r A TYTA7 /"i IP A B > H A AND mm of E . BALDWI Ulrich Middeldorf / EARLY CHRISTIAN ICONOGRAPHY AND A SCHOOL OL IVORY CARVERS IN PROVENCE Princeton Monographs in Art and Archaeology VI EARLY CHRISTIAN ICONOGRAPHY AND A SCHOOL OL IVORY CARVERS IN PROVENCE BY E. BALDWIN SMITH INSTRUCTOR IN ART AND ARCH.EOLOGY, PRINCETON UNIVERSITY PRINCETON UNIVERSITY PRESS PRINCETON LONDON : HUMPHREY MILFORD OXFORD UNIVERSITY PRESS 1918 Copyright, 1918 Princeton University Press Published June, 1918 PRINCETON .UNIVERSITY, \PRLSS / TO MY MOTHER Digitized by the Internet Archive in 2016 with funding from Getty Research Institute https://archive.org/details/earlychristianicOOsmit PREFACE No archaeological investigation is begun or successfully terminated in Princeton University which has not been inspired and made possible by the wisdom and kindness of Professor Allan Marquand. From the day when the present work was commenced, through the years of compilation and writing, Professor Morey has done all in his power to make it of scholarly value, and in the last stage of preparation, the author’s absence from the University has increased his obligation to his colleagues, for Professors Marquand and Morey have read all the proof and attended to even- detail of the publication. For the aid here acknowledged and for that which no words can acknowledge, the author now voices his gratitude to these two friends. Fie also wishes to express his thanks to Professor Butler, whose intimate knowledge of the Christian Orient has been freely placed at his disposal, and to the other members of the Department of Art and Archaeology, who have helped him in many generous ways. The illustrations are nearly all from tracings whose originals are cited in the notes or Tables; where plates or figures have been reproduced directly from other publications, acknowledgment has been made in the Fist of Illus- trations. E. Baldwin Smith. Fort Myer, Va. May 25, 1917. vii CONTENTS PAGE PART ONE— EARLY CHRISTIAN ICONOGRAPHY Preface. I — Introduction i - II — The Virgin in the Temple 7 - Ill — The Annunciation at the Spring n - IV — The Nativity 13 — - Y — The Appearance of the Star to the Magi 33 - YI — The Adoration of the Magi 36 -VII — The Massacre of the Innocents 59 jYIII — C hrist Among the Doctors 68 - IN — The Baptism 71 - N — The Miracle of Cana 85 “ NI — The Healing of the Blind 94 - Nil — The Healing of the Paralytic 102 — XIII — The Raising of Lazarus 108 XIV — The Entry into Jerusalem 121 XV— The Last Supper 129 — XVI — The Presentation of the Crowns 141 “ XVII — The Widow’s Mite 144 XVIII — Christ Triumphant 146 XIX — Summary of the Schools of Early Christian Art 159 PART TWO — A SCHOOL OF IVORY CARVERS IN PROVENCE I — Introduction 187 II — The Orientalizing of Gaul 192 III — Ivory Book Covers in Milan Cathedral 206 I\ — Werden Casket in the Victoria and Albert Museum 221 V — Diptych in the Library of Rouen 231 \ I — Two Ivory Fragments in the Berlin and Nevers Museums. . 237 \ II — Ivory Book Cover in the Bodleian Library, Oxford 248 IX ILLUSTRATIONS FIGURE 1 — Virgin in the Temple. Book Covers, Milan Cathedral 2 — Virgin in the Temple. Werden Casket; Victoria and Albert Museum, S. Kensington 3 — Annunciation at the Spring. Greek Gospels; Bibl. nat., gr. 74, Paris 4 — Annunciation at the Spring. Book Covers, Milan Cathedral. . . 5 — Annunciation at the Spring. Werden Casket; Victoria and Al- bert Museum, S. Kensington 6 — Adoration of the Ox and the Ass. Fresco; San Sebastiano, Rome 7 — Adoration of the Ox and the Ass. Sarcophagus, Milan 8 — Adoration of Animals and the Shepherds. Sarcophagus, Rome 9 — Adoration in a Manger Shed. Sarcophagus, Rome 10 — Virgin Introduced into the Adoration. Sarcophagus, Mantua. . 11 — Adoration Including the Magi. Sarcophagus, Rome 12 — Adoration Type of Nativity. Book Covers, Milan Cathedral. . 13 — Adoration Type of the Nativity. Werden Casket; Victoria and Albert Museum, S. Kensington 14 — Combination Scene of Adoration. Sarcophagus, Arles 15 — Adoration Type of Nativity, and Flight into Egypt. Encolpium, Constantinople (from Strzygowski : Byzantinische Denk- mdlcr, I) 16 — Nativity. Rabula Gospels; Laurentian Library, Florence 17 — Nativity. Panel from Chair of Maximianus, Rome 18 — Nativity. Book Cover, British Museum 19 — Nativity. Pyxis, Vienna 20 — Nativity. Pyxis from Minden, Berlin Museum 21 — Nativity. Fresco; Bawit, Egypt 22 — Nativity. Ampulla, Monza 23 — Nativity. Coptic Panel; Golenisheff Collection, Petrograd . . . . 24 — Nativity. Greek Lectionary, Berlin 25 — Nativity. Greek Gospels; Bibl. nat., gr. 74, Paris 26 — Appearance of the Star to the Magi. Sarcophagus, Arles 27 — Appearance of the Star to the Magi. Book Covers, Milan Cathe- dral 28 — Appearance of the Star to the Magi. Werden Casket; Victoria and Albert Museum, S. Kensington 29 — ‘‘The Coming of the Orient to Christ.” Mosaic; S. Maria Mag- giore, Rome 3° — Adoration of the Magi. Fresco: Catacomb of Callixtus, Rome xi PAGE 8 9 1 1 12 12 14 14 15 U 15 16 16 1 7 21 23 24 25 26 26 26 27 29 30 31 32 34 35 36 37 41 ILLUSTRATIONS xii FIGURE 31 — Adoration of the Magi. Fresco; Catacomb of SS. Pietro e Marcellino, Rome 32 — Adoration of the Magi. Sarcophagus, Ancona 33 — Adoration of the Magi. Sarcophagus, Rome 34 — Adoration of the Magi. Wooden Doors; S. Sabina, Rome. . . . 35 — Adoration of the Magi. Book Covers, Milan Cathedral 36 — Adoration of the Magi. Werden Casket ; Victoria and Albert Museum, S. Kensington 37 — Adoration of the Magi. Sacramentary of Drogo; Bibl. nat., lat. 9428, Paris 38 — “Epiphany Type” of Adoration. Lombard Relief, Berlin (from Kehrer : Die Heil. drci Konige ) 39 — Adoration of the Magi. Encolpium, Constantinople (from Strzygowski : Byzantinischc Denkmaler, I) 40 — Adoration of the Magi (part). Chair of Maximianus, Ravenna 41 — Adoration of the Magi. Coptic Textile from Akmim 42 — Adoration of the Magi. Homilies of Gregory of Nazianzus ; Bibl. nat., gr. 510, Paris 43 — -Adoration of the Magi. Ampulla, Monza 44 — Adoration. Etschmiadzin Gospels; Monastery of Etschmiadzin, Armenia 45 — Adoration and Nativity. Fragment of Murano Book Covers; John Rylands Library, Manchester 46 — Adoration of the Magi. Fragment of Five Part Coptic Dip- tych, British Museum 47 — Adoration and Appearance of the Star to the Magi. Codex Egberti, Trier 48 — Adoration of the Magi. Menologium of Basil II; Vatican Li- brary, gr. 1613 (from Kehrer: Die Heil. drci Konige ) 49 — Massacre of the Innocents. Mosaic; S. Maria Maggiore, Rome 50 — Massacre of the Innocents. Rabula Gospels ; Laurentian Li- brary, Florence 51 — Massacre of the Innocents. Fresco, Antinoe 52 — Massacre of the Innocents. Fragment of Five Part Diptych, Berlin Museum 53 — Massacre of the Innocents. Book Covers, Milan Cathedral. . . . 54 — Massacre of the Innocents. Ms. Cim. 2, Munich 55 — Massacre of the Innocents. Carolingian Book Cover; Bibl. Nat., Paris 56 — Massacre of the Innocents. Sarcophagus ; S. Maximin, Provence 57 — Christ Among the Doctors. Greek Gospels; Bibl. nat., gr. 74, Paris 58 — Christ Among the Doctors. Book Covers, Milan Cathedral . . . 59 — Baptism. Fresco; Cemetery of Lucina, Rome 60 — Baptism. Sarcophagus, Rome 61 — Baptism. Sarcophagus, Soissons 62 — Baptism. Book Covers, Milan Cathedral 63 — Baptism. Werden Casket; Victoria and Albert Museum, S. Kensington PAGE 41 4 2 42 43 44 44 45 45 48 50 5 1 5 2 53 54 55 56 58 59 60 61 62 63 63 64 65 66 70 71 73 74 75 76 77 ILLUSTRATIONS xiii FIGURE PAGE 64 — Baptism. Rabula Gospels; Laurentian Library, Florence 78 65 — Baptism. Etschmiadzin Gospels; Monastery of Etschmiadzin, Armenia 79 66 — Baptism. Panel from Chair of Maximianus, Milan 80 67 — Baptism. Ivory; British Museum (from Dalton: Cat. Early Christ. AnL Brit. Mus .) 80 68 — Baptism. Fresco; Bawit, Egypt 81 69 — Baptism. Mosaic; Arian Baptistery, Ravenna 82 70 — Baptism. Ampulla, Monza 83 71 — Baptism. Coptic Panel; Golenisheff Collection, Petrograd . . . . 84 72 — Miracle of Cana. Sarcophagus, Algiers 86 73 — Miracle of Cana. Sarcophagus, Rome 87 74 — Raising of Lazarus and Miracle of Cana. “Roman Glass,” Rome 88 75 — Miracle of Cana. Chair of Maximianus, Ravenna 89 76 — Miracle of Cana. Medallion from Egypt ; Cans Collection, Berlin 90 77 — Miracle of Cana and the Samaritan Woman at the Well. Mosaic, Baptistery of Naples '. 91 78 — Miracle of Cana. Book Covers, Milan Cathedral 91 79 — Miracle of Cana. Rabula Gospels ; Laurentian Library, Florence 92 80 — Miracle of Cana. Fresco, Antinoe 93 81 — Miracle of Cana. Soissons Gospels; Bibl. Nat., Paris 94 82 — Miracle of Cana. Greek Gospels; Bibl. nat., gr. 74, Paris 95 83 — Healing of the Blind. Sarcophagus, Rome 96 84 — Healing of the Blind and the Halt. Chair of Maximianus, Ravenna 97 85 — Healing of the Blind. Murano Book Cover, Ravenna 98 86 — Healing of the Blind and the Halt. Rabula Gospels ; Laurentian Library, Florence 99 87 — Healing of the Blind. Sinope Fragment; Bibl. nat., Paris. ... 99 88 — Healing of the Blind. Book Covers, Milan Cathedral 100 89 — Healing of the Blind. Codex Rossanensis, Rossano 101 90 — Healing of the Blind. Homilies of Gregory of Nazianzus; Bibl. nat., gr. 510, Paris 102 91 — Healing of the Paralytic. Fresco; Catacomb of Callixtus, Rome 103 92 — Healing of the Paralytic. Sarcophagus, Rome 104 93 — Healing of the Paralytic. Book Covers, Milan Cathedral 105 94 — Healing of the Paralytic. Book Covers; Bibl. nat., Paris 106 95 — Healing of the Paralytic. Murano Book Cover, Ravenna 107 96 — Healing of the Paralytic. Fresco; S. Saba, Rome 107 97 — Healing of the Paralytic. Greek Gospels; Bibl. nat., gr. 74, Paris 108 98 — Raising of Lazarus. Fresco; Catacomb of Callixtus, Rome. ... in 99 — Raising of Lazarus. Book Covers, Milan Cathedral 112 100 — Raising of Lazarus. Sarcophagus, Aix 113 101 — Raising of Lazarus. Sarcophagus, Rome 114 102 — Raising of Lazarus. Relief from Phrygia, Constantinople 1 1 5 103 — Raising of Lazarus. Sarcophagus, Ravenna 116 104 — Raising of Lazarus. Book Covers; Bibl. nat., Paris 117 105 — Raising of Lazarus. Ivory Comb from Antinoe, Cairo t 1 8 xiv ILLUSTRATIONS FIGURE PAGE 106 — Raising of Lazarus. Pyxis; Cluny Museum, Paris 118 107 — Raising of Lazarus. Pyxis; Basilewsky Collection, Petrograd.. 119 108 — Raising of Lazarus. Pyxis, Vatican 120 109 — Raising of Lazarus. Murano Book Cover, Ravenna 120 no — Raising of Lazarus. Codex Rossanensis, Rossano 121 in — Raising of Lazarus. Homilies of Gregory of Nazianzus; Bibl. nat., gr. 510, Paris 122 1 12 — Entry into Jerusalem. Sarcophagus, Rome 123 1 13 — Entry into Jerusalem. Book Covers, Milan Cathedral 124 1 14 — Entry into Jerusalem. Panel from Chair of Maximianus, Rome 125 1 15 — Entry into Jerusalem. Book Covers; Bibl. nat., Paris 126 1 16 — Entry into Jerusalem. Wooden Lintel, Church of el Mu’allaka near Cairo 126 1 17 — Entry into Jerusalem. Rabula Gospels; Laurentian Library, Florence 127 1 18 — Entry into Jerusalem. Codex Rossanensis, Rossano 128 1 19 — Entry into Jerusalem. Homilies of Gregory of Nazianzus; Bibl. nat., gr. 510, Paris 129 120 — Eucharistic Symbols. Fresco; Cemetery of Lucina, Rome.... 130 121 — Eucharistic Symbols. Fresco; Catacomb of Callixtus, Rome. . . 131 122 — Miracle of Loaves and Fishes. Sarcophagus, Algiers 132 123— — Miracle of Loaves and Fishes. Sarcophagus, Rome 132 124 — Miracle of Loaves and Fishes and Feeding of the Multitude. Fresco, Catacombs of Alexandria (from Cabrol : Diet, d'archeologie chret.) 133 125 — Miracle of Loaves and Fishes and Feeding of the Multitude. Chair of Maximianus, Ravenna 134 126 — Eucharistic Feast. Fresco; Catacomb of SS. Pietro e Marcel- lino, Rome 135 127 — Transitional Type of Last Supper. Mosaic; S. Apollinare Nuovo, Ravenna 136 128 — Transitional Type of the Last Supper. Book Covers, Milan Cathedral 137 129 — Communion of the Apostles. Rabula Gospels; Laurentian Li- brary, Florence 138 130 — Communion of the Apostles. Codex Rossanensis, Rossano. ... 139 131 — Last Supper. Codex Rossanensis, Rossano 141 132 — Presentation of the Crowns. Book Covers, Milan Cathedral. . . 142 133 — Traditio Legis. Mosaic; S. Costanza, Rome 143 134 — Widow’s Mite. Book Covers, Milan Cathedral 145 135 — Widow’s Mite. Mosaic; S. Apollinare Nuovo, Ravenna 146 136 — Christ Triumphant. Fresco; Catacombs of Alexandria 147 137 — Horns Relief, Cairo 149 138 — Christ Triumphant. Lamp from Akmim 150 139 — Christ Triumphant. Lamp from Akmim 151 140 — Christ Triumphant. Vase from Orleans 152 141 — Christ Triumphant. Pignatta Sarcophagus, Ravenna 153 142 — Christ Triumphant. Ivory Book Cover of the Lorsch Gospels, Vatican 155 ILLUSTRATIONS xv FIGURE PAGE 143 — Christ Triumphant. Genoels-Elderen Diptych, Brussels (from Laurent: Les Ivoires pre-gothiques ) 156 144 — Annunciation. Chair of Maximianus, Ravenna 171 145 — .Annunciation. Fragment of the Murano Book Covers; Stro- ganoff Collection, Rome 172 146 — Annunciation. Medallion from Egypt; Gans Collection, Berlin 172 147 — Annunciation. Ampulla, Monza 172 148 — Testing of the Virgin. Chair of Maximianus, Ravenna 173 149 — Testing of the Virgin. Fragment of the Murano Book Covers; Stroganoff Collection, Rome 174 150 — Joseph Assured by an Angel and the Journey to Bethlehem. Chair of Maximianus, Ravenna 175 15 1 — Joseph Assured by an Angel and the Journey to Bethlehem. Fresco, Antinoe (from Cabrol: Diet, d’archeologie chret.) 176 152 — Journey to Bethlehem. Pyxis from Minden, Berlin Museum. . . 177 153 — Virgin in the Temple. Inscribed Slab, St. Maximin 20 2 154 — Joseph Assured by an Angel and Reconciliation of Joseph and Alary. Sarcophagus, Le Puy 203 155 — Ivory Book Cover. Milan Cathedral (from Labarte: Hist, des arts industrials ) 207 156 — Ivory Book Cover. Milan Cathedral (from Garrucci : Storia del!' arte crist.) 209 157 — Werden Casket (Front). Victoria and Albert Museum, S. Kensington (from Garrucci: Storia dell' arte crist.) 222 158 — Werden Casket (Back). Victoria and Albert Museum, S. Kensington (from Jahrb. Preuss. Kunsts., XXIV) 226 159 — Werden Casket (Side). Victoria and Albert Museum, S. Ken- sington (from Garrucci: Storia dell’ arte crist.) 228 160 — Sts. Gervasius and Protasius before the City of Le Mans. En- graved Gem 230 161 — Ivory Diptych. Public Library, Rouen (from Gazette Archeo- logique 1886) 233 162 — Christ and Apostles. Sarcophagus, Marseilles (from Le Blant: Les Sarcophages chret. de la Gaule) 234 163 — Fragment of Ivory Diptych. Berlin Museum (from Jahrb. Preuss. Kunsts. XXIV) 239 164 — Xativity and Adoration of the Magi. Ivory, Nevers Museum. . 243 165 — Diptych of Probianus. Berlin Museum (from Jahrb. Preuss. Kunsts. XXIV) 245 166 — Sarcophagus in Marseilles Museum (from Le Blant: Les Sar- cophages chret. de la Gaule) 246 167 — Sarcophagus at St. Maximin (from Le Blant: Les Sarcophages chret. de la Gaule) 247 168 — Sarcophagus of Arles (from Le Blant: Les Sarcophages chret. de la Gaule) 248 169 — Ivory Book Cover. Bodleian Library, Oxford (from West- wood: Fictile Ivories) 250 y PART I EARLY CHRISTIAN ICONOGRAPHY I INTRODUCTION In the course of the following pages I have attempted to attain three results : in the first place to group, on a basis of iconography, the monu- ments of early Christian art into schools; then, as a result of this forma- tion of distinct and consistent art centres, to prove the provenience of certain monuments whose origin has been vigorously disputed ; and lastly, I have sought in the Second Part to establish a school of early Christian ivory carvers in Provence. In treating early Christian art I have under- stood the term “early Christian" to include the artistic efforts of Christian- ity from its inception down to the ninth century, when Byzantine art was formulated and generally adopted throughout the East. Thus I have en- deavored to distinguish, wherever it was possible, the iconographic types of Egypt, Syria, Asia Minor, Italy, and Provence and, on the basis of these types, have built up the artistic centres which flourished on the shores of the Mediterranean during the first eight centuries of our era. With the exception of the Provenqal, the schools have been constructed wholly from the iconography and the scenes used have been only those that occur on the monuments which I would attribute to Provence. Although many brilliant works have been written on the subject , 1 the art of the early Christian period is still so badly confused by many con- flicting theories that any systematic study in the field must go contrary to some authorities. Ever since the predominance of Rome in the forma- tion of Christian art was questioned by Strzygowski 2 the problem of the origin, spread, and development of this art has become increasingly diffi- 1 See Dalton’s Byzantine Art and Archaeology for a bibliography. 2 Strzygowski, Orient oder Rom, igoi, Leipzig. i EARLY CHRISTIAN ICONOGRAPHY 2 cult. While the position of Rome has been more and more weakly de- fended as new finds and excavations have shown that the great creative centres of Christian art were in the Orient, the question has become in- volved by the contradictory viewpoints from which the archaeologists have viewed the characteristics of these Eastern schools and by the confusing manner in which they have attributed the same or similar monuments to different schools. The principal reason for this confusion has been the lack of sufficient monuments of known origin from which to construct the schools. Other causes, however, lie not only in the gradual transforma- tion through which Christian art went in passing from a Hellenistic to an Oriental phase, but also in the uniform and vaguely defined schools to which the scholars attribute the monuments, and in the unsystematic use of iconography. The change that took place in Christian art as it discarded Hellenistic models and adopted Oriental forms has complicated the problem. At the time when Christianity was born, an Eastern cult in an Eastern land, the Mediterranean world was under classical domination. The only forms of expression which were available to the new faith were Hellenistic and, therefore, Christianity for the first four centuries of its existence followed Hellenistic models in its literature and art. In the fifth century a change took place in the Empire, the state crumbled under the barbarian invasions, the centre of gravity was shifted from Italy to the East, and the classical system which had until then unified the civilized world was shattered. Hellenism was dying, the dormant spirit of the Orient re- awakened. Great waves of national feeling swept over the Eastern prov- inces with the result that in literature and in art the old, native traditions began to assert themselves against the waning classicism. For about four centuries Hellenistic Christianity had accepted the canonical Gospels and by symbolism in art had sought to portray in classical forms the Biblical lessons. With the dissolution of classical restraint, however, the East turned to its apocryphal versions of the New Testament stories wherein were related more personal and narrative accounts of Jesus and the Virgin. These apocryphal gospels became a new source for artistic representation and the new scenes which were then created were expressed in indigenous Eastern methods. In Asia Minor, Syria, and Egypt the art took on the character of its Sassanian, Assyrian, and Nilotic prototypes. The figures tended to become stylistic, hieratic, and frontal; the substantial back- EARLY CHRISTIAN ICONOGRAPHY 3 grounds of Hellenistic design were eliminated; the composition began to set into hard and fast formulas; and the execution became linear and coloristic in treatment. The Christian art of the Hellenistic period, with general classical and ca- nonical models, was strikingly homogeneous throughout the Mediterranean and while the forms and the types of the new art, like the religion itself, were probably first established in the great Hellenistic cities of the East, they were at an early date adopted at Rome and became the traditional style of the West. With the assertion of Oriental tendencies in the fifth century art broke up into local schools throughout the Eastern provinces. No sharp line of demarcation, however, divides the Hellenistic from the Oriental and it was only gradually, and often only in part, that the change in Christian art took place. The centres of Hellenistic art, like Antioch, Alexandria, and Ephesus, were not invariably the centres of the indigenous arts of the countries. In Syria the artistic activity shifted to Palestine and in Egypt it passed from Alexandria to the monastic centres in the Thebaid. Thus an endless confusion arises when one attempts to unravel the threads of early Christian art while it was passing through its period of readjustment. The solution of the problem can only come when archaeologists agree on the characteristics, the boundaries, and the limitations of the different artistic centres. At present the monuments, after the Hellenistic period, are variously assigned to Constantinople. Asia Minor, Syria, Palestine, Upper Egypt, and the eclectic school of Ravenna. In many cases thev are also attributed in a general way to broader divisions, Syro- Anatolian, Syro-Palestinian, Syro-Egyptian, and Palestinian-Coptic ; and in the West, which the Orient soon invaded, minor places, as Milan and Monza, are cited as productive. From these possibilities each authority, after his own conception of their character and in many cases after his own intuition, has built up by attribution his own set of schools. Unless, therefore, the student treads carefully through this maze, the result can only be greater chaos. The baffling nature of these Eastern schools is clearly realized when a single monument is studied from the various authorities. No more im- portant example of early Christian art exists than the famous ivory chair at Ravenna, which is supposed to have been the cathedra of the Archbishop Maximianus. It has been attributed to nearly every artistic centre. About 4 EARLY CHRISTIAN ICONOGRAPHY fifty years ago it was claimed by Labarte 3 that the chair was commenced at Constantinople and finished at Ravenna by Greek artists. Its origin in Ravenna has been supported by Stuhlfauth , 4 while Venturi , 5 without deny- ing its Byzantine origin, preferred to consider it a work analogous to the mosaics of Santa Maria Maggiore at Rome. Of recent years Strzygowski , 6 who has perhaps written most on the subject, attributed it to Antioch. Its Alexandrian origin, however, supported by Ainaloff , 7 Diehl , 8 Leclercq , 9 and Graeven 10 has been most widely accepted. Dalton leaves the choice open between Alexandria and Antioch . 11 While it is evident that the establishment of these schools by attribution is a difficult and baffling problem, the lack of uniformity in the results is partly due to the unsystematic use that has been made of the means of attribution. These are mainly style and iconography. Of the two I am inclined to believe that style, except in the Palestinian and Coptic schools, is the less reliable. This is certainly true when the monuments preserve to any marked degree a trace of the rather uniform Hellenistic tradition. The iconography, however, which was dependent on indigenous customs, local liturgies, distinctive ceremonies, and special apocryphal gospels, tended to formalize into types and become peculiar to the separate artistic centres. Even when these resultant types and the use of particular scenes in specific localities were transmitted from one centre to another at the hands of itinerant artists, a study of the local customs, liturgies, apocryphal gospels, and monuments as a whole will often locate the centre from which the type emanated and will help to disclose the lines of inter- development along which the schools progressed. Too often, however, the iconography, as well as the style, has been used by writers in a small and partial way. An attribution is made on a few iconographic types and stylistic features; the attribution is then used independently as an estab- lished fact for further attributions. Since many of the schools preserved certain features in common and since there was inevitably an interchange 3 Labarte, Histoire des arts industriels, I, pp. 12-14. 4 Stuhlfauth, Die altchristliche Elfenbeinplastik, p. 86. 5 Venturi, Storia dell’ arte Italiana, I, p. 468, figs. 278-307. 8 Strzygowski, Journal of Hellenic Studies, 1907, p. 115 (trans. by Mrs. Strong). 7 Ainaloff, see Dalton’s Byzantine Art and Archaeology, p. 206. 8 Diehl, Manuel d’art byzantin, p. 281. 9 Leclercq, Manuel dc l’ art chretien, II, p. 352. 10 Graeven, Bonner Jahrbucher, 1900, pp. 159, 162. 11 Dalton, op. cit., p. 206. EARLY CHRISTIAN ICONOGRAPHY 5 of iconographic types, this method makes possible an unlimited addition of monuments to almost any artistic centre and assures no uniformity in the “centre” when “established.” In all cases the ultimate soundness of attributions must depend on comparison with those monuments which are known to have come from particular regions. The nuclei about which the schools are to be constructed are often very meagre. For Asia Minor or the proto-Byzantine “school” we have the Sidamara type of sarcophagi, 12 the Rossano Gospels, 13 the Sinope fragment of Matthew, 14 and possibly the mosaics of the Church of the Holy Apostles at Constantinople as they are described by writers of the tenth and twelfth centuries 15 Save for the fact that the Byzantine types appear to have been derived from the scenes of the Rossano Gospels, the Gospels them- selves were only attributions until the Sinope fragment was discovered. The Rabula Gospels, 16 certain miniatures of the Etschmiadzin Gospels, 17 the Syrian Gospels of the sixth century from Mardin in the Bibliotheque Nationale at Paris, and the decoration of the Syrian churches give the basis of the Syrian school wherein the transitional, semi-Hellenistic tradi- tion is represented by the Rabula Gospels and the Oriental tradition by the Etschmiadzin miniatures. The style and types of the Palestinian school are exemplified by the Monza phials 18 from the Holy Land, and by what is known of the mosaics of the Churches of the Nativity and the Ascen- sion 19 For Upper Egypt we have the frescoes of Antinoe, 20 Bawit, 21 12 Reinach, Monuments Plot, 1902, p. 189; Strzygowski, Journal of Hellenic Studies, 1907, and Orient oder Rom; A. Munoz, Nuovo Bullettino di arch, cristiana, XI, 1905, and L’Arte, 1906, p. 130; Dalton, op. cit., p. 129; G. Mendel, Bull. Corr.Hell. XXXIII, 1909, pp. 333-334. and Catalogue des Sculptures (Musees Imp. Ottomans), I, p. 312; to these should be added the sarcophagus of the same type found by the American Expedition under Mr. Butler at Sardis. 13 Gebhardt and Harnack, Erangeliorum Codex graecus purpureus Rossanensis, 1880; Haseloff, Codex Rossanensis. 14 Omont, Monuments Piot, VII, 1900, pp. 175-185, pis. XVI-XIX. For the relation of the Sinope Fragment to the Rossano Gospels and of both the manuscripts to Asia Minor, see J. A. Herbert, Illuminated Manuscripts, pp. 22-30. 15 Heisenberg, Graheskirche und Apostelkirche, 1908; Diehl, Manuel, pp. 449-541. 18 Garrucci, Storia dell’ arte cristiana, III, pis. 128-140. 17 Strzygowski, Byzantinische Denkmaler, I, pis. II- VI. 19 Garrucci, Storia, VI, pis. 433-435; Dalton, op. cit., p. 623 sq. 19 Heisenberg, op. cit. ; Dalton, op. cit., pp. 384, 414. 20 Leclercq in Cabrol’s Dictionnaire d’archeologie chretienne et de liturgie, s. v. “An- tinoe,” col. 2326-2359; Cledat, Bulletin de I’Institut frangais d’archeologie orientate, 1902, II. 21 Leclercq, op. cit . s. v. “Baouit,” col. 203-251; Cledat, Comptes Rendus de I'Academie des Inscr., 1904. 6 EARLY CHRISTIAN ICONOGRAPHY el Bagawat , 22 the miniatures of the World Chronicle , 23 the covers of the Freer Gospels , 24 the decorative stuffs and minor objects from Akmim , 25 and excavations at other sites . 