and on either hand stand Adam and Fig. 66. — Orans with girdled tunic and veil, fresco in the catacomb of Thrason. Fourth century. PAINTING — Noah 205 Eve, endeavoring to hide their nakedness with a leaf. Very commonly the man and woman are accompanied by the sym- bols of their respective labor : a sheaf of wheat represents the work of the husbandman, a lamb signifies the woman's work of weaving (Fig. 100). In some cases (Tigs. 95, 98) the figure of Christ appears between them, allotting to each their labor. Eve appears sometimes upon the gold-glasses with hair richly dressed, and wearing bracelets and arm bands (Fig. 163). Several sarcophagi represent the exclusion from paradise. The subject of Noah appeared as early as the end of the first century, and it was frequently employed in the art of the catacombs during the second and third. The earliest instance is in the vestibule of the catacomb of Domitilla, where, how- ever, the only vestige of the picture which remains is the dove bearing the olive branch. The subject was so simple in its treatment, and it was re- peated with so little variation, that it requires few words of description. The representation is in the highest degree ab- breviated: Noah stands alone and in the posture of an orans in a small cubical chest which floats upon the waves, while the dove flies toward him with the olive branch. In only one instance, a sarcophagus in Treves, is the whole family repre- sented in the ark, with various birds and beasts. Sometimes a lad, or even a female orans, appears in the place of Noah, more clearly indicating the typical character of the subject. It is very curious that the ark is represented in this same form on a coin of Septimius Severus stamped in Apamea in Phrygia (Fig. 79). In this case Noah and his wife are twice depicted, at one moment receiving the dove in the ark, in another standing upon the land and offering a prayer of thanks- giving for deliverance. It has been shown that there was a strong Jewish colony in this town, and the fact that it lay under the shadow of Mt. Ararat suggested this theme. It is not unlikely that the form which is here given to the ark was suggested by some prototype in Classical art, perhaps by the chest in which Danae and Perseus were set adrift. There is nothing in the monuments themselves to indicate whether any other symbolism was intended beyond that which we have traced. But it is not to be forgotten that in the first Epistle of S. Peter the ark is treated as a symbol of baptism, 206 PICTORIAL AST and that Tertullian, elaborating this figure, made the ark a sym- bol of the Church. The dove and the olive branch — or the olive branch alone — became a still more abbreviated symbol of deliverance and of peace. We have seen that Moses' deliverance from thehand of Pharaoh was one of the subjects which appears in the liturgies as a type of the resurrection. There is evidently no single dramatic in- cident which expressed this deliverance ; various episodes of Moses' life, however, were frequently depicted in fresco and sculpture. His colloquy with God at the burning bush was conventionally abbreviated almost beyond recognition;- even the bush was dispensed with, and Moses appears with one foot raised upon a rock bending over to take off his shoe. This, according to Wilpert, stands for the believer who prepares after death to appear before the presence of God. Especially common on the sarcophagi was the scene in which Moses re- ceives the tables of the Law from a hand stretched out of a cloud. Most frequent of all, and especially in the art of the catacombs, was the striking of water from the rock. The mode of representation was practically the same in painting and in sculpture ; it is illustrated in Fig. 96, and at the lower right- hand corner of two sarcophagi which are shown in Pigs. 95 and 98. On the sarcophagi this subject often balanced the resur- rection of Lazarus. This was doubtless due chiefly to the desire for symmetry ; but it can hardly be doubted that Moses was regarded as a type of Christ, as the water also was a 'sym- bol of baptism. It is claimed by Roman archaeologists that in the fourth century Moses was represented with the facial traits which were traditionally ascribed to Peter. A gold-glass in the Vati- can which represents the striking of water from the rock is inscribed with the name Petrns. The artisans who made such wares frequently fell into error in attaching names to the subjects which they represented ; but as this is not the only instance in which the name of Peter is attached to the figure of Moses, it is probable that a parallel was sometimes drawn between the chief of the Apostles and the chief figure of Jew- ish history. No subject was so popular in the art of the catacombs as the PAINTING — Jonah 207 story of Jonah. It not only furnished the most striking sym- bol of the resurrection, and one which Christ himself appealed to, but it provided the best field for the exercise of the artist's fantasy. There are about forty instances of it among the paint- ings, and it was repeated also in sculpture, although in the latter case it presented manifest difficulties. The artistic ex- pression of this theme was subject to a good deal of variety in detail, but the sarcophagus illustrated in Fig. 94 shows sub- stantially the mode in which it was treated both in fresco and in sculpture (see also Fig. 164). The story was represented in three scenes : the first was that in which the prophet was cast out of the ship into the mouth of the sea monster; in the second he is cast up by the monster upon the land ; and in the third he reposes under the gourd. It will be observed that the monster is not a fish, but one of the fabulous dragons which were common in Eoman art; it is probably most closely related to the dragon from which Andromeda was delivered. This theme, too, was some- times abbreviated by dropping one or more of the three scenes ; or space was economized, as in Fig. 98, by crowding the scenes so together that Jonah is spit out by the monster directly under the gourd. It is remarkable, however, that the scene which is generally retained and often quite alone (Fig. 155) is that in which the prophet rests under the gourd. It is evident that this was the most essential feature of the symbolism, and we have therefore to recognize that the original significance of the gourd, according to the text, was neglected, and the naked fig- ure of Jonah reposing under its shade was regarded as a sym- bol of the soul in the joy of paradise after being delivered from the dangers and the pains of death. It is to be remarked that even under the gourd Jonah is always represented naked. That he should be so represented when cast into the sea was alto- gether natural, but the story furnishes no reason for it in the case of his sleep under the gourd, since that followed upon his successful preaching to the Mnevites. It seems probable that the nakedness was in this case altogether in the interest of the symbol, to represent the ideal state of bliss in paradise. The posture suggests Endymion. When one considers the great number of nude or half-nude figures in Classic art, one cannot but be struck with their rarity 208 PICTORIAL ART among the Christian monuments. There was certainly no rigid scruple against the art of the nude, but there was as little pre- dilection for it. Eve was the only nude female figure known to early Christian art. Susanna was represented clothed, not surprised at her bath, as the artists of the Eenascence de- lighted to depict her. Daniel was generally, though not always, naked ; Adam and Jonah were always so. We have to add to this, occasional pictures of Tobias, or of a fisherman, several representations of Isaac sacrificed, and of the vision of Ezekiel, one of the martyrdom of Isaiah, one of Christ baptized, and two of Christ upon the cross. These rare representations of the nude are interesting, especially in sculpture, as affording a gauge of the technical capacity of art ; they prove what it was still capable of in the fourth century, and how utterly impotent it had become by the fifth. A subject equally familiar in the catacombs and on the sar- cophagi was the sacrifice of Isaac. The moment commonly represented is that in which Abraham stands ready with a knife to slay his son, who with hands bound kneels upon or beside an altar. A hand from heaven arrests the act, and the ram appears to replace the sacrifice. The mode in which this subject was represented is shown in the upper left-hand corner of the sarcophagus of Junius Bassus (Fig. 100). It was very often balanced by the receiving of the Law, as in Fig. 99. This, we have seen, was one of the symbols of deliverance. It is well known that the subject was also regarded as typical of the sacrifice of Christ, but how far this idea was associated with the artistic expression of the theme we have no means of knowing. It seems, however, that this must have been in the mind of the artist who in the third century represented this theme in connection with the Eucharistic symbols in the Sacrament Chapels of S. Callistus, and it certainly was in- tended in the sixth-century mosaic of S. Vitale (Fig. 133). Still more common was the subject of Daniel among the lions. It appears for the first time at the end of the first century in the vestibule of S. Domitilla. The picture repre- sents a young man clad in a short tunic, with arms extended in prayer and eyes raised to heaven, while a lion approaches on either side. With this early example, the form was already fixed, as it appears constantly in the frescos and in sculpture PAINTING — Daniel 209 (Figs. 95, 98, 99, 100) , — only, as was remarked above, the figure of Daniel was commonly naked. Generally the prophet Habakkuk appears at his side offering him bread, — a whole' loaf, not bread broken into pottage, as the story has it. As the loaf is often marked with the cross (Fig. 98), it seems evident that it had a mystic reference to the Eucharist. The great popularity of this subject was probably due to the fact that it represented the fate of so many Christian martyrs. The manner in which the martyrs were exposed to the beasts in the Roman amphi- theatre sometimes influenced the mode in which the subject of Daniel was depicted. It is not certain whether the clay lamp illustrated in Fig. 151 represents Daniel or a Christian mar- tyr ; but it is evident that it represents realistically the man- ner in which Christians were actually exposed to the lions. In Fig. 160, which illustrates a fragment of a cut-glass vase of the early part of the fourth century, there appears at the right a part of just such a platform as is shown in the previous fig- ure ; and the intention here is evidently to represent Daniel, for Habakkuk appears above, while on the left is the sacrifice of Isaac, and below the Israelites marching through the desert with their eyes fixed upon the cloud of fire. Another scene in which Daniel frequently appears is that in which he slew the dragon (Figs. 163, 166). The