26 For Alexandria the catacomb frescoes , 27 the numerous pieces of wood and ivory carving in the Cairo, Berlin, and Paris Museums, and the compositions in the miniatures of the Cosmas Indicopleustes 28 are characteristic. In the West at Ravenna are the sar- cophagi and mosaics , 29 at Rome the catacomb frescoes , 30 the mosaics , 31 and sarcophagi , 32 and for Provence we have only the sarcophagi. Fhe schools (or rather groups, for in many cases I have not attempted to locate certain types) which I have constructed from these monuments are the Hellenistic, the Oriental-Hellenistic, the Syrian, the Alexandrian- Coptic, the Palestinian-Coptic, the Provencal, and the proto-Byzantine. The work at Ravenna, which is not uniform, was found to be eclectic and I have only sought to show the origin of some of the iconographic types occurring there. The classification, with the exception of the Provenqal group, has been made up largely on the basis of iconography and, in order that I should not use this means in a partial or limited way, I have taken the scenes occurring on those monuments which I would attribute to Provence and, by grouping together all the examples of these scenes ac- cording to the types, I have attempted to let them fall naturally into what- ever groups they would. In this way I thought if there was trustworthy evidence in the iconography, as I believed there was, the schools would preserve a marked consistency. For this purpose I have used the classified tables in the back of the book, in which the more complicated scenes, show- ing many and varied types, come out more clearly. The dates assigned to the monuments are for the most part those which have been generally 22 Leclercq, op. cit., s. v. “Bagaouat,” col. 31-62; W. de Bock, Matcriaux pour scrvir a Varcheologie dc I’Egypte chretienne, 1901. 23 Strzygowski, “Eine alexandrinische Weltchronik,” Denkschriften dcr kaiserlichen Akademie dcr Wissenschaften (Philosophisch-historische Klasse), 1906, p. 169 sq. 24 C. R. Morey, East Christian Paintings in the Freer Collection, 1914, pp. 63-81. 25 Forrer, Die fruhchristlichen Alterthiimer aits dem Gr'dherf elde von Achmim-Panopolis. 26 Strzygowski, Koptische Kunst, Vienna, 1904; Crum, Catalogue general du musee du Caire, 1902. 27 Neroutsos-Bey, L’ancienne Alexandria, Paris, 1888. 28 Garrucci, Storia, III, pis. 142-153. 29 Dalton, op. cit., pp. 342-368. 30 Wilpert, Lc pitturc dclle catacombe Romane. 31 De Rossi, Musaici cristiani di Roma. 32 Garrucci, Storia, V ; Marucchi, I monumenti del Museo cristiano Pio-Latcranense, 1910. EARLY CHRISTIAN ICONOGRAPHY 7 accepted by the authorities. In other cases I have endeavored to fix the period of execution on the basis of the evidence available. While some- times it has been necessary to accept theoretical attributions, I have in the main been able to evade the danger of building up my schools from one attribution to another and from forcing the attributions on only a few stylistic and iconographic analogies. The definition, by the method above described, of the schools of early Christian art has isolated the iconography of a group of ivories which I have assigned to Provence. The style as well as the iconography of this group shows such a peculiar mixture of the Hellenistic art of Rome and the Oriental types of the East that they do not fit into any of the schools already recognized. By comparing the iconography and style of these ivories with the sarcophagi of Provence, which manifest the same eclectic blending of Hellenistic traditions and Oriental types as occurs on the ivories, I have been able to point out such analogies as prove the exist- ence of a school of ivory carvers in Provence. II THE VIRGIN IN THE TEMPLE Monuments : (i) Relief in crypt of St. Maximin, V Century (2) Ivory book covers, Milan cathedral, c. 500 Provencal Group ( 3 ) Werden casket, S. Kensington. VI Century J The representations of the Virgin in the Temple are peculiar in the sixth century to that group of ivories whose origin in southern Gaul will become evident in the course of the discussion of this and succeeding scenes. I shall hereafter refer to it as the Provencal group. That the scene on the Milan book cover 1 represents the Virgin in the Temple (Fig. 1 ) is not open to doubt, although it has been interpreted as the Holy Women at the Sepulchre, after the account in John. Such an interpretation is impossible when the costume of the Virgin is considered. Instead of being wrapped in a long mantle and heavily veiled, in token of grief, as are all the Holy Women in the scenes of the Visit to the Sepulchre, 1 Garr., op. cit., VI, pi. 455. 8 EARLY CHRISTIAN ICONOGRAPHY the Virgin is depicted as a young girl elegantly clad in a long gown, bound at the waist by a girdle and handsomely enriched with heavy embroidery around the neck and down the front of the skirt. Her hair, instead of being bound in mature manner, is arranged in a most youthful and modish covers. The Virgin in the Temple. fashion. Never in the sixth century or later would one of the mourning women, sadly visiting at dawn at the tomb of the Savior, have been repre- sented as a young, handsomely clad maiden, as is the female figure on the Milan ivory. Furthermore, this interpretation is confirmed by the scene on the Wer- den casket 2 in the South Kensington Museum (Fig. 2), which in all scenes that also occur on the Milan covers almost duplicates their iconog- raphy. Although the representation and arrangement of the figures of the Virgin and the angel, as well as the odd little temple, are identical on the two ivories, there is introduced on the Werden casket a priest reading by the steps of the temple. Aside from this difference, the two scenes are so much alike as to necessitate the same interpretation. The scene on the Werden casket must be taken as a representation of the Virgin in the Temple, not only because of the introduction of the priest, but also because the range of its scenes covers only the early life of the Virgin; hence the scene on the Milan ivory must depict the same subject. Interpreted thus, the episode appears to be derived from the account in 2 Garr., op, cit., VI, pi. 447/1. 9 EARLY CHRISTIAN ICONOGRAPHY the apocryphal gospel of Pseudo-Matthew. V hile the canonical Gospels do not treat the early life of the Virgin, both the Protevangelium of James 3 and the Gospel of Pseudo-Matthew 4 relate that from the age of three Mary was employed in the temple in the service of the Lord. \\ hereas the Protevan- gelium gives the main facts of her presentation in the temple and of her subsequent life there, the Gospel of Pseudo-Matthew, in its detailed narra- tion not only of the presentation, but also of the devout life which Mary Fig. 2. S. Kensington: Victoria and Al- bert Museum, ivory casket. The Virgin in the Temple. led in the temple, reflects more clearly the full significance attached to this episode by the Eastern church and later by the church of Gaul. It is also the only account that explains the presence of the angel in the composition. Moreover, in its dependence on this gospel I do not feel that the scene rep- resents the Presentation in the Temple, as Leclercq says , 5 so much as the Virgin's actual service in the temple. Neither of the apocryphal gospels mentions an angel as having led the child to the temple. Yet on both the ivories the Virgin is represented looking at an angel who directs her gaze to a star hanging above the temple. With this variation from the text it does not seem justifiable to explain the absence of her parents on the ground of lack of space, as Leclercq does in considering this scene an Entry into the Temple. In my opinion the scene depicts the \ irgin during her service in the temple, after the account of Pseudo- Matthew, where we read : “At the age of three years she walked with a step so sure, she spoke so perfectly, and put so much ardor into praising 3 Michel, Evangiles apocryphes, I, p. 1 7. 4 Michel, op. cit., p. 73. Leclercq in Cabrol’s Dictionnairc d’archeologic cliretienne et de liturgie, s. v. “Apocryphes,” col. 2557. IO EARLY CHRISTIAN ICONOGRAPHY God, that one would not have taken her for a child, but for a grown person.” 6 We also read that “she did not cease to pray until the moment when the angel of the Lord appeared to her; she received nourishment from his hand; and she learned better and better how to praise God.” 7 Further on, the account relates : “Often angels were seen with her and she obeyed them with the greatest affection.” 8 This scene of the Service of the Virgin in the Temple, which does not appear on the monuments of the East, may be associated directly with Provence. When the crypt of the church of St. Maximin was unearthed there was brought to light an inscribed slab (Fig. 153) of the fifth century. On this slab 9 occurs a representation of the Virgin in the Temple which is labelled by an inscription reading : MARIAVIRGO MINESTERDE TEM P VLOGFRO SALE While the representation on this relief, in which the Virgin stands in the attitude of an orant, does not recall the scene on the ivories, it shows that the subject, which must have been drawn from apocryphal sources, was known and used in the art of Provence. It was not used, as far as we know, by the early Christian artists of any other locality. The dependence of the scene of the two ivories on the Gospel of Pseudo-Matthew, which was a Western rendering of an Oriental gospel, sustains the connection with the West. A feast 10 of the Virgin existed in Provence as early as the sixth century. In Syria, whence Provence drew so many elements of its Christian culture, a particular liturgy 11 had been conceived in memory of the life which Mary spent in the temple preparing herself to be the mother of the Saviour; in Provence itself her cult was evidently popular enough to create a special scene commemorative of her preparation in the temple. 0 Michel, op. cit., p. 75. 7 Michel, op. cit., p. 75. 8 Michel, op. cit., p. 77. 9 Le Blant, Inscriptions chretiennes de la Ganle, II, no. 542, A; Les sarcophages chretiens de la Ganle, p. 148, pi. LVII/i. 10 Duchesne, Origines du culte chretien, p. 258. 11 Leclercq, op. cit., col. 2557. EARLY CHRISTIAN ICONOGRAPHY 1 1 III THE ANNUNCIATION AT THE SPRING Monuments : (1) Ivory book covers, Milan cathedral, c. 500 (2) Werden casket, S. Kensington, VI (3) Terra cotta medallion at Monza, VIII “ Provenqal Group. The Annunciation at the Spring, when an angel of the Lord first ap- peared unto Mary, has always been a rare scene in Christian art. So far as the existing examples of the scene throw any light on the question, it seems to have been peculiar, during the early Christian period, to the ivories which we assign to Provence, although it may have been introduced from the East, since we find it used in the ninth century in Byzantine manuscripts. Though it is true that the scene, as Dalton says, 1 “is not favored by later Byzantine artists, ” there are, none the less, several exam- ples after the ninth century. 2 It appears in a Greek Gospel of the eleventh century (gr. 74) in the Bibliotheque Nationale, Paris 3 (Fig. 3), in the Fig. 3. Paris: Bibl. Nat., Ms. gr. 74. The Annunciation at the Spring. Homilies of the Monk Jacobus in the Bibliotheque Nationale, Paris, 4 among the mosaics of Daphni, 5 and the mosaics of the church of Kahrie Djami at Constantinople. 6 The scene in its whole history does not vary materially from the earliest examples. On the Milan (Fig. 4) and Werden (Fig. 5) ivories the Virgin kneels at the fountain to fill her pitcher as the angel accosts her from be- 1 Dalton, Byzantine Art and Archaeology, p. 653. 2 De Waal, Rom. Quar., 1887, p. 180. 3 Fleury, La Saintc Vierge, I, pi. XV. 4 Bordier, Description des pcintures, etc., dcs manuscrits grecs de la Bibliotheque A ationale, 1883, no. 1208, F" 159 v°, p. 165; a copy of this manuscript is in the Vatican Library (gr. 1162). 5 Diehl, Manuel, fig. 243. 6 Dalton, op. cit., fig. 246. 12 EARLY CHRISTIAN ICONOGRAPHY hind; the Virgin, startled by the heavenly messenger, pauses in the act and turns her head in the direction whence the voice comes. Likewise, on the two ivories the fountain is represented as a spring gushing from a crevice in the rock . 7 The terra cotta medallion in the treasury of the Fig. 4. Milan : Cathedral, ivory BOOK COVERS. THE ANNUNCIATION AT the Spring. Fig. 5. S. Kensington: Victoria and Albert Museum, ivory casket. The Annunciation at the Spring. Royal Basilica at Monza shows a scene of the same order as that on the ivories , 8 at the right is represented a tree from whose roots flows the spring and kneeling before the spring with a vase in her hand is the Virgin, who turns her head in surprise to the angel. The date of this medallion, however, has been put as late as the eighth century, and the character of the work indicates a Palestinian origin. As to the Provencal origin of the two examples on the ivories, the scene itself offers no evidence. It would seem that the rendering, as in the case of so many of the other scenes, was drawn rather from the Gospel of 7 Stuhlfauth, Die altchristliche Elfenbeinplastik, p. 70, says that an Annunciation at the Spring occurs on a sarcophagus of Syracuse (Garr., op. cit., V, pi. 365/1). The scene to which he refers is essentially different from any other examples of the An- nunciation, for it shows a woman kneeling to draw water from a spring which rushes forth from the side of a mountain on whose summit is the head of a bearded man, while behind the woman stands a man stretching forth his hand in a gesture of com- mand. Garrucci (op. cit., p. 95) calls the scene Moses Striking Water from the Moun- tain and such it seems to be ; as he says, it was not customary during the fourth century to personify God by a human head, whereas it was traditional to personify mountains and rivers in this manner. The personification, therefore, is to be understood as Mt. Horeb and the gesture of the man, as also the posture of the woman, are to be ex- plained accordingly. 8 Leclercq, in Cabrol’s Dictionnaire d'archeologie chretienne ct de liturgie, s. v. “An- nonciation dans l’art,” col. 2262, fig. 765. EARLY CHRISTIAN ICONOGRAPHY !3 Pseudo-Matthew than from the Protevangelium of James, which are the only two texts mentioning this annunciation. While the Protevangelium only speaks of a voice having accosted her , 9 the Gospel of Pseudo-Matthew says that “the angel of the Lord” appeared to the Virgin while she was at the fountain filling her pitcher . 10 This dependence in Provence on the apocryphal Gospel of Pseudo-Matthew will become evident as we proceed. IV THE NATIVITY The representations of the Nativity are usually classified into two large divisions : The Western type which, after the Gospel narratives, depicts the event in a shed sheltering a manger, and the Eastern type which fol- lows the apocryphal accounts and represents the Birth either in the open or in a cave. Although this rough classification may aid in distinguishing the monuments of the East from those of the West, a more accurate and practical division may be made by considering other features in the manner of representing the scene of the Nativity during the early Christian period. The main divisions of this reclassification are the Hellenistic type, the Oriental-Hellenistic, the Alexandrian-Coptic, the Palestinian-Coptic, and the Byzantine type. The Hellenistic type is actually an Adoration type because, in addition to the Nativity proper, the scene includes an Adoration either of the ox and the ass, of the shepherds, or of the Magi. This Hellenistic type of the Adoration is characteristic of the monuments of the West, although with certain variations it occurs on several later Eastern monuments (and one early one) which are classified under the head of the Oriental-Hellenistic type. Another reason for calling this Western method the Adoration type is that the artists of the Hellenistic period of Christian art, especially in the West, were primarily concerned with symbolizing the universal worship which was accorded the Saviour by Tews, Gentiles, and the natural world. The Adoration type may be divided into five groups, each characterized by special features of iconography . 1 9 Michel, op. cit., p. 23. 10 Michel, op. cit., p. 87. 1 Schmid, Die Darstellung der Geburt Christi, Stuttgart, 1890; in this treatise the author has carefully gathered together a large number of examples and in the case of the Western scenes has successfully grouped them according to the variations in type. 14 EARLY CHRISTIAN ICONOGRAPHY Hellenistic or “Adoration" Type. Group I. Adoration of Ox and Ass alone. In this first group only the ox and the ass do homage to the new born Babe. The earliest example is in a fresco in the catacomb of S. Sebas- tiano (Fig. 6) and dates from the fourth century. 2 That the scene is rather a symbolic than a literal representation of the Birth is shown by Fig. 6. Rome: S. Sebastiano, Fig. 7. Milan: sarcophagus. The fresco. The Adoration of the Adoration of the Ox and Ass. Ox and Ass. the introduction of the nimbed bust of Christ above the crib, where the animals worship the young Child, and by the absence of the Virgin and Joseph, or other distinguishing features which would mark it as an actual event. The other example of this group is in the tympanum of the end of a sarcophagus cover at Milan. 3 In this scene (Fig. 7) the Christ Child is depicted lying in His crib with the ox and the ass kneeling in adoration at His head and feet. Group II. Introduction of the Shepherds. Again in this symbolic manner, before there seems to have been any attempt to localize the scene in a manger, we have three sarcophagi of Rome 4 on which (Fig. 8) the Adoration of the Shepherds is joined to that of the animals. The fact that the Christ Child in two of these examples is on an altar and in the other lies on the ground, while no other figures save the two shepherds and the animals are present, shows that it To him I am indebted for the sub-groups of the Adoration type. Other works to be consulted on the Nativity are: Heisenberg, Grabeskirche und Apostelkirche, II, p. 234 sq.; Noack, Die Darstellung der Geburt Cliristi in der bildenden Kunst, Darmstadt, 1894; Dalton, Byzantine Art and Archaeology, p. 653. 2 See Table I, no. 1. 3 Table I, no. 2. * Table I, Group 2. EARLY CHRISTIAN ICONOGRAPHY was the idea of worship rather than of birtli which the artists were seeking to convey. Group III. Introduction of the Shed. In this group the same features are preserved as in the preceding groups without the addition of any new figures to the composition; the only difference is that the Child in His crib is placed under the traditional Western shed (Fig. 9). On the three 3 sarcophagi of this group, as on all the remaining sarcophagus representations, the Babe is represented in a manger which assumes the form of what Schmid 6 calls the “Roman crib,” a child's basket higher at one end than at the other. Group IV. Introduction of the Virgin. The seated Virgin is added to the scene. Save for this addition of the figure of Mary, who is represented sitting apart from the manger where the Child lies (Fig. 10), the sarcophagi of the fourth group show the same scene as those of the third. 7 Fig. 8. Rome: sarcophagus. The Adoration of the Animals and Shep- herds. Fig. 9 Rome: sarcophagus. The Adoration in a Manger-shed. Fig. 10. Mantua: sarcophagus. The Adoration, with the Virgin. “Table I, Group 3. 6 Schmid, op. cit., p. 83. Table I, Group 4. 16 EARLY CHRISTIAN ICONOGRAPHY Group V. Introduction of the Magi. On the sarcophagi 8 of this group the Adoration of the Magi is added to the Adorations of the Ox and Ass, and of the Shepherds, already present in the type. On all these examples, with only one exception, 9 the Adoration of the Ox and the Ass as well as that of the Shepherds is pre- served in the composition (Fig. n), although the shepherds are reduced to one. These sarcophagi are anterior in date to the middle of the fourth Fig. ii. Rome: sarcophagus. The Adoration, with the Magi. century, 10 for in their peculiar combination of the Adoration of the Magi with the Birth they reflect the usage of the church in Rome, which up to the middle of the fourth century celebrated the Nativity along with Epiphany on January 6 . About 350 a.d. two distinct and separate feasts were established. 11 Group VI. The Provengal Type. This division of the Western Adoration type of the Nativity is repre- Fig. 12. Milan: Cathedral, ivory book covers. The Nativity, Adoration Type. sented by that group of ivories which I have assigned to an early Christian school of southern France, consisting of the Milan book covers 12 (Fig. 8 Table I, Group 5. 9 Table I, no. 19. 10 See Adoration of the Magi, Group II, page 42. 11 Schmid, op. cit., p. 53. 12 Table I, no. 21. EARLY CHRISTIAN ICONOGRAPHY 1 7 1 2), the Werden casket 13 in the South Kensington Museum (Fig. 13), the ivory fragment in the Nevers Museum (Fig. 164), 14 and possibly a pyxis at Rouen. 15 The same type occurs on a book cover in the Bodleian Library at Oxford 16 (Fig. 169), which is frequently grouped with the Milan covers as a work of early Christian art. but which is really a Caro- Fig. 13. S Kensington: Victoria and Al- bert Museum, ivory casket. The Nativity, Adoration Type. lingian ivory dating shortly before or after 800. 17 In nearly all the scenes that the Bodleian cover has in common with the Milan covers, the Carolingian work continues their curious iconography. The scene on the Milan ivory shows Mary and Joseph sitting in a shed on either side of a manger, where the ox and the ass adore the Christ Child. While the distinguishing feature of the Milan, Werden, and Bodleian ivories which separates the scene from the other Western representations is the presence of Joseph with a curiously shaped carpen- ter's saw in his hand, it is the form of the crib which binds all the ivories together into a group. On the sarcophagi a marked type of crib was represented which conformed to the regular Roman cradle; here, on the ivories, is a simple and yet distinct manger built up of stones or brick, much like that of the Alexandrian-Coptic group (IX), but distinguished by being covered with hay or some soft material, the folds of which radiate from the Child as He lies upon it. The scene on the Milan cover is so closely paralleled on the Werden casket and the Bodleian cover that the three are undoubtedly bound together thereby, but on the Nevers fragment (Fig. 164) there is left of the type only the same form of manger with its radiat- 13 Table I, no. 22. 14 Table I, no. 20. 11 Table I, no. 24; this scene is less Provenqal in type than it is Oriental-Hellenistic. In all probability the pyxis was executed in Egypt. 16 Table T, no. 23. 17 See Bodleian Ivory, p. 248. EARLY CHRISTIAN ICONOGRAPHY ing cover, and the ox and the ass. The fact, however, that the scene on the Nevers ivory (which is a fragment of a five part diptych carved sometime before the Milan and Werden ivories) has been broken in two, accounts for the absence of Joseph and the Virgin, who must have been represented in the missing half of the panel. 18 Not only are there no other traditional figures which might have com- pleted the scene, but such an assumption as to the missing half of the panel is sustained by the representation on the Carolingian book cover in the Bodleian Library. This later ivory copied its iconography from an ivory fragment in the Berlin Museum, which in its turn was a part of the same or a similar early Christian diptych as the Nevers frag- ment (see p. 237) and it depicts the same type of the Adoration of the Ox and the Ass with Joseph and the Virgin crowded into the corners of the composition. The Adoration Type and Pseudo-Matthew The most likely textual sources for the Adoration type are the accounts of Matthew and Luke, and for the West in the fifth century the apocryphal Latin Gospel of Pseudo-Matthew. Regarding the Birth, the account of Matthew (i, 25) is too colorless to suggest a type. Further- more, it fails to mention either the Adoration of the Ox and the Ass or the Adoration of the Shepherds, but in the second chapter tells of the coming of the Wise Men out of the East and says that “when they were come into the house they saw the young Child with Mary his mother, and fell down, and worshipped him.” St. Luke (ii, 1-7) is more explicit and, after telling of the journey to Bethlehem and the birth of the Babe, says that Mary “laid him in a manger.” Although he omits the Adoration of the Ox and the Ass and fails to mention the coming of the Wise Men, he dwells (ii. 8) on the summoning of the shep- herds and writes that they “found Mary, and Joseph, and the babe lying in a manger.” The Gospel of Pseudo-Matthew is the most interesting of these texts, for it is the only account, either canonical or apocryphal, which includes the Adoration of the beasts. It is, as is pointed out in the Intro- duction to Part II, a Latin adaptation from Jewish-Christian and Gnostic sources, written at some time in the fifth century. 19 It agrees at 18 See Nevers Ivory, p. 241. 39 Michel, op. cit., p. XXI. EARLY CHRISTIAN ICONOGRAPHY i9 the outset with the purely Oriental apocryphal books and relates that the time for the delivery came while Mary and Joseph were on the way to Bethlehem and Mary went “into a cavern." 20 Although the writer makes no mention of the shepherds going to worship at the crib of the Saviour, he gives a careful account of the Adoration of the Magi and inserts a description of the Adoration of the Ox and the Ass. We read that “O11 the third day after the birth of our Lord Jesus Christ, the most Blessed Mary went forth out of the cave, and entering a stable, placed the Child in the manger, and the ox and the ass adored Him.” “And Joseph and Mary lived in this place with the Child for three days." 21 The Western scenes, however, of the Adoration type already had in the account of Luke a good source for the shed and a source in Matthew and Luke for a combi- nation scene with the Adoration either of the Shepherds or of the Magi. That the scenes of this Western type are rather Adorations than Nativities is consistent with the slight consideration which the early theologians gave to the actual birth and also with the highly symbolic character of Christian art in the West during the first four centuries. The relatively slight importance of the Nativity prior to 350 A. D. and the fact that it was seldom, if ever, separated in the Christian mind from some form of an Adoration is shown by the fact that it was celebrated in Rome on January 6 at the same time with the Adoration of the Magi and the Adoration of the Shepherds. Augustine and other of the Fathers, some of whom already recognized the twenty-fifth of December as the Natal Day, looked upon it not as a feast day and a sacrament, but simply as a day to be remembered. 22 Therefore, considering the minor importance of the Nativity, the absence of the Nativity in the frescoes of the cata- combs, save the late and manifestly symbolic scene in S. Sebastiano, and the fact that the scene depicted on all the sarcophagi has joined with it some form of an Adoration, it is to be supposed that these Western scenes typify either the Adoration of the Gentiles, who are symbolized by the Magi at the feet of Christ, or of the Jews, whom the shepherds represent, or the Adoration of the animal kingdom as denoted by the ox and ass. W hether or not it is admitted that all the scenes were conceived with only the idea of extolling to the faithful the divine kingship of the Saviour 20 Michel, op cit., p. 99. 21 Michel, op. cit., p. 105. 22 Schmid, op. cit., p. 53. 20 EARLY CHRISTIAN ICONOGRAPHY by revealing to them the universal worship and homage which was im- mediately accorded Him even at the moment of His birth, it is clear that the ox and the ass existed in art not as accessories of the manger but as a form of adoration. Although the two beasts were not mentioned in any of the Gospels before the fifth century, their symbolic presence at the Nativity had already been referred to by theologians. In the West, Pru- dentius , 23 St. Ambrose , 24 and Peter of Ravenna 25 speak of them as adoring their Master, and in the East the same significance is given them by Ephraem, Gregory of Nazianzus, and Gregory of Nyssa . 26 Hence in the fifth century when the Gospel of Pseudo-Matthew was being translated for Latin readers from its Eastern sources, the presence of the ox and the ass as symbols of adoration in the Nativity was a commonplace of Chris- tian imagery . 