dragon is represented in a form similar to the serpent which tempted Eve. One may readily notice how closely some of the other exam- ples of deliverance correspond with forms of martyrdom which were actually suffered by the Christians. It was only in this indirect way that scenes of Christian martyrdom were depicted, for in the first three centuries there were no express represen- tations of such themes. Scenes of pain and of horror were altogether foreign to the spirit of early Christian art. We shall have occasion to notice later that representations of our Lord's Passion do not appear before the fifth century and were altogether exceedingly rare throughout the early period. The earliest representation of a martyrdom belongs to the latter part of the fourth century; it is found on a fragment of one of the columns of the ciborium of the basilica of SS. Nereus and Achilleus (Fig. 22). It represents Achilleus standing with hands bound behind him ready to receive upon his neck the executioner's sword, In the background is the cross sur- 210 PICTORIAL AST mounted by the crown. Another scene was painted about the same time in the house of SS. Giovanni e Paolo in Eome : three martyrs are kneeling blindfold upon the ground, while the executioners stand behind ready to strike with the sword. Beside these there are almost no other such scenes within our period. Closely related to Daniel were the stories of Susanna and of the Three Children in the furnace. Susanna was often represented standing like an orans between the two accusing elders. Sometimes Dan- iel appears as judge and deliv- erer (Kg. 67). In the Capella greca in S. Priseilla the story is depicted with unusual dramatic interest and in several scenes. In the cemetery of Pre- textatus there is a painting which represents a lamb between two wolves, and the inscription informs us that we have before us Susanna and the elders. Far more frequent both in the frescos and on the sarcophagi were the Three Children of Babylon. They are commonly represented standing in the attitude of prayer amidst the flames of a furnace which is fed below by an attendant. Sometimes a fourth person, evidently the Lord, appears in their midst. They are always clothed in the Phrygian cos- tume, which was used in Roman art to denote the inhabitants of the extreme Orient. This was probably the suggestion which led to their association with the Three Magi from the East who followed the star to Bethlehem, for they also were represented in the same costume. Fig. 67. — Susanna and the elders judged by Daniel, fresco in the catacomb of Callistus. Second century. PAINTINa— The Magi 211 This brings before us a curious, but characteristic trait of early Christian symbolism ; for these two subjects were not only compared, but in a certain sense confused, in art. This appears in many instances upon the sarcophagi, and always in substantially the form which is illustrated in Fig. 89. In this case the Three Children are represented in the act of refusing to worship the image of Nebuchadnezzar. On the other side the Magi, similarly clad, offer their gifts to the infant Jesus. It will be noticed that the Three Children turning away from the image are pointing to a star, and are thus brought into still Fig. 68.- - Visit of the Magi, fresco in the catacomb of Petrus and Marcellinus. century. Third closer relation with the Magi. One point of connection was doubtless the refusal of the Magi to listen to Herod's com- mand. But besides this there was a deeper thought, for Neb- uchadnezzar was regarded as the type of the persecuting emperors, and it will be seen how admirably balanced these two subjects are when it is remembered that it was the refusal to worship the emperor instead of Christ which was the Chris- tians' chief offence against the State, and the cause of the majority of martyrdoms. The star to which the Three Chil- dren point represents the true worship as opposed to idolatry, for it is the symbol of Christ. The star was conventionally represented by three lines intersecting at equal angles, and in this figure the Christians saw at once a symbol of the cross 212 PICTORIAL ART and the initial letters of the name 'lr/a-ovs Xpioros. This form of monogram appears long before Constantine, and we may sup- pose that the Constantinian monogram was suggested by it. On one sarcophagus the star to which the Three Children point has the form of the Constantinian monogram. It is not unlikely that the number of the Magi was fixed chiefly with reference to this association. This is at least more probable than that it was due to the number of gifts which they bore — gold, frankincense, and myrrh. The sub- ject of the Magi was a very common one in the catacombs as well as in sculpture and mosaic. The number three appears in the earliest instance — the Capella greca in S. Priscilla (Fig. 3). But very commonly an even number of Magi were represented for the sake of symmetry — two (Fig. 68), four, or in one instance six. The number three became definitely fixed in the fourth century, and it was invariable on the sarcophagi and in the mosaics (Fig. 139). In Fig. 86 the wise men and a shepherd appear together at the manger, where the ox and the ass which are mentioned in an apocryphal Gospel have also their place. It is as an incident to the story of the Magi that we have the earliest representations of the Virgin and Child. It is probable that this subject was then understood, as it certainly was later, to represent the calling of the Gentiles. There were no other Old Testament subjects so frequently depicted as those which we have just studied. Lack of space prohibits the particular consideration of more of them here, though some will be noticed incidentally as they appear in sculpture. David with his sling appears in but one fresco, Job is seldom represented in the catacombs, only once the transla- tion of Elijah, though several times on sarcophagi (Fig. 97). Tobias with his fish is represented graphically in the catacomb of Thrason ; on the other hand the vision of Ezekiel, which was peculiarly appropriate to sepulchral decoration, does not appear at all in the catacombs, and only a few times on the sarcophagi. This very nearly exhausts the types of deliver- ance so far as they were drawn from the Old Testament. We have now to consider the miracles of Christ from this same point of view. PAINTING — Miracles of Christ 213 THE MIBACLES OF CHRIST In the art of the catacombs a somewhat arbitrary selection was made of the miracles of Christ; some themes were re- peated again and again, while others found only here and there a chance notice. It must suffice to notice here the subjects which were the most common, and, important as they were, they require but few words of description. It is neces- sary here again to rely for illustrations chiefly upon the sar- cophagi ; the scenes were treated with equal simplicity and with like abbreviation in the frescos. The raising of Lazarus was the most important of these subjects, since it most directly testified to the power of God to raise the dead at the resurrection. The form is always substantially the same (Figs. 18 and 95, upper right-hand cor- ner) : Lazarus wrapped like a mummy stands erect in the entrance of a temple-like tomb, and Christ stretches toward him the rod which symbolizes his power. It is to be observed that in almost all of the representations of his miracles Christ, like Moses, carries a magic wand or rod. Since the motive was a symbolical and not an historical one, this miracle stood for others of like character. The raising of the widow's son does not appear in the art of the catacombs. As the artists were interested (especially in the case of the sarcophagi) in crowding as many subjects as possible in a small space, a specimen of each hind of miracle sufficed. So it was in the case of the healing of the blind. Christ is represented placing his finger upon the eyes of a single blind man (Fig. 98), but which particular act of healing this was the artist had no interest in specifying. According to a well- established principle of Classic art, subordinate personages, such as those who were craving Christ's assistance, were repre- sented in minute proportions, and the blind men look conse- quently like little children. Christ's gesture in healing the blind man has been sometimes misunderstood, and the scene has been misinterpreted as a representation of Christ blessing little children, — a subject which, strange as it may appear, never occurs in early Christian art. A female figure, often exceedingly minute, crouching at the feet of Christ and touching his garment or his hand, is the 214 PICTORIAL ART woman who was healed of an issue of blood (Figs. 95, 96). Equally simple, but more distinctly characterized, was the figure of the paralytic. It was very common both in the frescos and on the sarcophagi, but there happens to be no illustration of it in this book, except on a gold-glass (Fig. 168), which groups together a number of the subjects which we have just been studying. 1 To represent this miracle nothing more was necessary than the figure of a man carrying his bed. There is nothing to denote whether it was the man healed at Capernaum, or him of the pool Bethesda; it probably was intended to represent both at once. The multiplication of the loaves and fishes was in the catacombs represented in various ways which we shall have occasion to study later; on the. sarcophagi the treatment is stereotyped : Christ stretches out his rod over the baskets, or he touches with his hands the bread and fishes which are presented to him by two Apostles (Fig. 95). In the transformation of the water into wine he touches with his rod the waterpots (Fig. 95). These two miracles had more especially a relation to the Eucharist, and they will be noticed again in another place. In Fig. 72 Christ is represented with the Samaritan woman at the well ; and the same scene is several times reproduced in fresco and in sculpture. THE GOOD SHEPHERD There were various myths current in early times, and com- monly believed, about miraculous portraits of Christ. There was, in the first place, the napkin with the impression of his own face which our Lord himself was supposed to have sent to Abgar, king of Edessa. Closely related to that was the Veronica portrait, about which there were divers traditions. As no one any longer supposes that these were actual portraits, their sole interest lies in the proof that the effort was actually made in early times to produce an ideal representation of the face of Christ. In the loss of these pictures it remains im- possible to determine whether or not they had any influence in the determination of the type of head which was commonly 1 After the paralytic comes Moses striking the rock, then the sacrifice of Isaac, Adam and Eve, and the raising of Lazarus. PAINTING— The Good Shepherd 215 represented in the fifth century. The presumption is that they had not, and that there was no fixed tradition to hamper the artists in the expression