27 Then, as so frequently happens, the established art influ- enced the literature of the period. Western Nativities regularly included the stable-shed, while Eastern tradition (see p. 22) insisted that the Child was born in a cave. Plence the Latin compiler of Pseudo-Matthew took pains to reconcile his Eastern sources with Western artistic usage by hav- ing Mary go forth from the cave and place the Child in the manger of a stable, where the “ox and the ass adored Him.” This last interpolation, the only account we have of the Adoration of the Ox and Ass, is a similar enrichment of the narrative with a view to reconciling it with current images of the Nativity. This description in Pseudo-Matthew of the Adoration of the Ox and the Ass, rather than of a Nativity, is in turn the most likely source for the 23 Leclercq in Cabrol’s Dictionnaire d’arclicologie chretienne et de liturgie, s. v. “Ane,” col. 2049, cites the following hymn of Prudentius : “Aeterne rex, cunabula, Populisque per saeclum sacra, Mutis et ipsis credita, Adorat haec brutum pecus.” ( Cathemerinon XI, 78, P. L. LIX, col. 896. ) 24 Leclercq, op. cit., col. 2053, cites the following passage from St. Ambrose wherein is brought out the symbolic meaning of the ox and the ass : “Noli hoc aestimare quod cernis, sed quod redimeris agnosce. Quia in pannis est, vides, quia in caelis est non vides. Infantis audis vagitus, non audis bovis dominum agnoscentis mugitus. Agnovit eniin bos possessorem suum, et asinus praesepe Domini sui.” (In Lucam, II, 7, P. L. XV, col. 1648 sq.) 25 Kehrer, Die Heiligen drci Konige in Litcratur und Kunst, 1908, I, p. 21. 26 Kehrer, op. cit., I, p. 21. 27 Leclercq, op. cit., col. 2053, says : “Ainsi, dans le courant du iv e siecle, on pouvait reconnaitre encore, au moins quarid on avait entendu saint Ambroise, qiie l’ane et le bceuf de la creche n’etaient autre chose que des symboles.” EARLY CHRISTIAN ICONOGRAPHY 21 Provengal type of the Nevers, Milan, Werden, and Bodleian ivories. This type is Western in that the event is represented under cover rather than in the open or in a cave as is the case with the Eastern scenes. Yet the type is different from any of those on the Western sarcophagi. It is a more definite attempt than any on the sarcophagi (on which we find only the most rudimentary kind of a shed covering the Child alone) to render a scene in a stable, for across the whole back of the composition, in the best examples (Milan, Nevers), extends a brick wall. Also, unlike the sarco- phagi scenes in which either the shepherds or the Magi are usually figured and Joseph is absent, it omits the shepherds but represents both Joseph and Mary seated beside the crib where the beasts worship the Babe. As a translation into pictorial form of the account in the Latin Pseudo-Matthew these variations from the customary Western types are explained by the description which says that Mary, “entering a stable, placed the Child in the manger, and the ox and the ass adored Him. . . . And Joseph and Mary lived in this place with the Child for three days.’’ We may have then, in this scene, that dependence on the Pseudo-Matthew text which is apparent in other scenes on the ivories of this group. Two peculiarities of the use of the scene in Provence lie in the ultra- symbolical way in which it was conceived, and in a persistent tendency to connect it with the Adoration of the Magi. On a sarcophagus cover of Provencal style at Milan, 28 which has already been cited in the first group, there is an undoubtedly symbolical representation of the Adoration of the Ox and the Ass (Fig. 7). On a sarcophagus of S. Trophime at Arles 29 (Fig. 14) there occurs an Epiphany scene in which Jesus, a child of 29 See page 14 . Le Blant, Les sarcophages chretiens dc la zdlle d’ Arles, pi. XXVI; Garr., op. cit., V, pi. 317/4- No. 19 (Table I), which omits the Shepherds and combines the adorations of the Magi and the Ox and Ass, is a Provencal sarcophagus (St. Maximin). 22 EARLY CHRISTIAN ICONOGRAPHY two years, sitting in His mother's lap, receives the gifts of the three Magi while at His feet lie the two animals looking up at Him with devotion. This last scene is a frank combination of the two Adorations in a sym- bolical sense, without any connection whatever with the Nativity. The same combination of the Adoration of the animals with that of the Magi was in the minds of the ivory carvers of this school of southern France, as is shown by the way in which they associate the two scenes. For example, the Adoration of the Ox and the Ass on one of the Milan book covers occupies the top panel, which is also the place assigned to the Adoration of the Magi on the other cover, in defiance of chronological sequence. The Werden and Nevers ivories combine the two scenes in one panel, and on the Bodleian book cover one occurs below the other. East. In the East the Oriental-Hellenistic, the Alexandrian-Coptic, the Pal- estinian-Coptic, and the Byzantine types stand out in clear distinction. They are all distinguished from the Western Adoration type by the ab- sence of the shed. All the apocryphal gospels of the Orient, as the Prot- evangelium of James, 30 the History of Joseph, 31 the History of the Blessed Virgin Mary, 32 and the book of Pseudo-Matthew which was derived from Eastern sources, 33 agree in stating that Mary gave birth in a cave on the way to Bethlehem. These apocryphal gospels were the sources fol- lowed by the artists of the East. Their Nativities therefore are depicted either in a cave or, as is more often the case, in the open, because a scene in a cave was too difficult a problem for their limited artistic expression. There are, however, two very small groups of Eastern monuments which, save as they represent the scene in the open, do not conform to any of the Oriental types. Group VII. The Oriental-Hellenistic Group. This group is a continuation of the Adoration type on Eastern monuments. It differs, however, from the Hellenistic type in that the Western shed is omitted and the event is figured in the open. It appears on two encolpia, one in the Museum at Constantinople 34 (Fig. 15) and the other in the Museum of Reggio, 35 both ascribed to Palestine by Diehl, 30 Michel, op. cit., p. 39. 31 Michel, op. cit., p. 201. 32 E. A. Wallis Budge, The History of the Blessed Virgin Mary, London, 1899, p. 32. 33 See p. 189. 34 Table I, no. 26. 35 Table I, no. 25. EARLY CHRISTIAN ICONOGRAPHY 2 3 Dalton and Strzygowski. While on both of these encolpia the scene takes place in the open after the Eastern manner, the Virgin and Joseph are omitted, and on the Reggio encolpium there is figured only the Adoration Fig. 15. Constantinople: Imp. Ottoman Museum, encolpium. Nativity and Flight into Egypt. of the animals with the Adoration of the Shepherds. Another gold en- colpium, from Cyprus, was recently published by Strzygowski in Oricns Christianas (V, 1915, p. 96 scj.) ; it adds to the above scene the seated fig- ure of Joseph and a saddled and bridled ass, reminiscent of the Journey to Bethlehem, — the Virgin is still absent (Table I, Errata, no. 27 a). These Eastern examples show that the Adoration type is “Western” only in that nearly all its examples occur on Western monuments. As in the case of so many other scenes, the Western type was simply a Hellenistic type common throughout the Mediterranean basin and surviving later chiefly in the West. It is then not in the least surprising to find it occasionally, with its special meaning of worship, on later monuments of the East as well. That it was prevalent earlier in the East is shown by its appearance in its simplest form on a marble relief found in Naxos (Table I, Err. no. 24 a) which cannot be later than the early fourth century. Group VIII. This group appears to depict a transitional type and is represented, so far as existing monuments go, only by the Syrian Gospels of Rabula. 30 Unlike the Eastern representations, Mary is seated, instead of lying on a mattress, while Joseph is placed behind the crib ; in the background is the faqade of a building (Fig. 16). In the scene are no adoring animals, shepherds, or Magi. That it may represent the Syrian type of the Nativity can not be assumed from such scanty evidence. Group IX. The Alexandrian-C optic Type. As a type it is characterized by the introduction of the midwife Salome Table I, no. 28. 24 EARLY CHRISTIAN ICONOGRAPHY into what would otherwise be a general Oriental type. In this group the Christ Child, wrapped like a mummy, lies on a stone or brick crib where the ox and the ass adore Him ; on one side of the foreground sits Joseph, on the other the Virgin reclines on a mattress, and either in front of the crib Fig. 16. Florence: Laurentiana, Gospels of Rabula. The Nativity. or before the Virgin kneels Salome extending her withered arm in suppli- cation to Mary. Three sets of proofs connect the Salome type of the Nativity with the Coptic art of Egypt. They are : the Coptic character of the style and ornament of most of the monuments on which the type occurs, the Coptic origin of Salome in apocryphal literature, and the moral and religious concepts of the Copts which required the introduction of the doubting Salome into the scene to prove the absolute spiritual nature of Christ. The reason for calling this type the Alexandrian-Coptic is the fact that it first occurs on the chair of Maximianus 37 (Fig. 17) and then con- tinues to appear on a number of unquestionable Coptic monuments which in other scenes as well are allied in iconography with the Maximianus chair. The chair is certainly Egyptian by its iconography; practically all its scenes are closely related to Coptic types. Its style, however, while afford- ing some parallels with Coptic work, is so generally Hellenistic as to justify the attribution bv Diehl and Leclercq to Alexandria ; the extensive evidence 37 Table I, no. 29. EARLY CHRISTIAN ICONOGRAPHY 25 in favor of their attribution is assembled in my article on the subject in the American Journal of Archaeology (1917, pp. 22-37). The close rela- Fig. 17. Rome: Stroganoff Collection, PANEL FROM CHAIR OF MAXIMIANUS. The Nativity. tions of its scenes to Coptic types simply reflect the natural derivation of the art of Upper Egypt from ateliers of Alexandria. The Coptic monuments on which this type with the doubting Salome occurs are : a fragment of the Murano book covers in Manchester 38 (Fig. 45), an ivory fragment in the British Museum 39 (Fig. 18), a pyxis in Vienna 40 (Fig. 19), one in the Berlin Museum 41 (Fig. 20), and one in Werden. 42 In a fresco of Bawit, 43 dating from the sixth or seventh century, there is also a Nativity with Salome, as the inscription shows, which omits the Christ Child and Joseph, a point whose significance will be discussed later. In the eighth century on a Coptic ivory of Bologna 44 the scene appears as a transition to the ’'Table I, no. 31. 39 Table I. no. 32. 40 Table I, no. 35. 41 Table I, no. 33. 42 Table I, no. 34. 43 Table I, no. 30. 44 Table I, no. 36. 26 EARLY CHRISTIAN ICONOGRAPHY Byzantine type representing Salome about to wash the child. In a curious eclectic rendering Salome occurs in the scene of the Nativity on one of the ciborium columns of San Marco . 45 Instead of lying on a mattress as she does in all the Coptic examples of the Nativity, the Virgin is here Fig, 18. British Museum: ivory. Fig. 19. Vienna: K. K. Munzkabi- The Nativity. nett, ivory pyxis. The Nativity. seated as on the Western sarcophagi. The first reason, then, for connect- ing this type with Egypt is its appearance in the Bawit fresco and the fact that all the above mentioned ivories with the exception of the Max- imianus chair are readily admitted as Coptic monuments from their style and ornament . 46 Second, with reference to the Coptic origin of the Salome motif, Fig. 20. Berlin: Museum, ivory pyxis from Minden. The Nativity. Reveillout 47 has found some Coptic apocryphal fragments belonging to the Gospel of the Twelve Apostles which prove that this detail originated in Egypt. In the Protevangelium of James and the Gospel of Pseudo- Matthew, which goes back to Oriental sources, Salome is only briefly mentioned in connection with the Nativity. On the other hand, in the Gospel of the Twelve Apostles her importance in Coptic eyes is brought out in the detailed account which is given of her whole life from the 45 Table I, no. 37. 46 Dalton, Byzantine Art and Archaeology, pp. 209-211; Strzygowski, Hellenistische und Koptische Kunst, pp. 85-88. 47 Reveillout, Journal Asiatique, X e ser. V. 1, 1905, “La sage-femme Salome,” p. 409 sq. EARLY CHRISTIAN ICONOGRAPHY 27 time when she first sold herself for money, through her participation in the Nativity, to her conversion by Simeon. I11 the Protevangelium, again, the midwife whom Joseph finds bringsjfth her the doubting Salome; in the Pseudo-Matthew this midwife is calle^Yelemi. Inasmuch as Salome and the midwife in the Coptic Gospels are one and the same person, Reveillout 48 holds that Salome in the Protevangelium has no reason to be there save for the purpose of reconciling two diverse texts. One of these texts he supposes to be Coptic, in which Salome was the midwife, and the other a Syrian text in which the midwife was called Zelemi. Since the earliest manuscript of the Protevangelium dates only from the ninth century 49 and manifests a blending of the Gnostic ( ?) Apocryphum Josephi and the Apocrvphum Zachariae, 50 it is quite possible that its text was interpolated to conciliate the divers Coptic and Syrian versions of the Nativity. Furthermore, the Nativity in the fresco which Cledat 51 unearthed in the ancient Coptic monastery of Apa Apollo at Bawit (Fig. 21) bears out the Coptic origin of Salome. The unusual character of this scene Fig. 21. Bawit: fresco. The Nativity. lies in the absence of Joseph and the Christ Child from the composition and the striking importance given to Salome, who stands before the Virgin as she reclines upon a mattress. Reveillout, 52 commenting on the Bawit frescoes, shows that the whole series dealing with the early life of the Vir- gin represents an unique tradition drawn from the Coptic Gospel of the Twelve Apostles. When one considers the minute and detailed emphasis laid upon the story of Salome in the Coptic gospel in connection with the extraordinary significance given her in the Bawit Nativity, it seems clear 4 *iReveillout, op. cit., p. 441. 50 Vichel, op. cit., p. VIII. 51 Cledat, Comptes Rendus, 1904, p. 517. 52 Reveillout, op. cit., p. 428. 49 Michel, op. cit., p. III. 28 EARLY CHRISTIAN ICONOGRAPHY that the doubting Salome is an addition to the scene which we owe to Egypt. Lastly, the motif of Salome is perfectly consistent with the moral and religious ideals of the Copts. The Coptic mind required the presence of the doubting midwife in the scene by reason of its fundamentally carnal character. 53 The introduction of Salome in the Nativity simply expressed the doubt which until it was dispelled prevented the Copts from accept- ing the divinity of the Saviour. In other words, to the degree that they were themselves material, they naturally demanded that their divinity be wholly spiritual, partaking of no polluting material attributes. For this reason they were Monophysites, combating the orthodox view of the Trinity and upholding the purely spiritual nature of Christ. The doubting Salome afforded a very obvious, however vulgar, means of demonstrating the immaculate birth of the Saviour, especially important to them in their struggle against the orthodox view of Christ’s double nature. It is interesting to note that this incredulous midwife occurs spas- modically in Carolingian and Ottonian art. In the Sacramentarium of Drogo, 54 which was executed in the ninth century and bears some very close analogies to Oriental art and especially that of Alexandria, Salome is depicted leaning over the Virgin as she reclines on her mattress. In the same manner the midwife is shown on a tenth century ivory of German workmanship in the Spitzer collection 55 and on two other German ivories 56 of the tenth or eleventh centuries. This Coptic type also penetrated into Rome where we find the doubting Salome in the eighth century in a mosaic of the chapel of John VII 57 and in a fresco 58 in the cemetery of S. Valentino. With these few exceptions, the Coptic midwife rarely ap- pears in the later art either of East or West. 59 53 Gayet, L’art Copte, p. 33 sq. One has only to read the history of the Coptic Fathers from the account of the self-infliction of Origen, through the sensual temptations of St. Anthony, to the confessions of Schnoudi to realize, as Gayet says, that restraint, purity, and love did not exist in them. 54 Bibl. Nat., Paris, lat. 9428; L. Weber, Einbanddecken usw. aits Metzer liturgischen Handschriften, I, Strassburg, 1913, pi. IX. 55 Molinier, Hist. gen. des arts appliques, I, Ivoires, p. 147. 56 Voge, Die Elfenbeinbildwerke der christlichen Epochen (Cat. Kaiser-Friedrich Mu- seum), 1900, PI. XIV, no. 40, and no. 35. 57 Table I, 110. 58. 68 Table I, no. 59. 59 Dalton, op. cit., p. 654, says, “The episode of the incredulous midwife Salome . . . has been commonly thought not to occur much before 800 A. D.” He cites for this statement Schmid ( Die Darstellung der Geburt Christi, p. 125) who really says that EARLY CHRISTIAN ICONOGRAPHY 29 Group X. Palestinian-C optic Type. This is the type which seems to have originated in Palestine and passed into Egypt. 60 It is characterized by the representation of the birth in the open where Mary, with Joseph seated beside her, lies in Eastern fashion on a mattress. The best examples of this type appear on a Monza phial 61 (Fig. 22) and the covers of the Etschmiadzin Gospels. 62 The first shows the type as it appeared in Palestine ; the second is an Egyptian work, being a later product of the same school that produced the Maximianus chair, and showing in its iconography that imitation of Syro-Palestinian types which is so frequent in the Christian art of Egypt (p. 179). The type was prob- ably derived from the mosaic of the Nativity which adorned the famous Church of the Nativity at Bethlehem. 63 There existed intimate relations between the minor arts and the great works of monumental style. The large historical compositions of the scenes of the New Testament in the great church of the Holy Land were very soon universally known through Christendom. 64 These illustrious prototypes were willingly imitated and no doubt constituted a tradition for a certain number of scenes. Small mobile objects, as manuscripts, stuffs, ivories, censers, 66 and Salome does not appear in Eastern or Western art after the year 800. This is untrue of the West, but so far as I have been able to trace it, the actual doubting Salome does not occur in the East after that date. 60 Table I, nos. 38-53. 61 Table, no. 39. 82 Table, no. 38. 83 Heisenberg, Grabeskirche und Apostelkirche ; Dalton, op. cit., p. 624; Diehl, Manuel d’art byzantin, p. 300; Strzygowski, Hell, und Kopt. Kunst, p. 92. e * Diehl, op. cit., p. 10. 63 Table, nos. 42-49. Fig. 22. Monza : metal ampulla, detail. The Nativity. 30 EARLY CHRISTIAN ICONOGRAPHY especially the Palestinian ampullae of which those at Monza 66 are examples, may well have spread throughout the Mediterranean the iconographical types fixed by this monumental art. At any rate the Palestinian type of the Nativity passed into Egypt where it occurs on a Coptic painted panel (Fig 23) now in the Golenisheff Collection 67 at Petrograd and on a Coptic textile 68 in the Sancta Sanctorum at Rome. While a few later examples 69 exist of this type, for practical purposes of distinction we may say that the Palestinian-Coptic type fell into disuse at the end of the eighth century. The Byzantine type is characterized at the outset by the presence of two midwives as a customary part of the composition. 70 By the beginning of the ninth century Byzantine art had commenced to formulate its iconography along lines of supposedly historical fact. Seeking as it did to give some precise and comprehensive details to the rather pale events of the life of Christ as they were related in the canonical Gos- pels, the artists more and more depended upon the apocryphal books 66 Diehl, op. cit., p. 301. 67 Table, no. 40. 68 Table, no. 41. 69 Table, nos. 51, 52, 53. 70 This apocryphal corruption in the East of the canonical account, which in Luke specifically stated that Mary herself enveloped the Child, aroused a great controversy dur- ing the earlier centuries when Eastern ideas were being taken up in the West. Pope Gelasius rejected all apocryphal writing which gave any place to the testimony of the two midwives and proved so inconvenient for the theologians to justify. The popular demand was too strong for the Church and by the eighth century there was introduced into the West as well as in the East the two midwives washing the young Saviour. The Church, when it realized that Zelemi and Salome were in the scene to stay, explained Fig. 23. Petrograd: Golenisheff Collec- tion, panel. The Nativity. EARLY CHRISTIAN ICONOGRAPHY 3i and especially upon the Protevangelium of James. While in the literature of the East, outside the apocryphal gospels, the two midwives are first mentioned only in the tenth century by Simeon Metaphrastes, 71 from the eighth century on there are very few cases either in Byzantine or Italo- Byzantine art in which the two women do not appear in the Nativity as assistants washing the Child. 72 It is to be noted, however, that an abbreviated Byzantine type is found in the eighth or ninth century. Group XI. A Byzantine Sub-type. While the Byzantine type is generally to be characterized by the pres- ence of the two midwives, this small group represents a rare sub-type 73 where the midwives are omitted and the new feature of angels hovering over the crib is introduced (Fig. 24). Group XII. Byzantine Type. The actual Byzantine type is represented by this group, which, besides the distinctive feature of the midwives, also shows the adoring angels Fig. 24. Berlin: Kupferstichkabinett, Greek lectionary. The Nativity. who manifestly represent the Adoration of the Heavenly Hosts for the Son of God. \\ hi le this type does not begin to appear universally on Byzantine monuments until the ninth century, its inception may go back to earlier proto-Byzantine or Syrian models. The description' 4 of the lost mosaics their presence on the ground that they existed as witnesses of the event and the function which they performed originally referred to the sanctification by water which God decreed that His Son should undergo that man might see the sacramental necessity of baptism. 71 Heisenberg, op. cit., II, p. 226. 72 Table I, Group 12. 73 Table, nos. 54, 55, 56. 74 Heisenberg, op. cit., II, p. 223. The Church of the Holy Apostles was built and decorated by Justinian (op. cit., page 2), and by the end of the ninth century had lost all its old beauty and was in ruins. Under the Emperor Basil I (867-886) it was re- EARLY CHRISTIAN ICONOGRAPHY 3 2 of the Church of the Holy Apostles at Constantinople says that the event was pictured in a cave, that the Virgin was depicted reclining on a mat- tress, that at least one midwife was represented “washing the Child,” and that the Adorations of the beasts, the shepherds, and the angels were in- cluded in the scene. To this description Heisenberg, who attempts to restore the mosaics and date them in the sixth century, would add the figure of Joseph and another midwife. But these mosaics are now dated by the majority of scholars in the ninth century. With few exceptions Fig. 25. Paris: Bibl. Nat., Ms. gr. 7 4. The Nativity. from the eighth century on in the East the Adoration of the angels is combined with the Adoration of the animals, 75 with the Adoration of (or in some cases more properly the Annunciation to) the Shepherds, 76 and often with the Adoration of the Magi. 77 Mary on all the examples of this group reclines on a mattress, and the Nativity itself, though for a time fig- ured as having taken place in the open, 78 soon came to be depicted in a cave. A fair sample of the Byzantine type, exceptional nevertheless in omitting the angels, is afforded by Figure 25. decorated. Of the edifice thus restored and of the mosaics which adorned its walls there exist two descriptions : one, the poem of Constantine the Rhodian of the tenth century; the other, more complete, is the work of Nicholas Mesarites, which was com- posed between 1199 and 1203 {op. cit., p. 8) before the church was finally destroyed to make place for the mosque of Mahomet II. Although the mosaics in this church are usually attributed to the ninth century (Dalton, Byzantine Art and Archaeology, p. 392; Diehl, Manuel, p. 448-451 ; Millet, L’art byzantin, in Michel, Histoire dc I’art chretien, I, p. 190), Heisenberg assigns them to the sixth century largely on the ground of their similarity to the miniatures of the proto-Byzantine Gospels of Rossano (Heisenberg, op. cit., pp. 140-170). Whether of the sixth or ninth century, although the latter date is by far the more convincing, the scenes in the mosaics of this famous church would have been influential in the final establishment of Byzantine iconography which did not be- come traditional until the ninth century. 75 Table I, nos. 62-66, 71-74. 76 Table I, nos. 61-64, 66, 71-74. 77 Table I, nos. 66, 69, 73, 74. 78 Table I. Baumstark has recently ( Oriens Christ. 1915) pointed out a possible Syrian prototype for the Byzantine Nativity. EARLY CHRISTIAN ICONOGRAPHY 33 V THE APPEARANCE OF THE STAR TO THE MAGI Monuments : Sarcophagus of Arles. 1 Sarcophagus of Arles. 2 Ivory book covers, Milan. 3 Ivory casket in the S. Kensington Museum. 4 Sarcophagus of San Celso in Milan. 5 Ambo of Salonica. 6 Coptic ivory, Bologna, eighth century. 7 Codex Egberti at Trier. 8 Gospels of Henry III, Bremen. 9 Gospels of Brussels, Cod. lat. 9428. 10 Gospels of St. Bernward at Hildesheim, 11 No. >- Early Christian - Ottonian 18. Fresco of S. Urbano alia Caffarella, an. ion. 12 Bronze doors of Pisa and Monreale. 13 Hortus Deliciarum of Herrad von Landsberg. 14 Italo-Byzantine The Appearance of the Star to the Three Magi was primarily Syrian in conception and origin. The event, as a separate scene in early Christian art, occurs only on the monuments of communities which were closely connected with Syria. The largest number of examples come from Provence, which in many ways, as will be shown in a subsequent chapter (see p. 192 sq.), was a Syrian colony closely connected with the Syrian church. Although the star appears to the Magi on the sarcophagi of Rome it is always in the representations of the actual Adoration and never as a separate scene. As a distinct scene, usually combined with another and separate scene of the Nativity or Adoration, it is rare, and as far as my knowledge goes is limited, in early Mediaeval art, to the examples cited above While its importance in the Adoration was recognized in the West, it had no liturgical significance and does not appear to have received the symbolical interpretation that it had in the East. Prudentius simply de- 1 Le Blant, Lcs sarcophages chrcticns de la ville d’ Arles, p. 35, pi. XXI. 2 Le Blant, op. cit., p. 32. pi. XVIII. 3 Garr., op. cit., VI, pi. 454. 4 Garr., op. cit., VI, pi. 447/2. 5 Garr., op. cit., V, pi. 315/5. 9 Kehrer, Die Heiligen drci Konige in Literatur und Kunst, II, figs. 17, 18. 7 Stuhlfauth, Elfenbeinplastik, pi. III/2. 8 Kehrer, op. cit., fig. 19. 9 Kehrer, op. cit., fig. 20. 10 Kehrer, op. cit., fig. 22. 11 Kehrer, op. cit., fig. 23. 12 Kehrer, op. cit., fig. 24. 13 Kehrer, op. cit., fig. 25. 14 Kehrer, op. cit., fig. 26. EARLY CHRISTIAN ICONOGRAPHY 34 scribes the star and says that it effaced all others and shone so bright that Lucifer himself dare not struggle against it . 15 It was in Syria that the im- portance of this heavenly guide was symbolically developed and introduced into the liturgy of the church. The star was the text of exegesis and ser- mons and in the Syrian church on the night of January 6 , besides the P>irth of the Saviour and His Adoration by the Shepherds, the Appearance of the Star was also celebrated . 