of their own ideal, for this Yaried freely, and was marked by three distinct types which belong to the fourth, the fifth, and the sixth centuries respectively. Historically attested are the images of Christ (made of glass, gold, or silver) which one of the Gnostic sects venerated along with images of the philosophers Pythagoras, Plato, and Aris- totle. And it is known that the Emperor Alexander Severus placed in his lararium and honored with sacrifice the images — probably medallions — of Christ, of Abraham, and of Or- pheus. In neither of these cases is there any suspicion of a portrait. Of more " importance is Eusebius's account of the bronze statue in Csesarea Philippi. It represented the standing figure of a man clothed in the pallium and stretching out his hand to a woman who kneeled at his feet. There is no doubt that at the beginning of the fourth century this was regarded as a representation of the Lord, and that it was believed to have been erected out of gratitude to him by the woman whom he had healed of an issue of blood. This derivation of the statue is commonly rejected by historians, and it is explained rather as the statue of an emperor receiving the submission of a province. But it is difficult to believe that Eusebius, who himself saw the statue, could have failed to detect in it so common a theme of Roman art as the statue of an emperor. The statue was removed from its public position by Maxi- mums Daza out of hatred to Christianity, and was finally destroyed by Julian. It is hardly to be thought that these emperors could have mistaken its true character, if it were actually of pagan origin ; and there must remain at least the suspicion that the tradition was correct. When we turn from mythical or historical accounts to the exist- ing monuments of early art, the spirit in which the early Church regarded the representations of Christ seems obvious and un- equivocal. There were not only no representations which had any suggestion of realism or of portraiture, but till the middle of the fourth century not even an idealistic conception seems to have been so much as attempted. In the earliest cases in which our Lord appears he is depicted as an infant in the arms of his mother, and it is evident that he appears only as an inci- 210 PICTORIAL AIIT dent to the scene — as in the visit of the Magi, etc. It may be said, too, that in the representations of the miracles his pres- ence was equally incidental ; it was required to determine the subject, just as the presence of God in the Old Testament won- ders was symbolized by the hand stretched from heaven. In this case the artists implicitly intimated that they had no in- tention of representing Christ as he actually was, or as the Church might like to conceive him. They represented him Fig. 69, — The Good Shepherd, fresco decoration of a ceiling in the catacomb of Petrus and Marcellinus. Third century. under the figure of a very young man, beardless, and without any such marked personal traits as were early ascribed to Peter and Paul. The figure of Christ is, in fact, to be distin- guished only by the situation in which he is represented — the performance of his miracles — and by the pallium, in which generally only he and the Apostles and Moses were clad. But beside these incidental expressions it is evident that the Church did desire an artistic representation of Christ, and that this was attained in the symbolical figure of the Good Shep- PAINTING— The Good Shepherd 217 herd. This, it is well known, was one of the commonest fig- ures in early Christian art ; and from the second to the fourth century it was beyond comparison the most favorite represen- tation of Christ. It appears for the first time in the early part of the second century in the crypt of Lucina ; and later very fre- quently in frescos, on sarcophagi, and as an independent statue, upon gold-glass, upon rings, upon medals, and in almost every branch of art. Nothing could have comported better than this symbol with the spirit of early Christian art, and the mode in which it was treated was highly characteristic. When this subject is represented in our modern art the character of the shepherd is but a slender disguise ; it may be denoted by noth- ing more than by the fact that he carries a lamb and has the shepherd's crook, while for the rest it is the ideal face and figure with which we are accustomed to depict Jesus in all the scenes of his earthly life. There was nothing like this in the art of the catacombs ; the symbol was frankly carried out, — it was a symbol and no whit more. The insufficiency of art to portray in realistic terms the figure of the God-man was clearly recognized, as we have seen, and this feeling was frankly ex- pressed by a form of representation which could not even sug- gest the idea of a physical likeness. What was represented in the catacombs was just such a shepherd as might be seen in the campagna about Eome, a beardless lad of some sixteen years, clad in the short sleeveless tunic of the laborer, with his right shoulder bare, sometimes with feet and legs bare also, sometimes with shoes and stout leggins. There was a perfect realism about the representation, and yet at the same time a certain idyllic graciousness — it has been not unjustly called an hieratic trait — which distinguishes it from the same subject as it appeared in pagan art. The statue in the Lateran is alto- gether the finest example of the Good Shepherd (Kg. 117) : with both hands he holds a sheep upon his shoulders ; one