10 Basilius, Bishop of Caesarea in Cappadocia, has left a sermon on its appearance 17 and St. Ephraem gave the first full interpretation of this “Herald on high, a Herald of Divinity .” 18 More- over, he mentions an angel in connection with the star, and while there remain no early monuments from Syria where the Appearance of the Star is figured as a separate scene, on one of the Palestinian phials 19 at Monza an angel points out to the Magi a large star which hangs over the head of the Virgin. The rendering in Provence is distinctive. On two sarcophagi of Arles (Fig. 26) the Magi, dressed in Phrygian caps, short chitons, chlamydes, Fig. 26. Arles: sarcophagus. The Magi GAZING AT THE STAR. and anaxyrides, are represented gazing with gestures of surprise at a brilliant star above their heads. The characteristic feature is the odd notched chiton which they wear. The scene on the Milan sarcophagus in the church of San Celso, while similar in motif, is entirely different in composition and in this particular detail of the Magi's costume. The three Magi, all in a line, are looking, not over their heads at the star as 15 Duchesne and Bayet, Mem. sur line Mission an Mont Atlios, 1876, p. 264: Prudentius, Cathemerinon, hymnus XII, Dc Epiphania, V, 25-73. 10 Kehrer, op. cit., II, p. 27. 17 Kehrer, op. cit., I, p. 14. 18 Kehrer, op. cit., I, p. 14. 19 Garr., op. cit.. VI, pi. 433/9. EARLY CHRISTIAN ICONOGRAPHY 35 on the Provencal sarcophagi, but toward the right; their chitons instead of being notched around the bottom, so that they appear to hang in three triangular pieces as on the Gallic monuments, are cut with a straight edge and girded about the waist. This feature of the notched chiton was common, if not peculiar, to the costume of Orientals on the monu- ments of Provence. On a sarcophagus of Saint-Gilles 20 in the scene of the three Hebrews before the bust of Nebuchadnezzar the young men, instead of looking at the bust, are concerned with a star above their heads; they use the same gestures and attitudes as the Magi on the Arles sar- cophagi and they also wear the notched chitons. The scene on the Milan book covers (Fig. 27) and on the Werden casket (Fig. 28), which is always to be associated with the covers, is Fig. 27. Milan : Cathedral, ivory book covers. The Magi gazing at the Star. nearly identical, even to the gestures and notched chitons of the Magi, with the representations of the sarcophagi of Provence. The similarity in fact is so marked as undoubtedly to connect the ivories with southern Gaul. While the intimate religious bond between Provence and Syria relates the scene on the sarcophagi and ivories to the Syrian feast on January 6, the text of Pseudo-Matthew, 21 which we have seen in- fluencing the rendering of other scenes on the Milan and Werden ivories, offers a source for the actual representation. In the apocryphal account we read that “when the Magi were going on their way, there appeared 2r - Lc Blant, Les sarcophages Chretiens do la Gaule, pi. XXXVI/2. 21 Michel, Evangiles apocryphes, Paris, 1911, p. hi. 36 EARLY CHRISTIAN ICONOGRAPHY to them the star, which was, as it were, a guide to them, going before them until it came to where the Child was. And when the Magi saw the star, they rejoiced with great joy.” The only other example of the scene in the early Christian period is on the ambo of Salonica whose artistic connections with Syria have been Fig. 28. S. Kensington: Victoria and Al- bert Museum, ivory casket. The Magi gaz- ing at the Star. pointed out by Bayet 22 and are shown by my own classification of the Adoration of the Magi. 23 The presence of this scene on the ambo, while its composition is quite distinct from that of the Provencal sarcophagi and ivories, confirms the Syrian inspiration of the work. In the eighth century the scene occurs on an ivory from Egypt at Bologna in a form resembling the Provenqal type. The same scene appears after the ninth century in several Ottoman manuscripts (see Fig. 47) and in Italy among the frescoes of S. Urbano alia Caffarella and in panels of the bronze doors of the cathedrals at Pisa and Monreale. VI THE ADORATION OF THE MAGI The Epiphany is as complicated in art as it is involved in liturgy. With few exceptions the distinctive feature which separates the scenes of the West from those of the East during the first eight centuries of Chris- tian art is the aspect of the Magi, who are beardless on the monuments of the West and on those of the East are either all bearded or are differenti- 22 Duchesne and Bayet, op. cit., p. 249 sq. 23 See Adoration of the Magi, p. 55. EARLY CHRISTIAN ICONOGRAPHY 37 ated in age by the first Magus having a heavy beard, the second Magus hav- ing a light beard, and the last being beardless. Also the frontal Madonna in the scene is an Oriental characteristic never seen on the monuments of the West save as it was introduced directly from the East. The main types 1 of the Adoration in art are the Hellenistic, the Hellenistic-Oriental, the Alexandrian-Coptic, the Palestinian-Coptic, and the Byzantine. Before discussing the classification of the scenes of the Adoration of the Magi, it is necessary to mention the highly symbolic composition in the fourth cen- tury arch mosaics of Santa Maria Maggiore, 2 which defies all classification. It is called by Richter “The Coming of the Orient to Christ’’ (Fig. 29). Fig. 29. Rome: S. Maria Maggiore, mosaic. Adoration of the Magi. Christ is figured as a youth dressed in a long white tunic with purple clavi and seated on a large, jewelled, quadrangular throne with a gemmed suppe- daneum. The miraculous star is poised above His head and four angels stand behind His throne. To the right of the central group is a small throne on which is seated a woman enveloped in sombre purple, who has been variously interpreted as a Sibyl, an attendant, and as the personifica- tion of Ecclesia ex Circumcisione. The seated Byzantine figure at the left, apparently a Madonna, is an interpolation of the eighteenth century. 3 1 Kehrer, Die Heiligen drei Konige in Literatur und Kunst, Leipzig, 1908 ; Heisenberg, Grabeskirche und Apostelkirche, Leipzig, 1908, II, p. 229; Liell, Die Darstellung der aller- heiligsten Jungfrau, Freiburg \/b, 1887; Rohault de Fleury, L’Bvangile ; Lehner, Die Marienverehrung in den ersten Jahrhunderten, Stuttgart, 1886; Dalton, Byzantine Art and Archaeology, p. 654; Duchesne and Bayet, Mem. sur une Mission au Mont Athos, 1876, p. 249-300. 2 Table II, no. 1 : Richter and Taylor, The Golden Age of the Classic Christian Art part II, ch. v. 3 Richter and Taylor, op. cit., p. 331. EARLY CHRISTIAN ICONOGRAPHY 38 In all probability one of the Magi has been destroyed for now only two Magi to the right of the throne advance with gifts. West. The Hellenistic Type. The general characteristics of the Hellenistic type are : the beardless Magi advancing in a line, the young Jesus of two years sitting in His mother's lap, the Virgin at first represented sitting in a three-quarters attitude and later in profile, and the absence of Joseph or any figures to give the scene the appearance of an actual event. This Hellenistic type may he subdivided as it presents different variations into several groups : an early group where the Virgin sits in a Hellenistic three-quarters at- titude, a rare “Epiphany type” where the Magi worship the young Jesus in His manger, a later group where the Virgin is represented in profile, and the Provencal type where the Magi usually wear the peculiar notched chiton referred to in the discussion of the preceding scene. The reasons for calling this type Hellenistic are : the early appearance of the scene in Christian art, the Hellenistic character of the Virgin’s attitude, and the derivation of the Magi from the representations of the Mithraic cult so numerous in late Hellenistic art throughout the Mediter- ranean world. In the first place this type began as early as the second century when Christian art had only Hellenistic models to which to turn. Second, the easy posture of the Virgin, who on all the monuments of the first group sits in a three-quarters pose, belongs rather to the art of the catacombs than to that of the sarcophagi. The relation of the Christian Magi in the scene of the Adoration to the Hellenistic representations of Oriental priests has been shown by a first century relief from the Villa Borghese published by Bienkowski . 4 The costumes of the two priests on this relief, including the Phrygian caps, the short chitons, the chlamydes and the anaxyrides, are identical with the garments which are worn by the Magi on the sarcophagi; furthermore, the priests bear gifts to a god- dess, the one at the right extending a large wreath on veiled hands while the other priest at the right presents a vase in the same manner. This form of presenting gifts on veiled hands is usually followed on the sar- cophagi and even the wreath and the vase are frequently figured among the gifts. 4 See Von Sybel, Rom. Mitt., 1912, p. 311 sq., fig. 1. EARLY CHRISTIAN ICONOGRAPHY 39 Besides this first century prototype for the Magi and their gifts, an- other instance of the influence of classical art upon the Christian scene is afforded by a sarcophagus relief at Rome from the Villa Doria Panfili. 5 Here in a scene of the Adoration the Virgin with the young Jesus on her knees lies in classical fashion on a couch. Although unique among the representations of the Adoration, this example shows the tendency of the artists even of the fourth century to turn to Hellenistic models. Another Hellenistic feature of this type is its general symbolic char- acter. The absence of Joseph or any other figures to mark the scene as an actual event emphasizes the symbolism which depicts the Magi as the first converts from paganism. 6 During the first five centuries Christian art, besides being Hellenistic in character, was not primarily narrative, but symbolic. To sum up, tbe Hellenistic character of this first type is shown by its dependence on Hellenistic models, its symbolic rather than narrative conception, and its early inception when Christian art was purely classic in its forms. This Hellenistic type was not peculiar to Rome and the West but was generally known throughout the Mediterranean basin. While all the West- ern examples of the Adoration prior to the eighth century belong to this type, all the examples of the Hellenistic type are not Occidental. As far as its Oriental and Hellenistic prototypes go, it might well have originated in any one of the large cities of the East or even in Rome itself; but it is more likely that the type came from Alexandria, the great centre of Hellenistic art and the most important centre of Hellenistic Christianity. The fact that the Feast of the Epiphany first originated in Alexandria sustains the probable Alexandrian origin of the Hellenistic type. In Alexandria, early in the second century, the Feast of the Epiphany was celebrated on the eleventh of Tybi or the sixth of January in the Gregorian calendar; 7 from that centre it spread in the East over Syria, Asia Minor, north to Constantinople and in the W est to Gaul and Spain. 8 The first patristic reference to the triple gift of gold, myrrh, and frankincense comes in the second century from Origen 9 of Alexandria. St. Irenaeus, 1 " bishop of Lyon from 170, developer] the complicated •’Table II, no. 2. '• \\ ilpert, Le Pitture delle catacombe Roinane, p. 176. ’ Kehrer, op. cit., I, p. 24. * Kehrer, op. cit., I, 24, 25. 9 Kehrer, op. cit., I, p. 13, note 7. 10 Kehrer, op. cit., I, p. 12. EARLY CHRISTIAN ICONOGRAPHY 40 theological significance in these gifts. In the gold, incense, and myrrh Irenaeus saw the symbols of Christ as King, God, and mortal. While Clement 11 of Alexandria propounded that the gold itself was the incor- ruptible Logos, proof against the corruption of rust, the symbolism of Irenaeus was widely accepted and in the West was restated by Sedulius 12 and Prudentius . 13 And yet, while the Gospel of Matthew and practically all the theologians of both the East and the West state that the triple gift of the Magi was gold, myrrh, and incense, the gifts in the scenes of the Adoration on the Western sarcophagi show a wide range of objects . 14 In fourteen sarcophagus representations the first Magus carries a wreath resting on a cloth . 15 A Hellenistic prototype of this wreath has already been seen in the wreath offered by the priest on the first century relief from the Villa Borghese. Some such image seems to have been in the mind of Clement of Alexandria , 16 who, in developing his symbolism of the Logos, frequently refers to the “golden wreath” which the Magi presented to the Christ Child. It is significant that the literary parallel for this persistent and peculiar detail of the Hellenistic Epiphanies is found in a Christian writer of Alexandria. Group I. The Virgin seated in a three-quarters posture. This group 17 includes both catacomb and sarcophagus representations and the consistent characteristics are the beardless Magi, the Virgin seated in a three-quarters attitude, and the Child of two years in His mother’s lap. While on a majority of the examples the Magi are three in number and advance with their gifts in a line, among the catacomb representations there are three scenes where the number of Magi is only two, one scene where they are four, and two examples where they do not advance in a line but are arranged symmetrically. It is to be noted, however, that in the second century fresco of the Greek Chapel 18 the scene commences in the form which later becomes universal. This form is illustrated by 11 Kehrer, op. cit., I, p. 13. 12 Kehrer, op. cit., pp. 33-34. 13 Wilpert, op. cit., p. 176; Prudentius, Catliemerinon, XII, 19 in Migne, P.L., LIX, 905 - 11 Many of the gifts of the Magi can not be recognized and may well be taken to rep- resent any one of the triple gifts, which have no marked distinguishing features. In most cases the gifts represented appear to have been a matter of choice with the artists. On three sarcophagi (Table II, nos. 15, 20, 29) one of the Magi carries a dish containing two doves which suggests that the idea of purification was implied. 15 Table II, nos. 19, 20, 21, 24-30, 33, 37, 42, 43- 16 Kehrer, op. cit.. I, p. 13. 17 Table II, Group 1. 18 Table II, no. 3. EARLY CHRISTIAN ICONOGRAPHY 4 1 the early fourth century fresco in the cemetery of Callixtus shown in Fig. 30. In the two frescoes in the catacomb of SS. Pietro e Marcellino 19 where there are only two Magi, the number was reduced, as Wilpert 20 says, be- Fig. 30. Rome: Catacomb of Cal- Fig. 31. Rome: Catacomb of SS. Pietro e lixtus. Adoration of the Magi. Marcellino. Adoration of the Magi. cause of the lack of space and from the desire to balance the panel with a scene on the other side of the cubiculum. In the case of the fresco (Fig. 31 ) in SS. Pietro e Marcellino, 21 where the two Magi are arranged one on either side of the Virgin, and in Domitilla, 22 where the four Magi are placed two on either side of the Mother and Child, the reason was a frank desire to obtain symmetry. Moreover, during the earlier centuries there was no absolute obligation to represent the number of Magi as three. While the Christian Fathers agreed in stating that the Magi brought a triple offering and it was assumed that each Magus brought one gift, it was not until the sixth century that the three Magi were named in the Syrian “Spelunca Thesaurorum.’' 23 Therefore, it is not surprising to find in Rome during the third century and the early part of the fourth century some freedom, caused by the exigencies of space and the desire for symmetry, in the number of the Magi represented. Group II. The “ Epiphany Type ” 24 (Fig. 32). This group, which is composed of six sarcophagus representations of the Adoration, is characterized by the beardless Magi advancing in a line with their gifts to the young Jesus who lies in His crib under a manger 19 Table II, nos. 5, 6. 20 Wilpert, op. cit., p. 179. 21 Tab,e H. no. 4. 22 Ta bi e n, no . - -Kehrer, op. cit., I, p. 65; Bezold, Spelunca Thesaurorum, Leipzig, 1883, p. 57. - 4 Kehrer, op. cit., II, p. 21 sq . ; while isolating this “Epiphany type” Kehrer also at- tempts to establish another combination type which he calls the “Christmas type.” ThL last, as he describes it, shows the Adoration of the Magi combined with a con- tiguous scene of the Nativity which would, therefore, date the monuments on which it occurs after 354, for after that date the Nativity, formerly celebrated by the Roman 4 2 EARLY CHRISTIAN ICONOGRAPHY roof. 25 Kehrer’s reason for calling this group the “Epiphany type” is that it depicts the Adoration as happening at the time of the Birth and because at the feast of the Epiphany on January 6 the Adoration and the Nativity were celebrated together. The date of this group, however, would be anterior to the middle of the fourth century or not much later, for the well known Calendar of 354 gives the Nativity as a separate feast Fig. 32. Ancona: sarcophagus. Ado- Fig. 33. Rome: Sarcophagus. Ado- ration of the Magi. ration of the Magi. on December 25, showing that by that time the two festivals had ceased to be synchronous in Roman usage. It is also to he noted that on this early group of sarcophagi the Virgin sits in a three-quarters attitude. Group III. The Virgin in pro file. 26 This group, which is largely composed of representations on sarcophagi of Rome (Fig. 33), offers no features differing from the scene on the sarcophagi of the first group save the fact that the Virgin is no longer represented in the Hellenistic three-quarters posture but in profile. Theo- retically the monuments of this group should be later than those of the first group, reflecting as they do the later tendency toward a frontal or a profile rendering of the figure. Group IV. Odd examples of the Hellenistic Type in the West which are connected with Eastern art centres . 21 That the Hellenistic type was not peculiar to the West, but was used throughout the Mediterranean basin, is shown by at least two examples of this group. A fifth century panel of the carved wooden doors of Santa church along with Epiphany on Jan. 6, was made a separate feast on December 25. While I have kept the Epiphany type in my classification, I have not isolated the “Christmas” type because on the three sarcophagi (Table II, nos. 25, 26, 31), on which Kehrer says that it appears, the scene that he calls a Nativity I have shown to be an Adoration of the Shepherds (see Nativity, p. 14) and not an actual representation of the Birth. 25 Table II, Group 2. 2,; Table II, Group 3. 27 Table II, Group 4. EARLY CHRISTIAN ICONOGRAPHY 43 Sabina 2S at Rome, the style of which is undoubtedly that of some one of the great Hellenistic centres in the East, represents (Fig. 34) the three Magi, beardless, advancing in a line toward the Virgin, who is seated in a three- quarters position on a high podium, holding the two year old Child in her Fig. 34. Rome : S. Sabina, panel of doors. Adoration of the Magi. lap. On the sarcophagus 29 of the Armenian Exarch Isaac (d. 643 A. D.), formerly in S. Vitale and now in the Royal Museum at Ravenna, occurs this same Hellenistic type. This sarcophagus, as is shown in the dis- cussion of the scene of the Raising of Lazarus which occurs on it, 30 may be connected with the East through an Oriental relief recently found in Phrygia. The two amulets in the Vatican which are as late as the sixth century, as the use of the nimbi on the Virgin and Child would show, can not be located, although their style is that of the amulets coming from Palestine and Egypt. 31 Group V-A. Proven gal Type. The main feature of the Adoration on the ivories assigned to Provence is the curious notched chiton (already noticed as a Provenqal characteristic in the Appearance of the Star 32 ) which the Magi wear. The same chitons are used in the Adoration on a Provengal sarcophagus at Saint-Gilles. 33 On the Milan book covers 34 (Fig. 35), in addition to such details as the beardless Magi advancing in a line toward the young Child of two years seated in His mother's lap, the scene is represented in a house which ap- pears as a brick wall extending across the back of the composition. It 2,5 Table II, no. 38. 20 Tab]e n, no. 41. 30 See Raising of Lazarus, p. 116. 31 Table II, nos. 39, 40. 32 See p. 34. Comparison of Fig. 34 with Fig. 26 will show how the “notched” chiton worn by the Magi on the doors of S. Sabina differs from the Provengal garment. 33 Table II, no. 43. 34 Table II, no. 46. 44 EARLY CHRISTIAN ICONOGRAPHY is not only the representation of the scene in a house but also the manner of representing it that connects the scene on the other ivories of the group with the rendering on the Milan covers. The scene on the Werden casket (Fig. 36) is similar to the Adoration of the Milan ivory 35 in that the Virgin, heavily veiled, is seated in a three-quarters position on a cathedra ; in her lap is seated the young Jesus clothed in a long tunic; and from left to right the three Magi, beardless and wearing the odd notched chiton, ad- Fig. 35. Milan : Cathedral, ivory BOOK COVERS. ADORATION OF THE MAGI. Fig. 36. S. Kensington : Victoria and Al- bert Museum, ivory casket. Adoration of the Magi. vance in a line with their gifts. The representation on the Nevers frag- ment (Fig. 164) is almost identical with that on the Milan ivory, even to the brick structure in the background. 36 Group V-B. Continuation of the Hellenistic Types into Carolingian Art. The Provenqal rendering, where the Magi, all beardless and wearing notched chitons, advance in a line with their gifts for the Christ Child of two years, who sits in His mother’s lap, was transmitted about 800 directly into Carolingian art. It appears, even with the brick build- ing in the background which characterizes the representation on the Milan and Nevers ivories, in the Bodleian book cover at Oxford 37 (see Fig. 169.) In the same century the type occurs not only in the Sacramentarium of Drogo 38 (Fig. 37), and a silver casket in the Sancta Sanctorum, 39 but also on four ivories. 40 The main difference in these scenes from that on the Bodleian book cover is the introduction of the plain nimbus for the Virgin and the cruciform nimbus for the Saviour. The Adoration on a Lombard relief in the Berlin Museum (Fig. 38), where the Hellenistic feature 35 Table II, no. 47. See also p. 226, note 14. 30 Table II, no. 45. 37 Table II, no. 48. 38 Table II, no. 50. 3n Table II, no. 53. 40 Table II, nos. 51, 52, 54, 55. EARLY CHRISTIAN ICONOGRAPHY 45 of the beardless Magi is preserved, differs from the other examples of the group in that the Child lies in His crib, and Joseph is figured in the composition. This Lombard scene is, therefore, more directly a continua- Fig. 37. Paris: Bibl. Nat., Sacramentary of Drogo. The Magi. tion of the so-called Epiphany type, seen in Group II and illustrated in Fig. 32. THE HELLENISTIC TYPE AND PSEUDO-MATTHEW The possible texts which would in any way influence or confirm the use of this Hellenistic type are rare. Apparently, as an artistic type, it Fig. 38. Berlin: Museum, Lombard relief. Adoration of the Magi. grew up in direct contradiction of an early tradition that the Birth and the Adoration were synchronous. On all the monuments of the Hellenistic type, save for the small Epiphany group, the young Jesus is represented as a child of two years. The fact that in Rome prior to the middle of the fourth century the Nativity was celebrated as a part of the Adoration and that all the Eastern writers speak of the Magi finding the Child in a cavern would show that the more prevalent tradition among the 46 EARLY CHRISTIAN ICONOGRAPHY theologians was that Jesus was just born when the Magi came to wor- ship Him. 41 Nevertheless, there must have existed also the tradition that the Adora- tion took place two years after the Birth, for the symbolic character of the Hellenistic type does not explain the Child sitting in His mother’s lap. The Gospel of Matthew (ii, n), which is the only canonical book that relates the Adoration, vaguely suggests, in saying that Herod killed all the babes of two years and that when the Magi “were come into the house, they saw the young Child with Mary His mother,” that Jesus was two years old when the Wise Men came to offer Him their gifts. The translator of Pseudo-Matthew, however, when he was turning his apocryphal sources into Latin evidently sought to reconcile his text with the prevalent custom of representing Jesus as a child of two. We read, 42 “Two years having passed, some Magi came from the Orient to Jerusalem,” and, “entering into the house, they found the Child Jesus rest- ing on His mother’s breast.” At the time when the book was translated the West had separated the Birth from the Adoration by establishing the new feast on December 25, and the Hellenistic type had become the regular Western type. Thus, as in the case of the Adoration of the Ox and the Ass, the Pseudo-Matthew, when it was translated, was brought into accord with an iconographic tradition already established in the West. East. The main distinction between the Adoration scenes of the East and those of the West is, as said before, the aspect of the Magi. The Magi in the West are beardless, while on the Eastern monuments they are either all bearded or, as is more often the case, are differentiated in age and importance, the first Magus being heavily bearded, the second lightly bearded, and the last beardless. This attempt to give individuality and separate distinction to each of the Wise Men was characteristic of the Oriental attitude toward all the events of the life of Christ. The whole 41 Justin Martyr tells of the Magi finding the Child in a cavern (Kehrer, op. cit., I, p. 12) ; John Chrysostom dwells on the humility which permitted the Son of God to be exposed to the gaze of shepherds and Magi in a cavern (Kehrer, p. 17) ; the Syrian Balaeus (p. 17) repeats the tradition, which is carried on by Petrus Chrysologus (p. 81), and is restated in the Spelunca Thesaurorum and in the Ethiopic Vita Adami (Kehrer, p. 18). 42 Michel, op. cit., p. 109. EARLY CHRISTIAN ICONOGRAPHY 47 tendency of Christian art from the fifth century on in the East was to make the impersonal, symbolical art of the earlier centuries more his- torical and narrative in character. It is for this reason that we may call the differentiation of the Magi an Oriental feature. That the Eastern types, with the exception of the Byzantine, disregard the account in the Protevangelium of James and of the Eastern theologians which says that the Magi came to worship on the night of the Nativity, and depict Jesus as a child of two in His mother’s arms, indicates that they are derivations and indigenous adaptations of the Hellenistic type. The types of the Adoration of the Magi which stand out on the mon- uments of the East are the Oriental Hellenistic type, the Alexandrian- Coptic type, the Palestinian-Coptic type, and the Byzantine type. As characterized by their most distinguishing features : the Oriental Hellen- istic type is the same as the regular Hellenistic type save that the Magi are differentiated in age; the Alexandrian-Coptic is characterized by the Yirgin in a three-quarters position and the introduction of an angel who points out the Child to the Magi, who are again differentiated; the Pales- tinian-Coptic is the same as the Alexandrian save that the Virgin is always represented in strict frontality; and of the Byzantine types, one is a continuation of the Alexandrian-Coptic, while the second is an original rendering derived from the liturgy. This differs from all the others in showing the Child new born in His crib, and in uniting with the Adoration of the Wise Men the Adoration of the Shepherds and the Angels. Group YI-A. The Oriental Hellenistic Type. This is the direct continuation in the East of the Hellenistic type; neither the angel nor Joseph occurs in the composition and the three Magi advance in a line to present their gifts to the young Jesus of two years in the lap of the Virgin, who sits either in profile or in a three-quarters pose. The only difference is the Eastern characteristic of the bearded Magi. The type appears on the encolpium in the Ottoman Museum at Constantinople 43 (Fig. 39), in very similar form on the Strzygowski en- colpium mentioned before 44 (p. 23), an ivory pyxis at Florence, 45 a sarcophagus 46 at Ravenna, and an ivory at Rouen. 