shoulder- is quite bare, for he wears the tunica exomis, which is girded up to the knees, as it commonly was ; at his side he carries the shepherd's scrip. The gold-glasses which are given in Fig. 162 show common varieties of the subject, and illustrate different ways of holding the sheep. They exhibit also the short cape which seems to have been a familiar feature of the shepherd's attire. In two instances the shepherd has leggins; in one, 218 PICTORIAL ART the feet and legs are bare. The head is always bare. As here, so also in the other art of the catacombs (Figs. 12, 13, 14, 16, 17, 69), the figure was often balanced by two or more sheep at the shepherd's feet. It will be noticed that in one of these gold- glasses the shepherd is not pictured carrying back the wounded sheep to the fold, but is sitting and tending his flock. Eepre- sentations of the shepherd tending his flock occur some twenty times in the catacombs, in frescos and graffiti. In Fig. 19, for example, he is represented with his shepherd's staff in his hand, seated under a tree and playing upon the pipes, while the single sheep in the scene looks attentively up at him. Nothing could be plainer than this symbolism, nothing more gracious and touching. This whole range of meaning was expressed by our Lord himself, 1 and it was appreciated in its fullest extent in early art. But there can be no doubt that, in the catacombs and upon the sarcophagi, the figure of the Good Shepherd had especially a sepulchral reference. It represented, of course, the faithful care of the divine Shepherd in seeking the souls which had strayed into sin, and bearing them back to his Church ; but it also represented his power to bear aloft to his heavenly kingdom the soul which was wounded, weary, and bruised with the struggle here below. This was suggested by the familiar words of the Twenty-third Psalm : " Though I walk through the valley of the shadow of death I shall fear no evil, for thou art with me, thy rod and thy staff they com- fort me." And this reference is expressed plainly in many of the prayers of the Church, for example, in " the prayer after burial " of the Gelasian Sacramentary : " Let us pray God to grant that the deceased, carried upon the shoulders of the Good Shepherd, may enjoy the fellowship of the saints.'' A somewhat different symbolism is expressed by the pictures which represent the shepherd playing upon the syrinx and herding the sheep. It was not the parable of the Lost Sheep which was reflected in this theme, but the richer symbolism which was expressed by the speech recorded in S. John's Gos- pel, according to which the shepherd protects his sheep from the wolf, so that they can safely go in and out and find pas- ture ; he calls them by name and they know his voice and fol- low him, but a stranger will they not follow, for they know 1 John x. 1-27; xxi. 15-17; Matt. xv. 24; Luke xv. 4, 5. PAINTING — Hie Good Shepherd 219 not the voice of strangers. It is this image which is expressed in the epitaph of Abercius (see p. 235). The fulness with which the Church appropriated this symbolism is strikingly- shown by the fact that the figure of Orpheus was occasionally represented in place of the Good Shepherd. That the repre- sentation of Orpheus was not an unthinking repetition of a subject which was common in pagan art, is placed beyond a doubt by the fact that it occurs in the central medallion of the ceiling, that is, in the place which was especially appropriated to the Good Shepherd. It will be remembered that Orpheus too was a shepherd. In the catacombs, as in Classic art, he is represented in Phrygian costume ; he sits surrounded by his sheep, or by various birds and beasts, which listen attentively to the music of his lyre. It is impossible not to recognize that it was the intention of those early artists to associate with the Good Shepherd the loftiest moral myth which was known to the pagan world, to represent under the figure of Orpheus the compelling music of Christ's voice, which drew all men after him and closed their ears to the song of the Sirens. It must be remarked that the scene of Orpheus and the Sirens does not occur in early Christian art, though that other fable of Ulysses binding himself to the mast is represented several times upon the sarcophagi (Fig. 88). But this was not the only nor the most express symbolism which gave the figure of Orpheus a place in the decoration of the Christian cemetery. In Classic art, also, Orpheus belonged to the funereal cycle, for it was fabled that through the power of his music he alone had been able to penetrate into the lower world and lead up from Hades a human soul. The symbolism which was attached to the sheep was a very simple one in the early centuries. Of the sheep which was carried upon the shoulder of the Good Shepherd enough has been said. The fact that it is sometimes a kid instead of a lamb is probably not significant. It is not till the fourth cen- tury that Christ is represented dividing the sheep from the goats (Fig. 87). The sheep which he tends and feeds represent either his flock on earth, or faithful souls in heaven. The lat- ter sense is expressed by several paintings in which behind the sheep the figures of the blessed dead are represented in the posture of prayer. In the cripta delle pecorelle in S. Callis- 220 PICTORIAL ART tus the Good Shepherd is represented surrounded by a number of sheep, and in the midst of them stand two men who stretch out their hands to drink of the water which abundantly flows