47 43 Table II, no. 56. 44 Table II, no. 57, and Errata. 45 Table II, no. 58. 4(1 Table II, no. 60. 47 Table II, no. 59. 4 8 EARLY CHRISTIAN ICONOGRAPHY Group VI-B. The Oriental-H ellenistic Type in the West. This group represents the continuation of the Oriental-Hellenistic type in the West. With the exception of the eighth century Lombard relief 48 at Zara, where the first two Magi are bearded and the last is beardless, Fig. 39. Constantinople: Imp. Ottoman Museum, encolpium. Adoration of the Magi (see Fig. 15). all the examples listed are Carolingian and Ottoman ivories 49 and show all three Magi bearded. The introduction of this type into the Western art of the ninth and tenth centuries may very likely have come either through Ravenna or from the numerous small objects which were brought back to the West by pilgrims from Palestine. Group VII-A. The Alexandrian-C optic Type. The chief characteristic which differentiates this type from the Hellen- istic type is the introduction of the angel who, standing between the Virgin and the first Magus, points out the Child to the advancing Wise Men. The Virgin on all the examples of this type sits in a three-quarters position, or at least her head is turned in that pose which was characteristic of the Hellenistic portraits of Alexandria. The Magi are represented advancing in a line and on the early Alexandrian examples are always differentiated one from another by their beards while on the later Coptic examples they are sometimes beardless. The reason for calling the type Alexandrian-Coptic is its appearance on the Maximianus chair, whose Alexandrian origin has already been pointed out (p. 24), and on several later monuments of unquestionably Coptic origin. The first example of the type is a fourth century relief, 50 found in Carthage. Although the relief is badly mutilated, it still shows great beauty of workmanship. The Virgin and Child sit in a most natural and 48 Table II, no. 62. 49 Table II, nos. 63-66. 50 Table II, no. 67. EARLY CHRISTIAN ICONOGRAPHY 49 easy three-quarters position and before the fine classic seat on which the Yirgin sits stands a winged angel pointing out to the three Magi (who unluckily are badly damaged) the young Jesus seated in His mother’s lap. Behind the Mother and Child are traces of two male figures in whom Delattre recognizes the two prophets’ 1 who foretold that a virgin would bring forth a child to the world. If Delattre is right, this intro- duction of the prophets reminds one of Alexandrian exegetics which per- sistently sought for parallels to the New Testament in the characters and incidents of the Old. The fine execution, the classical poise and draperies of the figures, and the bold, high relief show that the monument must have been the product of the best Hellenistic workmanship in the fourth century. 52 To this same century belongs a relief with the Adoration of the Magi at Kahrie Djami 53 at Constantinople which displays essentially the same Hellenistic workmanship and the same composition save for the fact that the two prophets are omitted and the figure of Joseph introduced in their place. After these fourth century examples, the type occurs first on the Maximianus chair. 54 The panel with the three Magi is lost but on the part that remains the Yirgin appears in the same three-quarters position, the angel, standing between the Madonna and the first Magus, points out the Child to the Magi, and behind the Virgin's cathedra is Joseph (Fig. 40). On the ivory book covers of the Etschmiadzin Gospels 55 the composi- tion ie strikingly similar to that on the Maximiamus chair. The high backed armchair of wicker in which the Yirgin is seated is identical with her seat on the Maximianus chair; on both ivories the winged angel stands in the same position between the first Magus and the Madonna, and Joseph appears on both monuments. According to the description of Trombelli 56 the Magi who are now gone from the ivory chair were prostrate on their knees. The tunics and anaxyrides of the Magi on the Etschmiadzin cover are identical with the costumes worn by the Egyptians in the Joseph 51 Delattre, Musecs dc I’Algcrie, III, p. 1. 52 Strzygowski, Dcnkschriftcn dcr Kaiserlichen Akademie der Wissenschaftcn, ( Philos Hist. Klasse), Vienna, 1906, p. 125, 141, describes a fragmentary Adoration in the Alex- andrian "World-Chronicle” (c. 400), of which only traces of the Virgin and Child remain. 53 Table II, no. 68. 51 Table II, no. 69. 55 Table II, no. 70. Garr., VI, p. 21; Trombelli, De cultu SS. diss., IX, p. 247. The panel was in the Baruffaldi Collection. 50 EARLY CHRISTIAN ICONOGRAPHY Fig. 40. Ravenna: Archbishop’s Palace, IVORY PANEL OF CHAIR OF MaxIMIANUS. PART of Adoration of the Magi. scenes 57 of the Maximianus chair, with the costumes of the gift bearing Orientals on the Barberini 58 diptych of Alexandria, and also with those worn by a figure on the Coptic pyxis of St. Menas 59 in the British Museum. In further evidence of Egyptian origin the Adoration occurs on two Coptic textiles (Eig. 41) and on a Coptic medallion 60 with the same essen- tial feature of the angel. In passing, however, from Lower Egypt to the more indigenous art of Upper Egypt, the type lost all its Hellenistic style. Also on the textile examples the Magi are beardless and instead of ad- 57 Garr., op. cit., VI, pi. 421, 1, 3. 58 Diehl, Manuel, fig. 141. 59 Dalton, Catalogue of Early Christian Antiqnitic s in the British Museum, 1901, no. 297, pi. IX. 00 Table II, nos. 71, 72, 73. EARLY CHRISTIAN ICONOGRAPHY 5 * vancing in a line they are scattered about in the composition. This casual arrangement of the Magi is due either to a primitive horror vacui of the crude Coptic art or to the fact that all the compositions are within circles which do not afford space for representing the Magi in a line. Group \ II-B. The Byzantine continuation of the Ale xandrian-C optic Type. This type, which was fundamentally Hellenistic in character, was not popular, and from the sixth century it encountered the competition of Oriental types which in time supplanted it. It appears in the eighth century Fig. 41. Strassburg: Forrer Collection, TEXTILE MEDALLION FROM AkMIM, EGYPT. Adoration of the Magi. in Italy in a mosaic of the chapel of John VII 61 in old St. Peter’s, a portion of which is now in the sacristy of Santa Maria in Cosmedin 62 and on a Lombard relief. 60 It also occurs in the ninth century in the Byzantine manuscript No. 510 in the Bibliotheque Nationale at Paris 64 (Pig. 42) and in another Greek manuscript of the tenth century. 65 In the eleventh century the type appears at Rome in the frescoes of Sant’ Urbano 66 and again at Athens in the mosaics of Daphni. 67 THE PALESTINIAN-COPTIC TYPE The dominating features of this Palestmian-Coptic type are its monu- mental character and its tendency to stylized, symmetrical compositions. The A irgin, instead of being represented either in profile or in three- 61 Table II, no. 75. 62 Table II, no. 76. 63 Table II, no. 77. 61 Table II, no. 78. 65 Table II, no. 79- 66 Table II, no. 81. 67 Table II, no. 82. 5 2 EARLY CHRISTIAN ICONOGRAPHY quarters, is always depicted in strict frontality. The tendency for stylized and symmetrical compositions appears in the fact that either the Adora- tion of the Magi is balanced by the Adoration of the Shepherds on the other side of the Madonna or that the Magi are arranged two on one side of the Virgin and the other Magus and an angel are placed on the other side to provide the needed symmetry. Invariably the Magi are differentiated in age by their varying beards and one or more angels Fig. 42. Paris: Bibl. Nat. Ms. gr. 510. Adoration of the Magi. are figured in the scene. When the Adoration of the Shepherds is joined with the Worship of the Wise Men there are two angels present, one summoning the Magi and the other calling the shepherds to the scene. The Oriental Christians, while they multiplied details in their endeavor to give to all the scenes of the Bible as much historical reality as possible, were not realistic in their artistic methods and always tended to stylize their forms of expression. As the ancient Egyptians and the Assyrians formalized their art, so the Oriental Christians, when their indigenous tendencies began to assert themselves against the older Hellenistic tradi- tions, always tended to represent their figures in their broadest aspect and to give their compositions a symmetrical, monumental character. The frontal Madonna, so hieratic in character, which is the distinguish- ing feature of the Palestinian-Coptic type, is purely an Oriental creation. The representation of the frontal Madonna with a nimbus, whether she appears solely as the Theotokos, in the scene of the Flight into Egypt, or in the Adoration of the Magi, is to be considered as a mark either of Eastern origin or of direct Eastern influence. EARLY CHRISTIAN ICONOGRAPHY 53 Group VIII-A. The Palestinian Group of the Palestinian-C optic Type. The type originated in Palestine. This is shown by its appearance on numerous monuments which may be definitely connected with the Holy Land, by its monumental character which recalls the mosaics of the Church of the Nativity at Bethlehem, and by its adherence to the Syrian liturgy. In the first place the type appears on three of the metal phials of Monza which, as their inscriptions show, were executed in Palestine toward the end of the sixth century and were brought back to the West by devout pilgrims who had visited the Holy Land. The scene on two 68 of the phials shows the strictly frontal Virgin nimbed and seated on a bolster while she holds the two year old Jesus, who has a cruciform nimbus, directly tiox of the Magi axd Shepherds. in front of her; over her head is figured a large star at which two winged and nimbed angels point, the angel at the left summoning the Magi and the one at the right calling the shepherds from their flocks; the three Magi at the left of the throne are dressed in the traditional Oriental costume, carry their gifts on dishes, and are differentiated one from the other by their beards; at the right of the throne the three shepherds give the balance to the composition. On the third 09 phial, while the scene preserves the same details and general arrangement, there is a more 09 Table II, no. 85. 68 Table II, nos. 83, 84. 54 EARLY CHRISTIAN ICONOGRAPHY stylized composition; the shepherds instead of being freely grouped at the right of the throne are depicted standing in a row which rigidly bal- ances the row of three Magi on the opposite side (Fig. 43). Another example of the symmetrical composition with the frontal Madonna occurs on a full page miniature of the Etschmiadzin Gospels 70 (Fig. 44) which Strzygowski 71 dates in the first half of the sixth century and attributes to Syria. A further connection of the Palestinian-Coptic type with Palestine Fig. 44. Etschmiadzin : Gospels. Adora- tion of the Magi. is to be found in the hieratic and monumental character of the composition which appears to reflect the larger design of a mosaic. In a polemic against the Iconoclasts issuing from the Synod of 836 at Jerusalem 72 it is stated that the Nativity church at Bethlehem was decorated by mosaics represent- ing the Birth of Christ, a Panagia with the Christ Child in her bosom, and the Adoration of the Magi. This Adoration on the western facade is said to have been spared in the sixth century by the soldiers of Chosroes because the Magi wore Persian caps. 73 It is then possible, as Dalton says, 74 that this type as it appears on certain of the smaller objects of Christian art of 70 Table II, no. 86. 71 Strzygowski, Bys. Dcnk., I, p. 74. 72 Strzygowski, Hell, und Kopt. Kunst, p. 92; Oriens Christ., V, 1915, p. 102. 73 Dalton, Bys. Art and Arch., p. 384. 74 Dalton, op. cit., 182. EARLY CHRISTIAN ICONOGRAPHY 55 the fifth and sixth centuries was copied from the mosaics of the Church of the Nativity at Bethlehem, an assumption borne out by the monumental aspect of the composition pointed out above. Furthermore, the adherence of this type to the Syrian liturgy confirms its Palestinian origin. The evident importance of the star and the sum- moning of the shepherds on these Palestinian monuments is paralleled by the importance given to the appearance of the star by the Syrian writers and by the fact that in the Syrian liturgy 75 the feast of the Adoration of the Shepherds was celebrated at the same time as Epiphany, on Jan- uary 6. In the fourth century both Ephraem Syrus 76 and Balaeus 77 (373- 431) write of the appearance of a new star in the heavens at the time of the Nativity and associate with it an angel. By the end of the fifth century the Syrian legend in the “Spelunca Thesaurorum” 78 develops even more fully the appearance of the new star and the summoning of the Magi by an angel. Passing directly from Palestine to other Mediterranean lands where the Syrian influence was very strong, the type occurs in the sixth century on the ambo of Salonica 79 and in the mosaics of Sant’ Apol- 75 Kehrer, op . cit ., I, p. 24. 7fi Kehrer, op . cit ., pp. 14, 15, 21. 77 Kehrer, op . cit ., p. 21. 78 Kehrer, op . cit ., p. 18. 79 Table II, no. 87. F:g. 45. Manchester: John Rylands Li- brary, PANEL FROM IVORY BOOK COVERS OF Muraxo. Epiphany and Nativity. EARLY CHRISTIAN ICONOGRAPHY 56 linare Nuovo at Ravenna 80 and later in the tenth century in the Armenian miniatures which were inserted into the Etschmiadzin Gospels. 81 Group VIII-B. The Egyptian Examples of the Palestinian-C optic Type. The Palestinian type of the Adoration is exactly duplicated on two ivory diptychs of the sixth century which are of Coptic workmanship and were probably executed in the Thebaic!. In the central panel of the Murano book covers in Manchester 82 (Fig. 45) and of a five part diptych in the British Museum 83 (Fig. 46), which have every mark of Coptic Fig. 46. British Museum: ivory diptych. Adoration of the Mag . art, 84 even to the Coptic Nativity with the Doubting Salome, is the same composition which appears in the Etschmiadzin miniature. On a Coptic pyxis 83 in the Vienna Museum is depicted an Adoration in which the Virgin sits in a frontal position, but the composition differs from that on the other Coptic ivories mentioned in that the Magi advance in a line with their gifts and no angel is represented. The similarity in type between the Coptic and the Palestinian examples, as Strzygowski says, 86 does not mean that the ivories were not done in Egypt, but only goes to show, as in the case of the Palestinian-Coptic type of the Nativity, 80 Table II, no. 88. 81 Table II, no. 89. 82 Table II, no. 91. 83 Table II, no. go. 84 Strzygowski, Hell, und Ko[>t . Kunst, p. 85. 8r ' Table II, no. 9 2. 86 Strzygowski, op. cit., p. 92. EARLY CHRISTIAN ICONOGRAPHY 57 that Syria and Egypt were going along hand in hand in the matter of iconography. There are two other unique and baffling examples of the Adoration which do not really classify under the Palestinian-Coptic type but inas- much as they show the frontal Madonna and were perhaps executed in Alexandria, I have included them in the Coptic group of this Palestinian type. The silver casket 87 in the church of San Nazzaro at Milan, which is dated as early as the fifth century and is supposed to have been executed in Alexandria, shows a composition of the Adoration of the Magi too individual and unique to be considered as representative of any type. The Virgin and Child are frontal and the worshippers are arranged on either side of the Madonna. Still the costume and the number of the figures adoring the Christ Child fail to give any idea of their character as Magi, if such they were really meant to represent, for there are eight of them and they are not dressed in the traditional Phrygian garb. The scene on the sixth century (?) column of the ciborium of San Marco 88 defies classi- fication because its restricted and crowded niche composition made it im- possible to represent adequately any actual type of the Adoration. The Madonna and Child are still represented in strict frontality; the Magi, who are differentiated in the usual manner by their beards, advance in a line, but no angel appears in the scene. Group IX. An Ottoman Manuscript Type. While the scene of the Adoration as it is represented in several Ot- toman manuscripts 89 of the tenth and eleventh centuries is in no wise a continuation of the Palestinian-Coptic type, it shows, as is to be ex- pected from the fact that these manuscripts drew much of their iconography from Byzantine art, many Eastern features. The nimbed Madonna is represented in more or less frontality, the Magi who advance in a line are differentiated by their beards, and in the Codex Egberti (Fig. 47) and the Gospels of Henry the Second the Appearance of the Star to the three Magi occurs along with the Adoration. The most interesting feature of this group is the appearance of crowns instead of Phrygian caps on the Magi in these two last mentioned manu- scripts of the tenth century, marking the earliest occurrence in the West 17 Table II, no. 93. See p. 87, note 7. 88 Table II, no. 94. “ Table II, nos. 93-99. 58 EARLY CHRISTIAN ICONOGRAPHY of the change 90 from the traditional caps. There is no doubt that this new feature was introduced into the West from the Orient, for in the Menologium of Basil II in the Vatican, 91 which dates about 976, the crowns are given to the Magi (Fig. 48). Furthermore, the earliest refer- ence to the crowns appears to be in the Nestorian Liturgy 92 which is at least older than the middle of the seventh century. In this account the three kings, the King of Persia, the King of India, and the King of China, are described as having come to worship bearing crowns. Group X. The Liturgical Byzantine Type. This Byzantine type is a combination of the Adoration with the Nativ- ity and the Adoration of the Shepherds. As early as the year 300 93 the East celebrated the triple feast of the Birth, the Adoration, and the Bap- tism on January 6. Moreover the Eastern theologians, as Justin Martyr, Gregory of Nazianzus, John Chrysostom, and Balaeus Syrus, 94 all con- sidered the Adoration of the Magi as synchronous with the Birth of the Saviour in the cave. When the Western Christmas feast on Decem- ber 25 was adopted in the East 95 it was influenced by older traditions; 90 Dalton in his Byzantine Art and Archaeology, p. 654, says that “crowns first appear in the West in the twelfth century” and that they appear earlier in the East in the Homilies of Gregory of Nazianzus (Paris, Bibl. Nat., gr. 510), and the Vatican Meno- logium. Both statements may be questioned : the crowns certainly appear in the West as early as the end of the tenth century, and the curious round caps on the Magi in the Gregory Codex can not be taken for crowns. 01 Table II, no. 79. 92 Kehrer, op. cit., I, p. 30. 93 Kehrer, op. cit., I, p. 23. 94 See page 46, note 41. 95 Kehrer, op. cit., I, p. 26. Fig. 47. Trier: Stadtbibliothek. Codex Egberti. Adoration of the Magi. 59 EARLY CHRISTIAN ICONOGRAPHY at Constantinople the Adoration, according to the Byzantine Liturgy, 96 was celebrated on December 25, not only with the Nativity but also with the Adoration of the Shepherds, the Conception, the Presenta- tion in the Temple, the Massacre of the Innocents, and the blight into Egypt. By the eleventh century, when Byzantine iconography was fully Fig. 48. Rome: Vatican, menologium of Basil II. Adoration of the Magi. developed and in the process of final stylization under liturgical influence, the Byzantine scene of the Adoration of the Magi had been permanently modified by this combination feast and had commenced to unite others of the events celebrated at that feast with the composition of the Epiphany (see Fig. 25). Hence these examples 97 of the Byzantine type represent not only the Adoration of the Magi, but also the Birth of the Saviour, the Adoration of the Shepherds, the washing of the Child by the midwives, which symbolizes the Baptism, and the Adoration of the Angels. Thus the young Jesus in these scenes instead of being seated in His mother’s lap, lies in a crib with the ox and the ass adoring Him while the Virgin reclines on a mattress. From the eleventh century on, Byzantine art quite faithfully adheres to this type and the same method of representation is followed in Italo-Byzantine monuments of the same period. VII THE MASSACRE OF THE INNOCENTS The Massacre of the Innocents, the earliest event in the long history of martyrdom which marked the struggle of Christianity against the forces 9? Kehrer, op. cit., 1, p. 27. 97 Table II, nos. 106-110. 6o EARLY CHRISTIAN ICONOGRAPHY of paganism, though frequently referred to in the literature 1 of both the East and the West, was seldom represented on the monuments of the early Christian period. The scene never occurs in the catacombs or on the sarco- phagi of Rome. It was in the East that such tragedies of the Gospel his- tory were first felt and depicted in a realistic way. While there is no Western type of the Massacre, there is a single example among the mosaics of Santa Maria Maggiore which, as a purely symbolic rendering of the scene, may be treated under the head of a Hellenistic type. In the East there is the Syro-Egyptian type which may be designated as the sword type in which the soldiers of Herod slay the children at the point of the sword. Finally, there is the Provenqal type which may be characterized in contrast with the Syrian form as the smashing type where the soldiers, instead of using swords, smash the children to the ground. Outside of these three types there are no others, for it was the sword type which became the traditional method of representing the scene throughout the whole history of Christian art. Group I. The Hellenistic or Symbolic Type (Fig. 49). The scene in the fourth century arch mosaics of Santa Maria Mag- giore, 2 from its unemotional character, falls in that class which in other Fig. 49 Rome: S. Maria Maggiore, mosaic. The Massacre of the Innocents. scenes we have called Hellenistic. In the left spandril of the arch Herod, enthroned upon a solium, is represented as a Roman general and his sol- diers as Roman legionaries. The mothers, their hair unbound and hanging 1 The Protevangelium of James, (Michel, Huang He s apocryphcs, p. 45), Pseudo- Matthew {op. cit., p. hi), History of Joseph the Carpenter {op. cit., p. 202), Gospel of Nicodemus {The Apocryphal Gospels, Acts, and Revelations, Edinburgh, 1890, p. 1 35 ) » and Matt, ii, 16. 2 Table III, no. 1. Richter and Taylor, The Golden Age of Classic Christian Art, p. 354: Ainaloff, see abstract in Richter and Taylor, pp. 415-419. EARLY CHRISTIAN ICONOGRAPHY 61 dishevelled down upon their shoulders in token of grief, clasp their doomed infants in their arms and stretch out their hands to Herod in dignified and restrained gestures of supplication. Dignity is the key note of the composition. It is in striking contrast with the barbaric cruelty of the Eastern representations. The whole conception is softened by the spiritual significance presented in the restrained grief of the mothers and in the apparent desire of the infants to be the first Christian martyrs. The composition of this scene, in its spiritual serenity and theological signifi- cance, was the product, as Richter says, of that current conception of Christian heroism and faith which, sustained by the hope of rewards in the life hereafter, calmly and gladly faced all earthly trials. In other words, it is the reflection of the classical restraint of the Hellenistic period of Christian art when symbolism, rather than realism, governed the ren- dering of a scene. Group II. The Syro-Egypticin or the Sword Type. As the previous scene was the product of the heroic period of Christian- ity, this Syrian type, where the children are killed at the point of the sword, is in its unqualified realism the product of the coarser cult of a declining Empire. The realistic type is seen in the Rabula Gospels 3 written in 586 A. D. at Zagba (Fig. 50). The tribune, clothed in a mili- Fig. 50. Florence: Laurextiana, Gospels of Rabula. The Massacre of the Innocents. tary costume consisting of a short chiton and a pallium caught over his left arm, leans forward on the solium in the same attitude 4 as in the mosaic of Santa Maria Maggiore. The soldier executing his order holds in his hand a child whom he is about to pierce with his sword. In frantic and unrestrained grief the mother, seeking to ward off the blow with her right arm, rushes forward and grasps the soldier with her left hand. 3 Table III, no. 2. 4 For the attitude of the tribune in art see Griineisen, Sainte-Marie-Antique, p. 319. 62 EARLY CHRISTIAN ICONOGRAPHY This form became also the type used in Coptic art. In a fresco of the subterranean church of Dei'r Abou Hennys at Antinoe 5 (Fig. 51), dating from the end of the sixth or the beginning of the seventh century, is a representation of the Massacre which appears to have been derived quite directly from the scene in the Rabula Gospels. Herod, seated on Fig. 51. Antinoe: fresco. The Massacre of the Innocents. a raised dais, is flanked at either side, as in the Gospels, by two soldiers who support a large shield behind the tribunal seat. One of the soldiers has the head of an ass. The actual execution of the children is the same as in the Rabula example. The scene of the Antinoe fresco appears in es- sentially the same form in a fresco at Bawit, 6 and the sword type is also used on a Coptic bronze censer, 7 on a Coptic pyxis in the Louvre 8 and on a lost panel of the Maximianus chair which is known to us from Trom- belli’s description. 9 These six examples, the miniature, the two frescoes, the two ivories, and the censer, include all the representations that I know of the Massacre of the Innocents occurring in the East prior to the ninth century. The Byzantine examples 10 from the ninth century on do not differ essentially from this Syro-Egyptian sword type. Group III. The Provengal or Smashing Type. The scene on the Berlin fragment 11 (Fig. 52), which is the earliest example of the group of ivories which we ascribe to Provence, depicts a soldier, at the command of Herod, about to smash to the ground a child whom he swings above his head by the leg; another child already smashed 5 Table III, no. 3. 0 Cledat, Comptes Rendus, 1904, p. 22; Table III, no. 4. 7 Table III, no. 7. 8 Table III, no. 5. 9 Garr., VI, p. 21; Trombelli, De cultu SS. diss., IX, p. 247. This panel was in the Baruffaldi Collection. 10 Ms. gr. 510 in the Bibliotheque Nationale at Paris (Fleury, L’Hvangile, I, pi. XXIX/2. 11 Table III, no. 9. EARLY CHRISTIAN ICONOGRAPHY 63 to death lies before the suppedaneum, on which Herod rests his feet, and behind the soldier stands a grief stricken mother with her hair hanging loose about her shoulders and with her arms raised above her head. Herod here is bearded, and seated in a chair with a back, in contrast to the Ori- Fig. 52. Berlin: Museum, ivory. The Massacre of the Innocents. ental bench ; no soldiers flank the tribunal seat. It is in the actual exe- cution, however, that the scene differs so radically from previous repre- sentations ; not satisfied with the massacre by the sword, the soldier is about to hurl the child against the ground. Cruel and unrestrained as are the Syrian and Coptic examples when compared to the scene on the arch Fig. 53. Milan : Cathedral, ivory book covers. The Massacre of the Innocents. of Santa Maria Maggiore, these Eastern representations become mild when contrasted with the barbarism of the smashing type of the Massacre. A somewhat later rendering of the same type occurs on the Milan book covers 12 (Fig. 53) where the tribune, as in the Eastern examples, 12 Table III, no. 10. 64 EARLY CHRISTIAN ICONOGRAPHY is flanked on either side by soldiers bearing circular shields, and his short tunic and chlamys caught over the right shoulder resemble the costume of the official in the Rabula Gospels. Like the rendering on the Berlin fragment, the tribune sits in a chair and the soldier swinging a child in the air is nearly identical on the two ivories. On the Milan covers, how- ever, the scene is made more pretentious by increasing the number of the soldiers to five. Far from being peculiar to these two early Christian ivories, the smashing type may be traced by successive steps in Gaul into Carolingian art. It appears 13 in a sixth century miniature of the Codex Purpureus (Cim. 2) at Munich 14 (Fig. 54). About 800 we find a rather crude copy of the scene of the Berlin ivory on the Carolingian cover (see Fig. 169) in the Bodleian Library at Oxford, 15 the iconography of which was largely based on the diptych of which the Berlin fragment was a part. Finally, 13 Table III, no. 11. 14 Griineisen, Saint e-Marie- Antique, p. 350; Beissel, Gcsch. der Evangelienbiicher, p. 84. 15 Table III, no. 12. Goldschmidt’s recent monograph on Early Mediaeval ivories ( Elfenbeinskulpturen aus der Zeit dcr karol. und sticks. Kaiser, Berlin, 1916, PI. XXVII, 67b) records another plaque (fragment) in the Staatsbibliothek in Munich which is, like Fig. 54. Munich: Cod. Purp. (Cim. 2). Massacre of the Innocents. EARLY CHRISTIAN ICONOGRAPHY 65 toward the end of the ninth century, or in the tenth century, the same type of the Massacre is figured on two Carolingian ivories, one in the Yictoria and Albert Museum 16 and the other in the Bibliotheque Nationale (Suppl. Latin, no. 642, 9393) at Paris 17 (Pig. 55). As a rule the grounds for connecting certain iconographical types with definite geographical centres are slight and unsatisfactory. In regard to the smashing type of the Massacre of the Innocents, however, we have Fig. 55. Paris: Bibl. Nat., ivory book cover. Massacre of the Innocents. not only a long line of examples ranging in certain sequence from the early Christian period to the Carolingian but we also have a still earlier example, or a prototype, indigenous to Provence. When the ancient crypt of St. Maximin near Marseilles was unearthed there were discovered four sarcophagi of the first part of the fifth century 18 and the fragments of a much earlier one. These sarcophagi were held by Abbe Laillon 10 and Le Blant to be those to which were transferred the remains of the Magdalene, Sidion, Marcella, and St. Maximin, who, ac- cording to tradition, 20 came to Provence after the Saviour’s death to preach the Bodleian cover, an adaptation of the early Christian five-part diptych of which the Berlin fragment was a portion. It repeats the scenes and iconography of the Berlin fragment; reading from top to bottom the scenes are the Massacre of the Innocents, the Baptism, and the Miracle of Cana. 16 Table III, no. 13. 17 Table III, no. 14. 18 Le Blant, Les sarcophagcs chrctiens de la Gaule, p. 147. If *Faillon, Monuments inedits de I’apostolat de Saint e Marie-Madeleine en Provence, I, PP- 339-341- 20 An ancient legend, going back to uncertain sources and frequently repeated in verse and prose, recounts that the Jews, after the Passion of the Saviour, abandoned on a rudderless ship Mary Magdalene, Martha her sister, Lazarus their brother, Sidion the man born blind, Marcella the handmaid, and Maximin one of the seventy-two disciples. The ship, taking its course without human guidance, grounded on the shores of Provence, 66 EARLY CHRISTIAN ICONOGRAPHY the Gospel throughout this region. On the cover of one 21 (Fig. 56), the sarcophagus which Faillon 22 calls the “Sarcophagus of the Holy Inno- cents,’’ occurs the smashing type of the Massacre of the Innocents. The scene on this sarcophagus, as is to be expected at so early a date, is much simpler than any of the examples I have described. The tribune, beardless Fig. 56. St. Maximin: sarcophagus. The Massacre of the Innocents. and clothed in tunic and chlamys, is seated upon a sella curulis instead of a chair; no soldiers flank the tribunal seat; and at his command one soldier swings a child by the leg while another holds in his arms a child that he has just snatched from the mother who stands with hands clasped and her hair hanging loose about her shoulders. The sarcophagus of the crypt of St. Maximin may be dated in the early part of the fifth century, and is a local product, as is shown by the heads on the corners of the cover, which are identical with the heads on the sarcophagi of Arles 23 where during the fourth, fifth, and sixth centuries there was an active school of sarcophagi carvers who spread their works through the whole region. 24 We find, then, that the earliest example of the peculiar scene which connects into a group the early Christian and Carolingian monuments cited above, occurs on a sarcophagus made in Provence. The provenience of the smashing type of the Massacre may be further defined by the close connection of the church of St. Maximin, where our sarcophagus was found, with the Abbey of St. Victor at Marseilles, where the little band landed, and began to preach the Gospel and to found churches throughout the region. Finally, four of their number, the Magdalene, Marcella, Sidion, and St. Maximin, were buried in the little church of St. Maximin which is situated about thirty miles from Marseilles. 21 Le Blant, op. cit., pi. LVI/i. 22 Faillon, op. cit., p. 734. 23 Le Blant, Lcs sarcophagcs chretiens do la ville d’ Arles, pis. XIX, XX. 24 Le Blant, Sarcophagcs chretiens dc la Gaulc, p. v. EARLY CHRISTIAN ICONOGRAPHY 6 7 of which it was a dependency. The Holy Innocents were worshipped at Marseilles long before a feast day for them was set aside by the Gallican church. Although the feast must have been instituted during the fifth century, it is not to be found in the Latin calendars and liturgical books before the sixth century. 25 Nevertheless, from the early part of the fifth century their worship existed at Marseilles, where their relics, brought from the Orient by St. Cassianus, had been placed by the monks of St. Yictor in the celebrated crypt of their abbey. 26 Although Cassianus re- tained in his church at Marseilles the greater part of the relics which he had brought with him from Syria, he distributed portions of them through- out Provence 27 and there thus grew up among the ancient churches of the region the custom of burying along with the remains of their Saints some relic of the Holy Innocents. 28 St. Cassianus, returning from the Orient, founded the church of Saint Victor shortly after the year 414, a date to which our sarcophagus, from its style, might easily be ascribed. Furthermore, the little church at St. Maximin was a dependency of the famous church at Marseilles. All this suggests that the sarcophagus owes the peculiar scene of the Massacre to the fact that it was to be used to contain a portion of the relics of the first martyrs, presented to the church of St. Maximin by Cassianus, which were buried along with the traditional remains of the first apostles of Provence when their bodies were transferred to the new sarcophagi in the early part of the fifth century. The influence of the scene of the Massacre in the crypt of St. Maximin upon the subsequent art of Provence and later on Carolingian art is ex- plained by the renown which this crypt, where the bones of the Blessed Magdalene were supposed to rest, enjoyed during the Middle Ages. The discussion of the preceding scenes has already indicated that the Berlin frag- ment and the Milan book covers were themselves executed in Provence. In the case of the Bodleian took cover, the Munich miniature and the two ninth or tenth century ivories, which as Carolingian work are to be assigned to northern rather than to southern France, all apparently derive their rendering of the Massacre of the Innocents from the “Sarcophagus of the Holy Innocents” at St. Maximin. 2 " Duchesne, Origines du culte chretien, p. 257. • Le Blant, op. cit., p. 156. 27 Faillon, op. cit., p. 741. 28 Faillon. op. cit.. p. 738. 68 EARLY CHRISTIAN ICONOGRAPHY Several mediaeval manuscripts 29 of the tenth to the fifteenth century which are copies of an earlier Life of Mary Magdalene give evidence of traditional reverence for the crypt. In a Life of Mary Magdalene in Magdalen College at Oxford, purporting to be a copy of her life written in the ninth century by Rabanus Maurus, is a passage which reads: 30 “This place has since [that is, since the deposition there of the remains of Maxi- min and St. Mary Magdalene] become so sacred that no king or prince, or any one else, however endowed with power or wealth, can enter into the church here in order to ask for a blessing, except he shall have first put down his arms and set aside all angry passions, so that, at length, he may enter with all humility and devotion.’’ It is related that in 935 Count William and his son Odo made a pilgrimage there 31 and that in 1254 St. Louis and the Sire de Joinville visited the famous crypt. 32 It was in Provence, then, where the relics of the Holy Innocents were first worshipped and where their remains, jealously guarded in the various churches, were reverenced through the whole region, that this striking manner of representing the Massacre was created. It was probably in the church of St. Victor at Marseilles, to which the relics of the Innocents were brought by Cassianus in 414 and whence the cult in Provence emanated, that the type originated. Linally, it is in the famous crypt of St. Maximin, a dependency of St. Victor, that we find the earliest example of the smashing type of the scene, and so far as the evidence shows, it was from this centre that the curious bit of iconography spread, at the hands of devout pilgrims, through Gaul, and thus found its way into the Munich manuscript and the Carolingian ivories. VIII CHRIST AMONG THE DOCTORS The first appearance of Christ in public, when He confounded the doctors in the temple, was not a popular scene during the early Christian period and until Byzantine times no type, in airy appreciable sense, ap- peared. The Byzantine type, when developed, followed the canonical account in Luke which relates that Christ as a child of twelve years, having 29 These manuscripts are discussed by Taylor, The Coming of the Saints, pp. 93, 123. 30 Taylor, op. cit., p. 121. 31 Faillon, op. cit., I, p. 805. 32 The Memories of the Lord of Joinville, trans. by Ethel Wedgwood, 1906, p. 341. EARLY CHRISTIAN ICONOGRAPHY 69 lost His way, was found by His parents disputing with the doctors in the temple. As a type it always represents the Christ Child seated upon a throne in the temple where the Virgin, usually followed by Joseph, finds Him among the doctors. Besides this formal and developed Byzantine type there also exists a unique apocryphal rendering of the scene. On the ivory book covers of Milan 1 the type is derived from the Gospel of Pseudo-Matthew and depicts an earlier episode when Christ, as a very young child, having been sent to learn His letters of a certain master Levi, confounded the master by His learning. The Byzantine Type. While the scene of Christ among the Doctors is absent from monu- ments of assured Eastern origin during the first eight centuries, its pres- ence in the ninth century on purely Byzantine works leads one to suppose that it was constituted, like other Byzantine types, about the sixth century in the East. Sporadic earlier examples of the scene occur in the West, as on a sarcophagus of Perugia, 2 ascribed by De Rossi to the fourth century, and another in Sant' Ambrogio at Milan. 3 On the Perugian sar- cophagus Christ is seated on a raised throne and holds in His left hand a volume, while He extends His right hand in a gesture of speech; between the columns to the right of Jesus is the Virgin, followed by Joseph, and in the seven other niches are eight doctors. The regular Byzantine form with Christ seated on a raised throne with three doctors on either side, and with the Virgin and Joseph entering the temple from left or right, occurs in the ninth century in the Homilies of Gregory of Nazianzus 4 and in the eleventh century in the Greek manu- script 74 of the Bibliotheque Rationale at Paris 5 (Fig. 57). After the ninth century when this rendering entered the Ottoman and Italo-Byzantine art of the West, certain minor changes took place in the composition. In the tenth century in the Codex Egberti 6 there are only four doctors de- picted : in the eleventh century the frescoes of S. Angelo in Formis 7 1 Garr., op. cit., VI, pi. 454. 2 Fleurj-, La Sainte Viergc, I. p. 183: De Rossi, Bull, di arch, crist., 1871, p. 127, pi. VIII. 3 Fleur} - , L’Evangile, I, p. 86: Allegranza, Sacri monument i di Milano, t. X. 4 Fleury, L’ Evan g He, pi. XXXI/i. ’ Fleury, op. cit.. pi. XXXI/2. '’•Kraus. Die Miniaturen des Codex Egberti, pi. XVII. 7 Kraus. Jh. Prcuss. Kunsts., XIV, p. 1, pi. TII. EARLY CHRISTIAN ICONOGRAPHY 70 reduced the doctors to two, and the Virgin and Joseph, instead of entering from the side, look down on opposite sides from behind the throne where Christ is seated; in the twelfth century on the doors of Benevento 8 Joseph is omitted from the composition. The Provengal Type. The Milan ivory (Fig. 58) depicts Jesus, small of stature and raised above the other figures on a high seat, disputing with a doctor who stands before Him; behind the master are two pupils carrying books and listening to the words of Jesus. This scene represents one of the early incidents related in the Gospel of Pseudo-Matthew, when Jesus was sent to learn Fig. 57. Paris : Bibl. Nat., Ms. gr. 74. Christ among the Doctors. His letters from a Jewish doctor. While the only canonical account of Christ among the doctors is the story of the twelve-year-old Child in Luke, the Gospel of Pseudo-Matthew gives the event a different and more amplified treatment, telling at some length of three different occa- sions when Jesus, before He was twelve, astonished the doctors with His learning. First, when He was scarcely five, having dumbfounded a certain Zachyas with His wisdom, He was confided by His parents, at the advice of this doctor, to a certain master Levi, of whom He might learn His letters; Jesus, under Levi, on refusing to name the first letter was struck by the master, whereupon He stupefied him by His knowledge of the letters and speaking with a tone of command to the children of Israel who were present, He silenced them all by the wonder of His words. 9 “There came a second time,” the account goes on to relate, that Joseph and Mary were sought by the people to send Jesus to learn His letters at school. 10 This time, when Jesus was punished by the master for His refusal to answer, the master died. Then, for a third time, the Jews demanded that He 8 Venturi, Storia, III, fig. 643. 9 Michel, op. cit., ch. XXX, pp. 133-139. 10 Michel, op. cit., p. 151. EARLY CHRISTIAN ICONOGRAPHY 7i be taken to another master and with fear His parents complied. 11 “Now when Jesus had entered the school, guided by the Holy Spirit, He took the book from the hands of the master, who taught the law, and in the pres- ence of all the people who saw and heard Him, He began to read, not that which was written in the book, but He spoke in the spirit of the Living God.” Thus He taught them with so much force and grandeur that when Fig. 58. Milan: Cathedral, ivory book covers. Christ among the Doctors. Mary and Joseph came for Him the master said unto them, “You have given me not a disciple, but a master.” It is obviously one of these oc- casions that the artist of the Milan ivory has sought to depict, for the small stature of Jesus, the gestures of Jesus and the master which show they are in a discussion, the switches against the base of Jesus's seat, and the two smaller figures carrying books, make it impossible to interpret the scene as other than young Jesus before the Jewish master. 12 IN THE BAPTISM 1 he representations of the Baptism on the monuments of the early Christian period do not group themselves into obvious local divisions. At 11 Michel, op. cif., p. 153. - A curious ivory in the Eritish Museum (Dalton, Catalogue of the Early Christian Antiquities in the British Museum, no. 293, p. 51), whose spurious appearance is marked, shows the young Jesus of small stature standing, with a book in His hand, on the drum of a column from which He is expounding the law to the master at His right, who raises his hand in astonishment, and to a pupil at the left, who stands with a book in his hand looking at the preaching child. Besides the unique composition of the scenes on this EARLY CHRISTIAN ICONOGRAPHY the same time that they show a marked uniformity of type on all monu- ments, they also present unique features of composition in nearly every specific example. This general similarity and specific individuality of the various scenes is due in part to the fact that the Baptism, which was usually represented in its symbolical significance bv scenes from both Testa- ments , 1 was uncommon during the first few centuries of Christianity. Like all the Hellenistic scenes, it was essentially symbolic but unlike some of them its significance was universally the same. Therefore in the East, where so many of the New Testament scenes were treated more in the light of historical events than as symbols, its representation tended to preserve the traditional composition. None the less, the broad division between the Eastern and W estern scenes is fairly manifest. Such features as the hand of God, with or without rays emanating from it, the per- sonification of the Jordan, ministering angels, nimbi on the Saviour and John the Baptist, and the bearded Christ do not occur in the Western representations. In the Western scenes Christ is never immersed in the Jordan above His knees, while in the East He is usually represented in the waters up to His waist. The types, however, do not stand out clearly and are not always consistent. This Western, or Hellenistic, type is charac- terized by such details as the childlike figure of Christ immersed only to His knees and by the manner in which the waters of the Jordan appear to fall like a cataract. The other features of the type are principally negative ones. The Provengal group, within the Hellenistic type, is ex- ceptional, for it is both obvious and distinct from the other forms of the scene. The distinguishing feature, unique to the group, is the pedum, or shepherd’s crook, which John the Baptist carries. The Alexandrian-Coptic type is characterized by the personification of the Jordan and the minister- ing angels, while the Palestinian-Coptic is essentially the same save for the absence of the fleeing Jordan. The two Syrian examples of the scene are similar enough to form a group but they can not be said to present an apparent type. ivory and the poor character of the work, the representation of a bearded angel in the Baptism convinces me that at best the work could not have been done in the early Christian period. 1 Fleury, L’Evangile I, p. 102 : the Baptism was often symbolized by the Passage of the Red Sea, by Moses Striking Water from the Rock, by Christ and the Samaritan Woman at the Well, etc. EARLY CHRISTIAN ICONOGRAPHY 73 Group I. The Hellenistic Type. Since the Baptism was so frequently represented by Old Testament scenes, there remain only a few catacomb examples of the actual Baptism. 2 These in themselves were not sufficient to have played any important part in the subsequent development of the scene on Western sarcophagi. One feature of these catacomb representations is interesting (Fig. 59). The dove in the frescoes prior to the fourth century, instead of descending directly upon the head of Jesus while He stands in the Jordan, appears to Fig. 59. Rome: Cemetery of Lucina. The Baptism. be flying toward the scene. 3 This manner of depicting the scene corre- sponds more faithfully with the Gospel accounts than do the examples after the third century. None of the Evangelists say that the Holy Spirit descended upon Christ while He was in the water, and Matthew (iii, 16), Mark (i, 10), and Luke (iii. 21) write that when Jesus had been baptized He went straightway out of the water and then the Spirit of God descended upon Him like a dove. In the fourth century, however, the sculptors of the sarcophagi brought the two incidents, of the Baptism and the descent of the dove, into one scene. On the sarcophagi Christ is always beardless, small like a child, and never immersed in the waters of the Jordan more than to His knees. John the Baptist is dressed in exomis or perizoma in the catacomb frescoes ; in - Table I\ , nos. 1-5. The following are some of the best discussions of the Baptism : Strzj gowski, Iconographie dcr Taufe Christi: Heisenberg, Grabeskirche und Apostcl- kirche, pp. 236-239; C. Jacoby, Ein bisher unbeachteter apokrypher Bericht fiber die Taufe Jesu C. F. Rogers, Baptism and Christian Iconography ; and Leclercq in Cabrals Dictionnaire d’archeologie chretienne ct dc liturgie, s. v. “Bapteme.” In the Lucina fresco it is to be noted that Christ is stepping out of the water. 74 EARLY CHRISTIAN ICONOGRAPHY the sarcophagi he generally wears a pallium instead of the shepherd’s gar- ment. The method of representing the Jordan as an apparent waterfall descending like a cataract behind the figure of Christ is characteristic of the Western scenes 4 (Fig. 60). Although this feature does not occur on all the sarcophagus representations of the Baptism, it is peculiar to the greater part of them and is in marked contrast to the Eastern manner of Fig. 6o. Rome: sarcophagus. The Baptism. depicting the Jordan as a river in which the Saviour stands waist deep. The Hellenistic scenes are all very simple and never include such details as the assisting angels and the personification of the fleeing Jordan seen on Eastern monuments. The main features of the Hellenistic manner of representing the Bap- tism are the childlike Christ and the frequent piling up of the waters of Jordan like a waterfall. The artistic tradition that Christ should be de- picted as a child in contrast to the figure of the Baptist was universal inasmuch as even the Eastern scenes not infrequently show the Saviour as very small of stature. The dates of Christ’s birth and crucifixion, implied by the Gospels, make Him more than twenty years at the time of the Bap- tism. This and the statement of St. Luke that there was only six months difference between the ages of John and Jesus are inconsistent with the 4 Table IV, nos. 7-13. EARLY CHRISTIAN ICONOGRAPHY /o apparent difference in their ages on the monuments. A possible explana- tion, suggested by Leclercq , 5 is that the artists were perhaps following a more authentic tradition than had the Evangelists in the Gospels. A more plausible suggestion, however, is Le Blant’s explanation of the small stature of the Saviour by the Christian concept of the Baptism as a birth into life immaterial — whence the name “infans” given to the baptized . 6 A more distinct and characteristic detail of this Hellenistic type is the representation of the Jordan as a waterfall. This feature, while it appears only on Western monuments and principally on those of Gaul, offers, perhaps, some evidence of the general Hellenistic character of the type and even of its connection with Alexandria. In the “Chronicon Paschale” Fig. 6i. Soissoxs: sarcophagus. The Baptism. of Alexandria' we read that when Jesus was baptized in the Jordan, “the Jordan turned back.” “The Lord said unto John, ‘say to the Jordan, stop, the Lord has come among us,’ and so the waters stopped.” This Alexan- drian description of the event seems to be exactly the phenomenon that the sculptors were seeking to represent on the sarcophagi. By what ap- 3 Leclercq, op. cit., col. 356. Le Blant, Etude sur les sarcophages chretiens de la ville d’ Arles, 1878, p. 27. ‘Leclercq, op. cit., col. 348; Chronicon Paschale, Bonner Corp., p. 420 sq ; the Ency- clopaedia Britannica, s. v. “Chronicum” says that the manuscript was compiled between 610-614 A. D. ;6 EARLY CHRISTIAN ICONOGRAPHY pears to be a waterfall they sought to show the waters of the Jordan piled back upon their source in order to leave a shallow space in the river bed where the first sacrament might be performed. On one sarcophagus the river instead of resembling a cataract or waterfall, as it usually does, rises behind Jesus like a huge wave 8 (Fig. 61). There could have been no direct relation between the text and the representations of the Baptism on the sarcophagi inasmuch as the “Chronicon” was written as late as the beginning of the seventh century. It is supposed, however, to have been compiled from earlier sources and may, therefore, preserve, in the case of the Jordan turned back upon its source, an older Hellenistic tradition which was also current among the sarcophagi carvers. Group II. The Provcngal Type. The Provenqal type, which is listed as a sub-group of the Hellenistic type, is distinct and uniform. The monuments which show this type are three ivories whose Provenqal provenience has become evident in the dis- cussion of the preceding scenes: the Berlin fragment, the Milan book cover (Fig. 62), and the Werden casket in the Victoria and Albert Museum. Fig. 62. Milan : Cathedral, ivory book covers. The Baptism. A fourth is the Carolingian book cover in the Bodleian Library at Oxford which appears to have been in part copied from the diptych of which the Berlin fragment is a portion (see also a Carolingian ivory at Munich, note 15, p. 64). All these ivories show Jesus with long hair, beard- less and small, and, with the exception of the Berlin fragment (Fig. 163), they show Him immersed in the water to His knees. 9 The fact 8 Table IV, no. 8. 9 Table IV, nos. 19-22. EARLY CHRISTIAN ICONOGRAPHY 77 that Jesus on the Berlin ivory wears a plain nimbus and on the Werden cas- ket wears a cruciform nimbus has little or nothing to do with the iconog- raphy, as there is every reason to believe that the nimbi were added long after the ivories had been carved. This assumption is borne out by the manner in which the waves of the Jordan on the Berlin fragment cut across the nimbus, which is only a double incised line on the background. John the Baptist on these four ivories is bearded, wears the exomis instead of the pallium, and carries in his hand a pedum, or shepherd’s crook, which is the most distinguishing feature of the group. In the same manner all the monuments represent the Jordan like a cataract falling behind the figure of Jesus on whose head descends the Spirit of God in the form of a dove. The Werden casket ( Fig. 63), which in so many of its scenes duplicates the iconography on the Milan covers and is even considered by some Fig. 63. S. Kensington : Victoria and Al- bert Museum, ivory casket. The Baptism. authorities to have been done by the same hand, adds a curious detail in the form of a personification of the Jordan seated at the left of the scene. With his arms full of water plants, the river god sits upon the top of a little mountain and from an overturned urn at his side the waters of the Jordan flow down to Christ. The pedum, which is the feature distinguishing the group from all others, does not appear in any of the representations on the sarcophagi. In fact, the only other example of this shepherd’s crook that John carries occurs in the fifth century mosaic in the Arian baptistery (Santa Maria in Cosmedin) at Ravenna. 