from the rock on either side. Here the shepherd is depicted refreshing the souls in heaven with the water of life. On a sarcophagus in the Lateran the Good Shepherd is surrounded by twelve sheep, and behind each of them stands an Apostle. After the fourth century the twelve sheep were very com- monly used to represent the Apostles, especially in the mosaics ; but the symbolism of the Good Shepherd rapidly fell out of favor and was replaced by the mystic lamb of the Apocalypse. In the mausoleum of Galla Placidia there is a mosaic of the fifth century which again represents the Good Shepherd (Fig. 141), but in a manner which reminds one of our modern art. It is no longer properly a shepherd, but an idealized figure of Christ adorned with the nimbus, and carrying a cross. It is strange to note how early and how completely the old symbol- ism disappeared, never to be revived again in Christian art. But before leaving this subject we have still to trace another trait of symbolism which was associated with the Good Shep- herd in its earliest representations. I refer to the milk with which the shepherd nourishes his sheep. This was represented by the jug or pail which he carries sometimes at his side (Fig. 78). The importance of this apparently trivial detail is shown by the fact that the sheep and the pail of milk sometimes appear apart from the shepherd. In the crypt of Lucina two sheep are represented on either side of a rustic altar of stone on which rests the jug of milk and the shepherd's staff. There are somewhat similar representations in the catacomb of Dorni- tilla and in the house of SS. Giovanni e Paolo. This is so delicate a symbol, and it is so remote from the terms of modern thought, that its sense could hardly have been understood but for an explanation which is furnished in the Acts of S. Per- petua, which was written in the early years of the third cen- tury. It describes a vision of the saint in these words : " I mounted and beheld a garden of vast extent, and in the midst of this garden a man seated, having white hair and the dress of a shepherd, milking sheep : and standing round about him thousands of men clothed in white. And he raised his head and looked at me, and said to me : Thou art welcome, my PAINTING— The Celestial Banquet 221 daughter. And he called me and he gave me some of the warm milk which he had just drawn, and I received it with folded hands and I ate it : and all about me said, Amen. And at the sound of the voice I awaked, with an indescribable taste of sweet in my mouth." From this it is evident that the jug of milk carried by the Good Shepherd is a mystic symbol of the Eucharist. The favorite representations of heaven or of paradise dur- ing the first three centuries, as we see from the vision of S. Perpetua, were in terms of the pastoral symbol. The olive tree serves here to denote the celestial garden, as the palm tree does in the art of the following centuries. THE CELESTIAL BANQUET A more definite representation of the refreshment of the soul in heaven was given by the pictures of the celestial ban- quet. The idea was obviously suggested by many of our Lord's sayings, and its representation in Christian art was perhaps facilitated by the fact that it had been a favorite theme in some of the pagan religions. The earliest example is in the vestibule of S. Domitilla. It represents two persons seated before a small table upon which fish and bread are dis- played ; an attendant stands ready to minister to them. The whole is so realistic that it might be taken for a scene of everyday life — perhaps for a funeral agape — were it not for the mystic fish upon the table. Most of these scenes belong to a single region of the catacomb of SS. Pietro e Marcellino, and are probably to be dated about the end of the third cen- tury. There are five of these pictures still discernible, and they have a striking similarity which proves either the hand of the same artist or at least a common influence. They rep- resent a family feast in which parents and children are seated about a semicircular table in front of which is a tripod bearing the mystic fish. Above the figures of two of the party are in- scribed the commands which they are supposed to be address- ing to the two maid-servants, " Peace, bring me warm water,' 7 "Love, fill my glass" (ibene da calda; agape misce mi). A painting of the same period in an arcosolium of the Ostrian cemetery represents the parable of the Wise Virgins. 222 PICTORIAL ART The scene is divided in two by an orans : on the right are the five Virgins holding lighted torches ; on the left they all appear seated at the marriage feast. The same parable is represented again in the fourth-century fresco of the cemetery of Cyriaca. In this case Christ with the nimbus appears in the middle ; on the left are five Virgins with lighted torches, while the five others on the right have their torches lowered Fig, 70. — Veneranda introduced into paradise by S. Petronilla, fresco in the catacomb of Domitilla. Fourth century. and extinguished. It is probable that both paintings marked the tombs of consecrated virgins. The introduction into heaven through the kind offices of a saintly " advocate " was sometimes represented in the third century ; it became a more common theme in the fourth, and was perpetuated in the apsidal mosaics of the basilicas. A picture of the fourth century in the catacomb of Domitilla (Fig. 70) represents a certain Veneranda introduced into PAINTING — The Eucharistic Banquet 223 paradise by