10 All the other details of the composition of the Provenqal type occur on Gallic sarcophagi and in many cases are peculiar to the West. The beardless and childlike Christ, which was the regular 10 Table IV, no. 30. EARLY CHRISTIAN ICONOGRAPHY 78 Hellenistic type in the Western representations, appears on two sarcophagi of Arles, one of Soissons, and on another at Mas d’Aire. 11 The Jordan depicted like a cataract occurs on all four of these monuments. In other words, the characteristic elements of the scene on the ivories, with the exception of the pedum, were customary parts of the composition on the sarcophagi of Gaul. While the analogies with the Provencal monuments only show that the connection of the type with Provence is possible, the important fact that is established by the scene of the Baptism on the ivories is the close relation of the group. In the case of the Bodleian cover (see Fig. 169), a Carolingian work of about 800 that was carved somewhere in the north of France, it is interesting to be able to show so intimate a connection on the part of a monument of this period with the early art of Provence. East. The features which appear to be peculiar to the Oriental representa- tions of the Baptism are the bearded Christ, the fleeing Jordan of the Egyptian variant, and the assisting angels. Minor characteristics which Fig. 64. Florence: Laurentiana, Gospels of Rabula. The Baptism (The Hand of God, OMITTED IN THE CUT, APPEARS IN THE ORIGINAL). are not so distinctive are the immersion of Christ to His waist, the nimbi on Jesus, John, and the angels, and the hand of God with the rays of light issuing from it. 11 Table IV, nos. 10, 13, 8, 9. EARLY CHRISTIAN ICONOGRAPHY 79 Group III. The Syrian Group. The only two examples, miniatures in the Rabula (Fig. 64) and the Etschmiadzin Gospels (Fig. 65), which form this group, while they Fig. 65. Etschmiadzin: Gospels. The Baptism. possess many features in common, are not sufficiently alike and distinct enough from other representations to form a type. 12 On both manuscripts Christ, who is immersed to the waist, is small, has dark hair, and though beardless in the Etschmiadzin Gospels is bearded in the Rabula Gospels. Both scenes show John the Baptist bearded and wearing a pallium and depict the hand of God above the dove to complete the Trinity. Group IV. The Alexandrian-C optic Type. The distinguishing features of this group are the personification of the Jordan as a fleeing river god and the appearance of one or more angels as attendants at the Baptism. The type, if such we may call it, occurs on the Maximianus chair 13 (Fig. 66), an ivory in the British Museum 14 (Fig. 67), and in one of the Bawit frescoes in Egypt 15 (Fig. 68). The fact that the iconography of the Baptism, as of so many other scenes, on the Maximianus chair is also represented on the Egyptian ivory in the British Museum and in Egyptian frescoes, tends to confirm not only the Alexan- drian origin of the Maximianus chair but also the existence of an Alex- andrian-Coptic type of the Baptism. 12 Table IV, nos. 23, 24. 33 Table IV, no. 25. 14 Table IV, no. 26. ir ' Table IV, no. 27. 8o EARLY CHRISTIAN ICONOGRAPHY To this Alexandrian-Coptic type must be added the scene on the gold medallion from Cyprus in Strzygowski s collection. 1 *’ While its general workmanship, as well as the consistent use of the nimbus on Christ, the Baptist, and the angels, recalls the Palestinian-Coptic type of the scene, the Fig. 66. Milan : Museo Archeologico, IVORY PANEL OF CHAIR OF MaXIMIANUS. The Baptism. Fig. 67. British Museum : ivory panel. The Baptism. essential features of the medallion classify the scene with the Egyptian ivory in the British Museum. In both scenes, besides the dove descending on the head of Christ, occurs the uncommon feature of the hand of God with rays of light emanating from it. On the Cyprus medallion, although three personifications appear, one of them represents the fleeing Jordan 10 Table IV, no. 32, and Errata. This medallion was first classified from a description in which the hand of God and personification of the fleeing Jordan were not noted. Upon verification with a photograph, it was found that these features did appear and that the medallion was to be classified in the Alexandrian-Coptic instead of in the Palestinian-Coptic group. EARLY CHRISTIAN ICONOGRAPHY 8 1 which is the distinguishing feature of the Alexandrian-Coptic type of the Baptism. In Egypt the Hellenistic type was gradually altered, as it became indigenous to Egypt, by the growing Oriental influences both traditional to the land of the Pharaohs and introduced into the country from Syria and Palestine. This tendency of the type to become more Oriental in character appears in comparing the scene on the Maximianus chair with the Bawit fresco; on the chair the figure of Christ is small and beardless after the Hellenistic representations; but in the fresco the bearded Christ recalls the figure of the Rabula Gospels. Although it may be presumed that the personification of the fleeing Fig. 68. Bawit: fresco. The Baptism. Jordan as it is represented in the scene of the Baptism on the Maximianus chair originated in Alexandria, there is no definite connection of the type, irrespective of the monuments on which it occurs, with that city. While the “Chronicon Paschale,’’ compiled in the early seventh century in Alex- andria, personifies the Jordan and describes the river as having fled at the words of the Almighty, 17 a similar story is related by several Syrian writers and Theophany sermons. 18 Besides the personification of the river, the regular appearance of one or more angels as attendants at the Baptism is another marked feature of the type. 1 he presence of these angels Leclercq considers to be a characteristic detail of Oriental, and particularly of Alexandrian ivories. 19 '“Jacoby, op. cit., p. 17. 1!> Leclercq, op. cit., col. 368. !S Jacoby, op cit., pp. 42, 43, 44, 45. 82 EARLY CHRISTIAN ICONOGRAPHY Group V. A Ravenna Group. The two scenes of the Baptism on the fifth century mosaics of San Giovanni in Fonte 20 and the Arian Baptistery ( Santa Maria in Cosmedin) 21 (Fig. 69) at Ravenna show a strange mixture of Eastern and Western fea- tures. The restorations of the mosaics of San Giovanni have greatly altered the iconography. It is probable, however, that the scene of the Baptism in Santa Maria in Cosmedin is a copy of the earlier scene in San Giovanni in Fonte. Such being the case, we may with Strzygowski ascribe to the re- storation of the earlier mosaic the patera in the hand of the Baptist, the sub- stitution of the cross for the pedum, and the placing of the nimbus on John’s head. These two scenes do not classify very satisfactorily. In the first place the fact that Jesus is immersed in the Jordan to His waist is more char- acteristic of Eastern representations and the presence of the personification of the Jordan in both scenes points to Alexandria. The figure of Jordan, however, is here used only as a symbol of locality, and is not represented as fleeing. Group VI. The P ale stinian-C optic Type. The characteristics of this group, with the exception that the personifi- cation of the Jordan does not figure in the compositions, are not very different from the Alexandrian-Coptic type. Aside from the absence of the river god the only really distinguishing feature of the group is the consistent use of the nimbus on all the figures. In the other groups there have been a very few scattered examples where the nimbus was used, but Fig. 69. Ravenna: Arian Bap- tistery. The Baptism. 20 Table IV, 110. 29. 21 Table IV, no. 30. EARLY CHRISTIAN ICONOGRAPHY 83 in this type all the figures, including the angels, receive the nimbus. The Palestinian monuments on which this type occurs are a Monza phial 22 (Fig. 70), an ampulla in the Ivaiser-Friedrich Museum. 23 and an eighth Fig. 70. Monza: ampulla. The Baptism. century (?) panel in the Sancta Sanctorum. 24 In the sixth century the type appears in Egypt on a panel of the Golenisheft Collection 25 (Fig. 71), on a medallion from Akmim, 26 and there is possibly an example on an ivory in the British Museum 27 of the seventh century ( ?, see p. 71, note 12). In the eighth century the type occurs in Italy in the frescoes of St. Pontianus at Rome 28 and San Gennaro at Naples. 29 Save for the addition of the at- tendant angel and the use of the nimbi, the scene on the Golenisheft" panel recalls the representation of the Rabula Gospels in that it is the only ex- ample in the group to show Jesus with a beard. The ivory in the British Museum presents a suspiciously curious scene in which the attendant angel is heavily bearded. 22 Table IV, no. 31. 28 This example, which was listed after the table in the Appendix had been set up, is of Palestinian-Coptic type (see Amtlichc Berichte aus den Koniglichen Kunsts., XXXV, no. 2, Nov. 1913, fig. 21, p. 40). Christ is small, beardless, with long hair, wears a nimbus, and is immersed to His waist in the Jordan. John the Baptist is bearded and nimbed, and wears a pallium A dove descends upon the head of Christ, while two nimbed angels at the left of the composition complete the scene. In the same collection {op. cit., fig. 22) is another scene of the Baptism on a piece of beaten bronze from Pergamon. Although much cruder and simpler than the rendering on the ampulla, this is also Palestinian-Coptic. Save for the absence of all nimbi and the presence of only one angel, it is essentially the same as the scene on the ampulla. 24 Table IV, no. 36. 23 Table IV, no. 33 - 28 Table IV, no. 34. 27 Table IV, no. 35. 28 Table IV, no. 38. 29 Table IV, no. 39. 8 4 EARLY CHRISTIAN ICONOGRAPHY Group VII. The Byzantine Group. Unlike so many other scenes in Byzantine iconography, the Baptism does not present a type that in its origin is clear and distinct from the Fig. 7 1. Petrograd: Golenisheff Collec- tion, panel. The Baptism. other types. A carved column-drum of the sixth century in the Museum of Constantinople may present what was the proto-Byzantine type. 30 It shows a bearded ( ?) Christ, the personification of the Jordan, and two angels assisting at the sacrament. Save for the bearded Christ, which is doubtful, it is the same as the Alexandrian-Coptic type. It is not unlikely that the Byzantine method of representing the Baptism may have had some such origin, for Constantinople in the sixth century was a clearing house for all the artistic ideas of the Orient. The personification of the Jordan as well as the number of the angels does not seem to have con- formed to any rule. 31 In the Menologium of Basil II (976-1025) there is no personification and only two angels, 32 while in the eleventh century Ms. 74 at Paris (Bibl. Nat.) there are five angels and a small figure of the Jordan. 33 30 Table IV, no. 28. 31 Dalton, Byzantine Art and Archaeology, pp. 654-655, says that two angels were usually represented until the eleventh century and that there were three or more there- after. 32 Strzygowski, Iconographie dcr Taufe Christi, II/n. 33 Strzygowski, op. cit., III/9. EARLY CHRISTIAN ICONOGRAPHY 85 X THE MIRACLE OF CANA The various methods of representing the Miracle of Cana in Christian art divide into four general types, the Hellenistic, the Alexandrian-Coptic, the Palestinian-Coptic, and the Byzantine. The Hellenistic type is purely symbolic and the disciples are reduced to only one, while Christ, like a magician, is represented touching with His wand the jars of water which van,- in number from one to seven. The Alexandrian-Coptic is transitional between the earlier symbolical type and the more indigenous and narrative type of the East ; it retains the symbolic character of the Hellenistic type by preserving in the scene the one disciple and by continuing to represent Christ with His wand as a kind of magician; yet to give a slightly more narrative cast to the scene it introduces the figure of one servant who pours the water into the jars. The Palestinian-Coptic type introduces the Virgin into the scene and appears to fix the number of jars, in accordance with John ii, 6, as six. The Byzantine is the fully developed and historical type in which two servants pour the water into six jars, the Virgin is always present, and the additional feature of the wedding feast is added to the composition. The Hellenistic Type In its purely symbolic character the Hellenistic type of the Miracle of Cana is in no wise different from other Hellenistic scenes. While recounted only in the Gospel of John (ii, 1-10), which differs so notice- ably from the Synoptic Gospels by the marked symbolic interpretation that it puts upon the miracles, and was one of the leading factors in mould- ing the entire cycle of early Christian symbolism, the episode was con- sidered by theologians as the first authentic miracle of the Saviour. During the third and the fourth centuries it appears to have gained great theo- logical importance, and became, in combination with the miracle of the loaves and the fishes, a symbol of the Eucharistic Transubstantiation. The groups of this Hellenistic type which are here constituted are primarily based on the number of jars represented in the various scenes. While the number 1 of jars was a subject of mystical contemplation and 1 De Mely, Vases de Cana, in Mon. Piot, X, 1903, p. 156. 86 EARLY CHRISTIAN' ICONOGRAPHY intricate symbolism during the Middle Ages, there is no evidence that it played an important part in the symbolism during the early centuries of Christian art. Nevertheless it offers an easy means of classifying the variations of the Hellenistic type. Groups I-VI. In these groups the only variation in the representation is the number of jars, varying from one to seven, which Christ touches with His wand. There is only one example, a sarcophagus of Numidia (Fig. 72), on which the jars are reduced to one. 2 Two jars occur on two sarcophagi of Arles, a sarcophagus of Toulouse, and on two sarcophagi of Rome Fig. 72. Algiers: sarcophagus from Numidia. The Miracle of Cana. which differ from the others in the detail that Christ does not carry a wand. 3 The group with three jars is by far the largest; it includes two sarcophagi of Arles, one of Gerona, one of Marseilles, one of Saragossa, and eight of Rome 4 (Fig. 73). On three sarcophagi of Rome the number off jars is increased to four. 5 Five jars appear on six sarcophagi of Rome, and on one of Saragossa; as a rare exception to the rule, one of the Roman representations omits the disciple who usually accompanies the Saviour. 6 The variation with six jars occurs first in a third century 2 Table V, no. 1. 3 Table V, nos. 2-3; 4; 5, 6. 4 Table V, nos. 7, 21: 8; 10; 11 ; 13-20. 5 Table V, Group 4. 0 Table V, nos. 24-28, 29, 30. EARLY CHRISTIAN ICONOGRAPHY 87 fresco in the catacomb of SS. Pietro e Marcellino and later, in the fourth century, in another fresco in the same catacomb ; it also appears on Fig. 73. Rome : sarcophagus. The Miracle of Cana. a sarcophagus in Madrid and on two of Rome where again the disciple is omitted. Included in this group is a copper cover from Alexandria and the silver casket of Milan which has sometimes been ascribed to the same centre. 7 Group VII. Seven Jars. The number seven is peculiar to the representations on the “Roman glass" s (Fig. 74). It is true that the scenes on these pieces of glass are more symbolic in appearance, for the disciple is always omitted and the solitary Christ, as a kind of thaumaturgist, is depicted touching with His wand the seven stone jars. In addition to these glass examples the seven jars appear on the doors of Santa Sabina at Rome 9 which are probably of Eastern execution. 7 Table V, nos. 31-32, Group VI. All the writers on the silver casket in the church of San Xazzaro at Milan have seen evidence in it of late Hellenistic workmanship. For my own part, I feel that it exhibits, especially in the scene of the Adoration of the Magi (?), a freedom of composition and a disregard of tradition incompatible with early Christian workmanship. H. Graeven, Zeitschrift fiir christliche Kunst, 1899, taf. I ; A. Riegl, Die spatromische Kunstindnstric , p. 106: Dalton, Byzantine Art and Archaeol- ogy, P- 563- 8 Table V, nos. 39-44. ■' Table V, no. 38. 88 EARLY CHRISTIAN ICONOGRAPHY East The scene of the Miracle of Cana does not appear to have been as popular in the East as it was in the West. As a rule the scenes of de- Fig. 74. Rome : Vatican, bottom of glass cup. Resurrection of Lazarus and Miracle of Cana. veloped theological allegory and of Hellenistic origin prevailed in the West as the Occidental type while in the Orient they were often supplanted by scenes more personally connected with the lives of the Saviour and the Virgin. This tendency of the East to give actual color to all the events of the New Testament disguised the symbolism of the older Hellenistic scenes. For example, in the Miracle of Cana the realistic expansion began by the addition to the Hellenistic type of one servant pouring the water ; then the Virgin was introduced into the composition; and finally in the Byzantine type the scene was made completely historical by the introduc- tion of the wedding feast. Group VIII-A. The Alexandrian-C optic or Transitional Type zvhere one servant is added to the Hellenistic Type. The Alexandrian-Coptic type marks a transition from the Hellenistic type of symbolic character to the more historic and indigenous types of Palestine and Byzantium. If Alexandria was the centre from which the old symbolic type originally spread, it is to be expected that any subse- quent tvpes originating there would preserve much of the earlier form. EARLY CHRISTIAN ICONOGRAPHY 89 Fig. 75. Ravenna : ivory panel of chair of Maximianus. The Miracle of Cana. The introduction of the one servant, on the other hand, shows the modifi- cation of the Hellenistic type in the direction of Oriental realism. This transitional type occurs on a sarcophagus at Civita Castellana 10 and on a silver vase formerly in the Bianchini collection at Rome 11 whose marked beauty of execution and Hellenistic figures afford some evidence of the Alexandrian origin which is indicated by its iconography. The type also occurs on the ivory chair of Maximianus at Ravenna. 12 On this ivory (Fig. 75), Christ wears a nimbus and bears the regular Coptic cross in- stead of the wand. Six jars are represented, into which a servant is pouring the water, while behind him stands a disciple with his hand raised in surprise at the miracle. This scene on the Maximianus chair is repeated on a Coptic medallion in the Gans collection at Berlin, 13 from Egypt (Fig. 76). 10 Table V, no. 45. 11 Table V, no. 46. 31 Table V, no. 47. 13 Table V, no. 48. 90 EARLY CHRISTIAN ICONOGRAPHY Group VIII-B. Continuation of the Transitional Type in Italy. There is nothing in the Western versions of this type which necessitates an Egyptian origin. It could as well have sprung up in a dozen places where Oriental influence was beginning to transform the Hellenistic models. What it always does presuppose, however, is the presence of Fig. 76. Berlin: Antiquarium, gold me- dallion from Egypt. The Miracle of Cana. Eastern influences. Therefore, it is not at all surprising to find it in the Baptistery of Bishop Soter at Naples 14 (Fig. 77) which manifests un- questionable Oriental inspiration, 15 and in the mosaics of Sant’ Apollinare Nuovo at Ravenna. 16 The affinities of the Baptistery of Naples are all with Syria; it is interesting, therefore, to note that in its scene of the Miracle of Cana, which is combined with the Samaritan Woman at the Well, there are represented two servants who are, as we shall see, charac- teristic of the Palestinian-Coptic type. Group VIII-C. The Provencal , Carolingian, and Ottoman use of the Alexandrian-C 0 ptic Type. In Provence the Transitional or Alexandrian-Coptic type occurs in the fifth century on the ivory fragment in the Berlin Museum 17 (see Fig. 163). In this scene Christ, accompanied by a single disciple, stands between four jars into which a servant is pouring the water. Without attaching undue importance to the observation, it is to be noted (in con- 14 Table V, no. 49. 15 Diehl, Manuel, p. 117. 16 Table V, no. 50. 17 Table V, no. 51. The commentary on the miracle by Maximus of Turin (first half of fifth century, Horn, 23) shows that the servant was an important feature of his mental image of the scene (cf. Zahn, Amtl. Ber., XXXV, p. 100). EARLY CHRISTIAN ICONOGRAPHY 9i sideration of the fact that, from stylistic reasons, I have suggested that Roman workmen, emigrating from Rome to Provence, carved the Berlin ivory 18 ) that four jars are found only on two sarcophagi of Rome. 19 Fig. 77. Naples: Baptistery, mosaic. The Miracle of Cana (after Garrucci; a second SERVANT APPEARS IN THE ORIGINAL MOSAIC). About 500 the same type, with a few changes, appears on the ivory book covers in Milan cathedral 20 (Fig. 78). Christ, accompanied by eight spec- tators instead of one, is here represented touching with His wand three jars into which a servant is pouring the water. In a way this rendering is even more transitional in character than the regular type seen on the previous ivory; for the figure of Christ with a wand is in the Hellenistic Fig. 78 Milan : Cathedral, ivory book covers. The Miracle of Cana. manner seen principally on the sarcophagi of the West, the single servant is an Eastern feature added to make the scene less symbolic, and the eight disciples are the effort of the particular artist of the Milan covers to make his composition more narrative. While there is nothing in these composi- tions to connect the ivories with Provence, it is noteworthy that the num- 18 See Berlin Ivory, p. 246 sq. 1 ' < Table V, group 4. 20 Table V, no. 52. EARLY CHRISTIAN ICONOGRAPHY 92 ber of jars, three on the Milan covers and usually six on the more literal Eastern monuments, is never more than three in the scenes of the Miracle of Cana on the sarcophagi of Gaul. 21 As in the case of so many Egyptian types, the Alexandrian-Coptic type was transmitted to Carolingian art where it appears on the book cover in the Bodleian Library at Oxford 22 (See Fig. 169). This Carolingian ivory, which is in part a copy of the Berlin ivory and has many icono- graphic affiliations with Provengal types, presents a scene that is nearly identical with the rendering on the Berlin fragment save for the fact that the jars are increased to six, thereby conforming not only to the Eastern usage but also to the Carolingian scenes of the ninth and tenth centuries. In the subsequent art of the North this type appears on an Anglo-Saxon drinking cup in the British Museum 23 and in the eleventh century in the Missal of Limoges. 24 Group IX-A. Palestinian-C optic Type with the addition to the composi- tion of the Virgin and with the servants increased to two. Besides the characteristic introduction of the Virgin and the addition of another servant pouring water, the Palestinian-Coptic type definitely fixes the number of jars as six and omits the disciple. The introduction of the Virgin, while attributable to Oriental fidelity to the Biblical account, Fig. 79. Florence: Laurentiana, Gospels of Rabula. The Miracle of Cana. may also be explained by the early acceptance and development of the cult of the Virgin in Syria. The type appears in Syria among the miniatures of the Rabula Gospels 25 of 586 A. D. (Fig. 79) and then appears to have been closely imitated in a fresco of Antinoe in Egypt 26 (Fig. 80). Un- luckily these are the only two examples of the early Christian period. 21 Table V, 110s. 7, 10, 21. 22 Table V, no. 53. See also the Carolingian fragment at Munich, p. 64, note 15. 23 Table V, no. 54. 24 Table V, no. 55. 25 Table V, no. 56. 20 Table V, no. 57. EARLY CHRISTIAN ICONOGRAPHY 93 Group IX-B. Continuation of the Palestinian-C optic Type in the West. The Carolingian art of the ninth century in the West derived a large part of its iconography and many of its ornamental motifs from Syria and Egypt. Therefore, it is not at all surprising to find the Virgin added Fig. 8o. Antinoe: fresco. The Miracle of Can a. to the scene in the Gospels of Soissons 27 (Fig. 81), the Gradual of Priim, 28 a Carolingian coffer in the Sancta Sanctorum 29 at Rome, the Codex Egberti 30 of the tenth century, and in Italy on an ivory of Palermo. 31 Group X. The Byzantine Type with the introduction of the zvedding feast. This type is primarily the same as the Palestinian-Coptic save for the fact that there is added the wedding feast with the bride, groom, and guests seated around the table while the Saviour, at the suggestion of the Virgin, performs the miracle. When this type originated in the East it is hard to say. On the left hand anterior column of the ciborium of San Marco 32 this amplified type occurs in a form suspiciously developed for the date in the sixth century which is regularly given to this column. It was from the second half of the ninth century on, when the venerators of images triumphed in Byzantium, that the historical type became devel- oped and popular on the monuments of the East. In the ninth century, on the Paliotto of Milan, 33 all the elements of this type appear, although the actual feast is represented by only the gov- ernor of the feast who receives the wine from the hand of a servant. By the end of the ninth century and through the eleventh century we have several examples of it on Byzantine ivories 34 and in the Greek manuscript 27 Table \, no. 58. 2S Table V, no. 59. 26 Table V, no. 61. 30 Table V, no. 60. 31 Table V, no. 62. 32 Table V, no. 63. 33 Table V, no. 64. 34 Table V, nos. 65-68. 94 EARLY CHRISTIAN ICONOGRAPHY Fig. 8i. Paris: Bibl. Nat., Gos- pels of Soissons. The Miracle of Cana. no. 74 in the Bibliotheque Nationale at Paris 35 (Fig. 82). In the twelfth century it also appears on such Italo-Byzantine monuments as the doors of Benevento, 36 the ivory panels of Salerno, 37 and the mosaics of Mon- reale. 38 XI THE HEALING OF THE BLIND The Gospels tell of three and possibly four occasions when Christ gave sight to the blind. In the artistic renderings of the miracle of the Healing of the Blind it is not only very difficult to tell which one of the events the artists have followed, but it is also only possible to derive from these representations certain general types. The classification is in part based on the local origin of the examples, as the variations in rendering do not justify a segregation into types on the ground of differing details. The Hellenistic type shows a single blind man, small of stature as compared to Christ and frequently carrying a staff ; the Saviour either touches him on the eyes or lays a hand on his head. The Oriental-Hellenistic division is a continuation of this same type on Eastern monuments, although on some of the examples Christ raises his hand instead of touching the blind man. The Alexandrian-Coptic is again a continuation of the earlier type with the addition of the Coptic cross which Christ usually carries in miracle scenes. The Syrian type is more distinct in showing two blind 30 Table V, no. 69. 37 Table V, no. 72. 36 Table V, no. 71. 38 Table V, no. 73. EARLY CHRISTIAN ICONOGRAPHY 95 men approaching Christ instead of one. It is into this group that the scene on the Milan ivory falls. The Byzantine type is the most historical, for it represents the blind man, after Christ has put clay upon his eyes, washing in the well of Siloam. While the Gospels speak of four miracles of the kind, they give the artistic details for only two distinct scenes. Matthew writes that when Christ was departing from the house of Jairus 1 two blind men followed Him and that on leaving Jericho 2 He met two blind men seated by the side of the road; in both cases after He had touched their eyes they were healed. Fig. 82. Paris : Bibl. Nat., Ms. gr. 74. The Miracle of Cana. Mark writes that Christ on His way to Jericho, when the blind man Bar- timeus who sat begging by the side of the road asked for mercy, told him that his faith had made him whole. 3 Unlike the other miracles it is related in the Gospel of John 4 that Jesus, meeting the man born blind, rubbed clay upon his eyes and sent him to wash them in the pool at Siloam. The Hellenistic type, where Christ usually touches the eyes of the blind, who is small of stature, appears to be a symbolic rendering of the last named miracle 0 which in the Byzantine type is depicted in a more narrative manner. This miracle is the only one wherein Christ touches the eyes of a single blind man and its dependence on the account in John is shown by a sarcophagus described by Bottari 6 where two miracles of the healing of the blind are figured, the one at the left showing Christ laying His hand on the head of a blind man who sits by the wayside and another scene which depicts Him touching with His fingers the eyes of a small figure in front of Him. If the touching of the eyes represents the 1 Matthew, ix, 27-30. 2 Matthew, xx, 29-34. “Mark, x, 46-52. 