S. Petronilla, whose tomb was near this spot. At the feet of the saint is a box containing the rolls of the Scrip- ture ; above it is an open book. THE EUCHARISTIC BANQUET The miraculous multiplication of the loaves and fishes was pic- tured very frequently in the catacombs, but in very various ways. This theme owed its great popularity to the fact that it stood as a symbol of the Eucharist, — a symbol which was suggested by Christ himself in the address which S. John reports after the miracle. A third-century fresco in the cripta delle pecorelle in S. Callistus represents it in the same manner as was com- mon on the sarcophagi : Christ stretches out his hands to bless the loaves and fishes which are presented to him by two Apostles ; on the ground stand six baskets filled with the fragments which remained. The figure of Christ has here been destroyed in order to construct a niche for a lamp, but his position and gesture were evidently the same as in Fig. 95 and on other sarcophagi. The gesture is important, for it is doubtless the same as that which was commonly used in the consecration of the Eucharist. In a picture of the latter part of the second century in one of the Sacrament Chapels of S. Callistus (Fig. 71) Christ is represented with right arm and shoulder bare — that is, clad in the philosopher's mantle — stretching his hand toward the bread and fish which are sup- ported upon a tripod. An orans standing by clearly expresses the soul which has departed from this world comforted with the Eucharist and confident in the almighty power of Christ. There were very many ways in which this subject was abbreviated ; it became a mere hieroglyph as it was treated in the graffiti of the gravestones ; the five loaves and the two fishes alone sufficed to represent at once the miracle and the sacrament. A second-century sarcophagus of travertine in the catacomb of S. Priscilla is adorned simply with five loaves on one side of the epitaph and an anchor on the other. Two fish and five loaves appear upon a tombstone which is preserved at Modena ; a similar monument is preserved in the Kircherian museum at Rome ; and one of the tiles covering a loculus in S. Priscilla has upon it two loaves moulded in mortar, — evi- 224 PICTORIAL ART dently a fragment of this same theme. On this last monument and on the epitaph at Modena the breads are marked with a cross, as they are also in several frescos which represent the seven baskets. On the sarcophagus above mentioned and on the Eoman epitaph they are marked by three lines crossing each other obliquely, which was a current symbol of the cross. Two terra-cotta lamps from Salona which represent the same Fig. 71. — Fresco of the third century, in one of the Sacrament Chapels, catacomb of Cfil- listus. Christ, clad in the philosopher's pallium, consecrates the fish and bread (symbol of the Eucharist) ; an orans stands beside the tripod. theme have the breads marked with the Constantinian mono- gram. But we must postpone other abbreviations of this theme till we come to consider the symbol of the fish. We are here engaged with the scenes which represent the banquet of the multitude. The most notable examples are in the Sacrament Chapels of S. Callistus. These so-called chapels are in reality ordinary cubicula which are notable only for the subtilty of their symbolical frescos. The name is due to the fashion which till lately has been prevalent of interpreting these frescos as symbols of all the seven sacraments of the Roman PAINTING — The Eucharistic Banquet 225 Church. It is now recognized that only baptism and the Eucharist are there expressed, and these subjects appear so often in the catacombs that they constitute no reason for giv- ing so distinctive a name to these chambers. There are six cubicula usually reckoned to this group ; Wilpert ascribes the first three of them to the same artist and to the latter part of Fig. 72. - Christ and the Samaritan woman at Jacob's well. Fresco in one of the Sacra- ment Chapels in the catacomb of Callistus. Third century. the second century; and the others, which are on a higher level, to the century following. The chief interest lies in the first group, and it is to be regret- ted that we cannot study in detail the numerous subjects which appear on their walls and ceilings, nor trace the deeply planned scheme which connects them. The scheme of decoration is substantially the same in all three chambers, in each the sym- Q 226 PICTORIAL ART bols of baptism and of the Eucharist are represented. In the third chamber is represented Christ's miracle, the multiplica- tion of the loaves and fishes (Fig. 71), and to the left of it the feed- ing of the multitude, a theme which was thor- oughly conventionalized in early Christian art. As the whole multitude could not be pictured, it sufficed to represent seven men seated at a table upon which were placed the loaves and fishes, and about which were arranged the seven baskets. The same scene was depicted in the first chamber. In both these cases, and in all others where this subject is represented, the men are clad in tunics. But in the second chamber, with but slight change of treatment, a different subject was represented, the meal of the seven disciples at the sea of Tiberias, when after his resurrection the Lord met them and put before them bread and fish. In this representation (Pig. 73) the seven baskets are naturally omitted, and the disciples, as fisher- men, are depicted naked. It may be that the num- -- A ^ a en e3 2 = v. a -3 -a r-.