4 John, ix, 1-41. Leclercq in Cabrol s Dictiomiawc d’archeologie chvctienne ct dc liturgie, s. v. “Aveug- les,” col. 3232. Pj Bottari, Sculturc e pitture, I, pi. XXXII. EARLY CHRISTIAN ICONOGRAPHY 96 healing of the man born blind, the other scene on the sarcophagus must then depict the healing of Bartimeus who sat begging at the side of the road until Jesus called to him. The symbolic nature, however, of all the Hellenistic types makes it uncertain in the case of this miracle which one of the Gospel accounts was being followed. The symbolic and didactic meaning of the episode was suggested and partly developed by the Evangelist John in his account of the miracle. As early as the second century St. Irenaeus 7 expounded the relationship between the restitution of the light of the world and the Fig. 83. Rome: sarcophagus. Christ Heal- ing the Blind. resurrection of the flesh by the introduction into it of the light of heaven, and this relationship was restated by the theologians of the succeeding centuries . 8 The Hellenistic type appears in the West as early as the beginning of the third century in a fresco of the catacomb of Domitilla where Christ is represented touching with His fingers the eyes of a blind man who kneels 7 Leclercq, op. cit., col. 3230; Irenaeus, Adi’, haeres. P. G. t. VII, col. 1163. 8 Leclercq, op. cit. EARLY CHRISTIAN ICONOGRAPHY 97 before Him. The regular rendering of the type on the Western sarcophagi 9 shows Christ touching the eyes of a blind man of small stature (Fig. 83), but in all the catacomb representations, with the exception of one in Domi- tilla, 11 and on a few sarcophagi 12 Christ lays His hand on the blind man’s head. On twelve sarcophagi 13 the blind man carries a staff, a detail more characteristic of the later scenes of Eastern origin than of the Hellen- istic type. Group II, the Oriental-Hellenistic division of the Hellenistic type, includes those monuments of Eastern origin on which the Hellenistic type occurs. While this type was uniformly maintained in the West for the first five centuries of Christian art, and for that reason might be called Western, it probably originated at the end of the second century in the East and thence spread through the whole Hellenistic world. With its presumably Oriental origin in mind it is not surprising either to find this type on Hellenistic monuments of Eastern inspiration, like the Brescia casket, 14 the doors of Santa Sabina, 15 the ciborium columns of San Marco, 16 the Alexandrian pyxis at Bologna, 17 and the gold encolpium at 8 Table VI, nos. n-47, 49. 10 Table VI, nos. 2-8. 11 Table \ I, no. 1. 12 Table VI, nos. 8, 9, 10, 71, 72, 48. 13 Table VI, nos. 42-51, 71, 72. 14 Table VI, no. 53. lo Table VI, no. 55. 16 Table VI, no. 56. 17 Table VI, no. 54. Fig. 84. Ravenna : ivory panel of chair of Maximianus. Christ Healing the Blind. 9 8 EARLY CHRISTIAN ICONOGRAPHY Constantinople; 18 or to find it the characteristic composition of the Alex- andrian-Coptic group. Group III, the Alexandrian-C optic group, is also a continuation of the Hellenistic type with the added feature of the Coptic cross which Christ carries in the miracle. This Coptic variation of the type appears on the ivory chair of Maximianus 19 (Fig. 84), the ivory book cover in the Bibliotheque Nationale at Paris, 20 the Murano cover 21 (Fig. 85), and the Fig. 85. Ravenna : Museum, ivory book COVER FROM MURANO. CHRIST HEALING THE Blind. Coptic pyxides in the Vatican Museum, 22 at Pesaro, 23 in the Cluny Museum 24 and the Micheli Collection 23 at Paris. The regular Hellenistic type without the cross occurs on three other Coptic ivories, two in Paris 26 and one in the Sancta Sanctorum at Rome. 27 Group IV, the Syrian type, represents two blind men seeking the mercy of Christ. This type, which follows the account in Matthew, occurs in the Rabula Gospels 28 from Zagba in Mesopotamia. The Gospels also depict a scene (Fig. 86) of the Healing of the Blind and the Halt which recalls the rendering on the chair of Maximianus. Curiously enough the 18 Table VI, no. 57. 19 Table VI, no. 58. 21 Table VI, no. 60. 22 Table VI, no. 61. 24 Table VI, no. 63. 23 Table VI, no. 64. 27 Table VI, no. 67. 28 Table VI, no. 68. 20 Table VI, no. 59. 23 Table VI, no. 62. 28 Table VI, nos. 65, 66. EARLY CHRISTIAN ICONOGRAPHY 99 Fig. 86. Florence : Laurentiana, Gospels of Rabula. Healing of the Blind and Halt. type also appears in the sixth century in the Fragment of Matthew from Sinope in Asia Minor 29 (Fig. 87) which from its relationship with the Rossano Gospels belongs to a proto-Byzantine group of Asia Minor, and in the ninth century in the Byzantine manuscript no. 510 30 in the Biblio- theque Rationale at Paris. I have called the type Syrian because it occurs in the Rabula Gospels and it may have been Syrian influence which carried Fig. 87. Paris : Bibl. Nat., Ms. fragment from Sinope. Christ Healing the Blind. it to Sant Apollinare Nuovo. 31 It made its way into Asia Minor, since we find it in the fragment from Sinope, and even into the early Byzantine 26 Table VI, no. 69. 30 Table VI, no. 70. 31 Table VI, no. 74. Cf. for a recent statement of the evidence for Syrian influence on Ravennate art, Strzygowski, Oriens Christ., V, 1915, p. 83. ioo EARLY CHRISTIAN ICONOGRAPHY art represented by Ms. gr. 510 in the Bibliotheque Nationale. But this same Greek manuscript contains the other version of the scene which was already present in the Rossano Gospels (see p. 101), an Asia Minor manu- script whose iconography is persistently Byzantine, and it was this second form which prevailed in subsequent Byzantine art, supplanting the sporadic Syrian version, and common enough to justify its title as the Byzantine type. Whether the Syrian type originated in Asia Minor or Syria, it was certainly Eastern. Therefore, its appearance on a sarcophagus in Rome 32 and in the mosaics of Sant’ Apollinare Nuovo at Ravenna is indicative of the introduction into the West during the fifth and sixth centuries of Oriental motifs. The scene on the Milan book cover 33 represents Christ, accompanied by a disciple, raising His hand in blessing toward the small figure of a blind man who comes forward leaning on a stick, while a second blind man emerges from an aedicula in the background (Fig. 88). At the out- set, this scene with two blind men associates itself with the Eastern type of either Asia Minor or Syria. This form appeared on the walls of Sant’ Apollinare Nuovo shortly after the time that the ivory was carved; it seems likely that the type had been introduced directly from the Orient into Provence as it had been brought from the East to Ravenna by 32 Table VI, no. 72. 33 Table VI, no. 73. EARLY CHRISTIAN ICONOGRAPHY ioi Oriental workmen. While the direct relation of the scene on the Milan ivory to any particular centre must remain problematic, the type was East- ern and goes to confirm the curious eclectic character of the scenes on the book covers. Group Y, the Byzantine type, as it appears in the Codex Rossanensis 34 (Fig. 89), shows Christ, accompanied by two disciples, touching the eyes of a single blind man who bends before Him ; to the right is the pool of Siloam where the blind man washes the spittle from his eyes while a large Fig. 89. Rossano: Gospels. Christ Healing the Blind. crowd of astonished people watch the restoration of his sight. This type was followed in the ninth century in the Byzantine manuscript no. 510 at Paris 35 (Fig. 90), in the eleventh century in the Byzantine mosaics of San Marco, 36 and passed in the tenth century from Constantinople to Ottonian art 37 ; in the following century it was used in the eclectic frescoes of Sant' Angelo in Formis. 38 It represents a typical Byzantine rendering of the Biblical account, the source in this case being the Gospel of John. The dependence on the account in John was most natural, for in Asia Minor, where the type first occurred in the sixth century in the Codex Rossanensis, the centre of religious inspiration was the church of Ephesus, founded by the Evangelist John and built upon his teachings. 34 Table \ I, no. 75. 83 Table VI, no. 76. 39 Table VI, no. 81. 37 Table VI, nos. 77, 78, 79. 33 Table VI. no. 80. 102 EARLY CHRISTIAN ICONOGRAPHY XII THE HEALING OF THE PARALYTIC I he Gospels refer to two different occasions when Jesus restored the use of his limbs to a man sick of the palsy: once at Capernaum (Mark ii, 1-12 > Luke v, 1S-26), and another time at the pool of Bethesda (John v, 1-15); in each case Jesus said to the afflicted man, “Take up thy bed and walk.” As it was the artistic tradition from the second to the ninth century to picture only the last act of the miracle, it is impossible to tell which of the two events is being depicted. It is also difficult to differentiate the representations of the various schools, as the artists of both the East and the West persistently show the paralytic departing, at Christ’s com- mand, with the bed on his shoulders. In a general way the following types appear in the representations : the Hellenistic type of simple, symbolic character where the paralytic is usually represented in profile and Christ is not always figured in the scene; an Oriental-Hellenistic group that is a direct continuation in late sixth century Eastern art of the early catacomb and sarcophagus type; two Coptic variations of the Hellenistic type; and a developed and his- torical Byzantine type which portrays Christ and the disciples approaching a house where men lower the bed-ridden paralytic through the roof. From the monuments on which they occur the two Coptic renderings may be designated as the Alexandrian-Coptic, where the paralytic, carrying his bed in profile, looks back at the Saviour who holds in His hand the Coptic Fig. 90. Paris: Bibl. Nat., Ms. gr. 510. Christ Healing the Blind. EARLY CHRISTIAN ICONOGRAPHY 103 cross, and the “Palestinian-Coptic,” wherein the paralytic is depicted in more or less frontality holding his bed with only the front end showing, and the Saviour carries a cross. Hellenistic Type. Like all the types of this class, the Hellenistic version of the Healing of the Paralytic was highly symbolic. As a scene of deliverance it sym- bolized the Baptism, for a cleansing from sin, which was the significance of the Baptism, was included in the physical restoration of the paralytic. 1 This dogmatic connection with the Baptism, which Tertullian 2 plainly as- serted, Wilpert has detected in the frescoes of the catacombs wherein the scene of the paralytic is combined with other symbols and with the actual representations of the Baptism. 3 The Hellenistic type appears as early as the second century in the Roman catacomb of Priscilla. 4 In all the catacomb frescoes (Lig. 91) down to the fifth century the paralytic is usually depicted 5 carrying Fig. 91. Rome: Catacomb of Callixttjs, fresco. Healing of the Paralytic. his bed upside down on his shoulders and with two exceptions Christ is omitted. On the “Roman glass,” 6 dating from the third to the fifth cen- tury, the scene is similar to that of the frescoes and Christ is only once introduced into the composition. 1 On the Western sarcophagi 8 a more consistent type is preserved; the paralytic is always small of stature, and stand? in profile at the feet of Christ holding his bed with the legs hanging Lamberton, St. John’s Gospel in Roman Catacomb Painting, p. 132. 2 Lamberton, op. cit., p. 127. ■ W ilpert, Le Pitture dclle catacombc romane, pp. 243-245; Fractio Panis, p. 66. * Table VII, no. 1. -Table VII, nos. 1-19. « Table VII, nos. 20-23. ' Table VII, no. 23. s Table VII, nos. 24-46. 104 EARLY CHRISTIAN ICONOGRAPHY Fig. 92. Rome: sarcophagus. Healing of the Paralytic. down his back (Fig. 92). There are three exceptions to this rule in which he holds the bed endwise on the ground before him. 9 The scene on the Milan book covers 10 is of the Hellenistic type (Fig. 93). The paralytic, facing Christ and a disciple, stands holding his bed upside down on his shoulders. The presence of this form of the scene in Provence confirms the eclectic character of these ivories, which were done under the combined influence of the Hellenistic art of the West emanating from Rome and the Eastern influences introduced by Syrians and other Orientals. The East. With the exception of the Byzantine type the Eastern representations are simply a continuation, with certain not too well defined variations, of the Hellenistic form. The Oriental-H ellenistic group represents a continuation in Oriental art of the Hellenistic method of depicting the miracle. The fact that one of its examples came from Syria and another presumably from Alexandria confirms the general character of the Hellenistic types and indicates an Eastern origin. The scene in the Rabula Gospels 11 reverts to the cata- comb method of depicting the paralytic carrying his bed upside down on his shoulders. On a pyxis of Bologna 12 is a representation which is similar 9 Table VII, nos. 27, 32, 46. 10 Table VII, no. 47. 11 Table VII, no. 48. 12 Table VII, no. 49. EARLY CHRISTIAN ICONOGRAPHY 105 to the scenes on the sarcophagi; save that the paralytic is as large as Christ. This pyxis, from its similarity to the fine Alexandrian pyxis in the Berlin Museum 13 and to an Alexandrian bone carving in the same museum, is ascribed to Alexandria. 14 Much the same representation also occurs on a pyxis from Darmstadt. 15 This continuation of the Hellenistic type ap- Fig. 93. Milan : Cathedral, ivory book covers. Healing of the Paralytic. pears on the ciborium columns at San Marco, 16 in one of the mosaics of _Sant’ Apollinare Nuovo, 17 and on an encolpium of the sixth or seventh cen- tury at Constantinople. 18 The Alexandrian-C optic group represents the paralytic holding tight against his body the two front legs of the bed, of which only the front end shows, and looking back at the Saviour. This manner of depicting the scene occurs on the Etschmiadzin book cover, 19 an ivory book cover in the Bibliotheque Nationale at Paris 20 (Fig. 94) and an ivory in the Micheli Collection at Paris, 21 all of which follow the iconography of the Max- imianus chair and in other scenes fall into the Alexandrian-Coptic group. The scene on the Carolingian book cover in the Bodleian Library at Oxford (see Fig. 169), instead of following the iconography of the Milan book covers as in many other scenes, preserves in this case the rendering of the Alexandrian-Coptic monuments. 22 It depicts the paralytic holding his bed in a frontal position. In the scene of the Miracle of Cana 23 this ^ oge, Elfenbeinwerke (Cat. Kaiser-Friedrich Museum), 1902, no. I. 4 Y\ulft, Altchristliche Bildwerke (Cat. Kaiser-Friedrich Museum), no. 428; Dalton, Byzantine Art and Archaeology, p. 195. 1 ' Table VII, no. 50. 16 Table VII, no. 51. 17 Table VII, no. 52. 19 Table VII, no. 54. 19 Table VII, no. 55. 20 Table VII, no. 5 7. 21 Table VII, no. 56. 22 Table V II, no. 58. 23 See p. 92. io6 EARLY CHRISTIAN ICONOGRAPHY ivory also followed the Alexandrian-Coptic type and in the representation of Christ trampling on the four beasts the Bodleian cover, as we shall see later, presents a type which was transmitted into Carolingian art from Alexandria 24 ; it is not surprising therefore to find here an Egyptian version of the Healing of the Paralytic. The Coptic type wherein the paralytic is represented in profie look- Fig. 94. Paris : Bibl. Nat., ivory book cover. Healing of the Paralytic. ing back at Christ, who carries in His hand the Coptic cross, occurs on the Murano book cover at Ravenna 25 (Fig. 95) and on four pyxides in the Youlgrave, 20 Vatican, 27 Cluny, 28 and Basilewsky 29 collections. In the case of this group it is true that there are no examples to show a Palestinian origin or continuation of the type, and yet on the table I have characterized the type, for consistency, as Palestinian-C optic. This dis- tinguishes the type from the Alexandrian-Coptic, and is consistent with the general trend of the iconography of the monuments on which this variant occurs, for the Murano cover, and the group it represents, show in other scenes the influence of Palestinian iconography. The Byzantine group presents the amplified or historical composition in which the paralytic on his bed is lowered through the roof of a house. There is no doubt that in this group it is the miracle of Capernaum, as related in Mark and Luke, that the artists are depicting, for at the pool of Bethesda there is no mention of a house in the account of John. 24 See p. 146 sq. 25 Table VII, no. 59. 26 Table VII, no. 60. 27 Table VII, no. 61. 28 Table VII, no. 62. 20 Table VII, no. 63. EARLY CHRISTIAN ICONOGRAPHY 107 Fig. 95. Ravenna : Museum, ivory book COVER FROM MURANO. HEALING OF THE PARA- LYTIC. Curiously enough there are two representations of the Healing of the Paralytic among the mosaics of Sant’ Apollinare Nuovo, of which one is Hellenistic and the other is the earliest example of this amplified render- ing. 30 Judging from most of the Byzantine types, this form of the scene must have originated in some proto-Byzantine monuments like the Rossano Gospels and have first been formulated in Asia Minor. This assumption Fig. 96. Rome : S. Saba, fresco. Healing of the Paralytic. is borne out by the appearance of the type in the eighth century in the very Byzantine frescoes of the basilica of S. Saba on the Aventine 31 fFig. 96). In the ninth century this form appears in the miniatures of the Homilies 30 Table VII, no. 64. 31 Table VII, no. 65. io8 EARLY CHRISTIAN ICONOGRAPHY of Gregory of Nazianzus (Bibl. Nat., gr. 510, Paris). 32 The eleventh cen- tury example in the Greek Gospels no. 74, Bibliotheque Nationale 33 (Fig. 97), shows essentially the same type, although more stylized and narrative in the rendering. As in the case of other Byzantine types, it passed into Fig. 9 7. Paris : Bibl. Nat., Ms. gr. 74. Healing of the Paralytic. Ottoman art where it occurs among the manuscript illuminations 34 and in later centuries became the usual composition employed in Italo-Byzan- tine art. XIII THE RAISING OF LAZARUS The Raising of Lazarus, from its intimate relation with the Christian hope of spiritual salvation in the life hereafter, was from the earliest days of Christian art the most popular and frequently represented of Christ’s miracles. Inasmuch as the methods of figuring the scene in the vari- ous art centres of the Mediterranean world were, with the exception of the Byzantine type, based on a general Hellenistic form and differed only in details, the examples divide more into groups than into types. The groups or types, whichever they may be, are the Hellenistic, the Oriental-Hellenistic, the Alexandrian-Coptic, the “Coptic,” and the Byzan- tine. The Hellenistic type, appearing on Western monuments, sub-divides into five variants of the general type. With exceptions in the first and fifth variants, the uniform characteristics of the Hellenistic type show Christ touching with His wand the figure of Lazarus who, wrapped like a mummy with his head covered with a cloth, stands in the doorway of a little gabled aedicula set upon a podium. The sub-groups are charac- 32 Table VII, no. 66. 33 Table VII, no. 68. 34 Table VII, no. 67. EARLY CHRISTIAN ICONOGRAPHY 109 terized as follows : In Group I, composed of the earliest examples of the scene, Lazarus sometimes lies on the steps of the aedicula and is seldom represented as a mummy; Group II presents only the uniform character- istics of the Hellenistic type as a whole; in Group III one of the sisters of Lazarus, kneeling at the feet of Christ, is added to the composition; Group IV introduces the second sister standing at the side of the Saviour; Group V is a unique and highly symbolic type appearing only on the “Roman glass,” where Lazarus, wrapped like a mummy with his head bare, leans against a kind of foliate growth. The Oriental-Hellenistic group, composed of Eastern examples of the foregoing type, shows in its principal minor division a Syro- Anatolian rendering wherein the tomb is represented in frontality as an arch supported on two columns. The Alexandrian-Coptic continues the Hellenistic type with the gabled aedicula in which Lazarus stands as a mummy with his head covered, but on the Coptic monu- ments Christ carries a cross instead of the traditional wand. A later Coptic group depicts the tomb in strict frontality with a faqade which has a kind of mansard roof, Lazarus with his head bare, and Christ carrying the Coptic cross. The developed Byzantine type shows, with certain variations, Christ, followed by disciples and spectators and with Mary and Martha lying prostrate at His feet, commanding Lazarus to come forth from the sepulchre or rock hewn tomb; servants are represented either holding back the entrance slab of the tomb or unwinding the swathed Lazarus, and they often hold their noses at the stench arising from the sepulchre. THE HELLENISTIC TYPE That the local methods of representing the scene throughout the Med- iterranean, with the exception of the proto-Byzantine in Asia Minor, vary from a general type only in slight details may perhaps be explained by the highly developed and universally accepted symbolical character of the scene and the fact that only the Evangelist John relates the event. 1 Typify- ing the resurrection of the body, it stood in the Christian mind as the guarantee of immortality. The Hellenistic manner of representation, appearing first in the cata- combs of the West, must have begun as a memory picture brought to Rome from the Hellenistic centres of the Orient by those Eastern Chris- 1 John, xi. 1 10 EARLY CHRISTIAN ICONOGRAPHY tians who first spread the new faith in the cities along the shores of the Mediterranean. 2 In accordance with the usage of the Jews, who adopted embalming from the Egyptians, and also with the account of John (xi, 44) the Western artists figured Lazarus as a little mummy “bound hand and foot with grave clothes; and his head hound about by a napkin.” Quite the opposite result took place in regard to the sepulchre; while the cus- tomary burial of Judea was in rock hewn tombs and the words of the * Evangelist say that “it was a cave, and a stone lay upon it,” nearly all the representations of the first six centuries, either from the technical difficulties of depicting a cave or, as is more likely, from the influence of the temple tomb widely used in Hellenistic times, represent the sepulchre in the form of an aedicula. Group I, in the Hellenistic division, includes the early second century frescoes in the Roman catacombs 3 and two pieces of decorated glass. 4 The iconography of the scene on these examples is unsettled and appears to antedate the establishment of any type. The Raising of Lazarus in the catacomb of Priscilla, which dates in the first half of the second cen- tury, 5 is unique in treatment. In the doorway of a regular gabled aedicula Lazarus, swathed in grave clothes like a mummy, lies diagonally across the door. At a little distance from the tomb stand two figures, one of whom is undoubtedly a woman. Wilpert says that in this first representa- tion of the scene Christ is not figured at all but that the male person is the resurrected Lazarus. His reason for refusing to identify the man with Christ is the familiar gesture of the woman who lays her hand upon his shoulder. This attitude seems to indicate the personages as Lazarus and his sister. Another reason given by Wilpert for interpreting the figure as Lazarus is his white raiment and the fact that he holds his arms crossed over his breast and looks back at the aedicula “mit grossen Augen.” 6 The double representation of Lazarus is consistent with the date. In a badly damaged fresco of the second century in Praetextatus appears the lower half of two figures standing before a tomb, who are usually interpreted to be Christ and one of the sisters of Lazarus. 7 In two frescoes in Callix- 2 Ainaloff, The Hellenistic Origins of Byzantine Art, 1900 (Russian). 3 Table VIII, nos. 1-4, 6. 4 Table VIII, nos. 5, 7. 5 Table VIII, no. 1. 6 Wilpert, Fractio Panis, p. 4; Lamberton, St. John’s Gospel in Roman Catacomb Painting, p. 27. 7 Lamberton, op. cit., p. 28, note 33. EARLY CHRISTIAN ICONOGRAPHY in tus (Fig. 98), a fresco in the Lateran Museum, and on a glass dish from Podgoritza. Lazarus is not depicted as a mummy. A less curious example occurs on a piece of ‘‘Roman glass,” included in the same group, where Lazarus, wrapped like a mummy, lies on the steps of the aedicula. Group II, which represents the simplest and most characteristic render- ing of the Hellenistic type, shows Christ, clothed in classical garb, touch- ing with His hand the mummylike figure of Lazarus, who stands with his head veiled in a napkin in the door of an aedicula resembling a Roman Fig. 98. Rome : Catacomb of Callixtus, fresco. The Raising of Lazarus. temple. 8 The aedicula, set upon a podium, is approached by a flight of steps leading up to the entrance which is usually flanked by two Corinthian columns. While Christ generally touches Lazarus with His wand, the wand is sometimes omitted and the Saviour simply raises His hand in an attitude of command. 9 10 The rendering of Group II occurs in the third century catacomb frescoes seven times and in those of the fourth century forty-one times. It also appears on sarcophagi of Arles 19 and Rome. 11 Group III in the Hellenistic division is peculiar to the sarcophagi of 8 Table VIII, nos. 8-65; the table gives only the characteristic examples as all the others are the same. Of the third century frescoes there are seven examples (Wilpert, Le Pitture, pis. 55, 45/1, 65/2, 71/1, 93, 108/2, and De Rossi, Bull., 1873, pi. 1). In the fourth century frescoes there are forty examples which are listed by Lamberton, of. cit., p. 105, note 142. s Table VIII, no. 61; W ilpert, Le Pitture, pi. 228/4. 10 Table VIII, nos. 59-60. 11 Table VIII, nos. 62-64. Two other scenes on the sarcophagi are often listed as Resurrections of Lazarus. That these scenes do not depict the resurrection, but repre- sent other miracles, is shown by a sarcophagus in the Lateran Museum at Rome (Garr., of. cit., V, pi. 367/1) where all three scenes are carved on the same face of the sarcophagus. The scenes on this sarcophagus are divided into two rows ; in the left hand corner of the upper row Christ is represented touching with His wand the figure of Lazarus who stands, wrapped like a mummy, in the doorway of an aedicula. In the same row appears the miracle of the Raising of the Son of the Widow of Nain II 2 EARLY CHRISTIAN ICONOGRAPHY the West 12 and the Milan book covers 13 (Fig. 99). Here one of the sisters of Lazarus, kneeling at the feet of Christ, is introduced into the composi- covers. The Raising of Lazarus. tion. The account in the Gospel of John (xi, 32) reads, “Then when Mary was come where Jesus was, and saw him, she fell down at his feet, saying unto him, Lord, if thou hadst been here, my brother had not died.” It is to be presumed, then, that the prostrate figure represents the sorrow- ing Mary. This single figure does not appear on Eastern examples but characterizes the scene on most of the Roman sarcophagi and on several of Gaul. The appearance of this scene on the ivory book covers of Milan is important, for the same rendering of the scene is found on several sar- cophagi of southern Gaul 14 (Fig. 100). All the features of the scene are identical with those on the Provencal sarcophagi, even to the single disciple wherein Christ, by the miraculous touch of His wand, brings into a sitting position a little mummy that lies at His feet. In the right hand corner of the lower row occurs the Raising of the Daughter of Jairus, a scene where Christ places His hand on the head of a little girlish figure who is seated on a bed behind which stand two figures watching the miracle. This last scene of the Raising of Jairus’s Daughter usually has the figure of the mother kneeling at the feet of Christ as is seen on another sarcophagus in the Lateran (Garr., pi. 376/4) ; in this representation the bed with the dolphins on the head piece exactly duplicates the bed on the other sarcophagus and is the regular type of the bed in this scene on all the examples. Another scene of the resurrection of the dead is the Awakening of Tabitha by Peter which is depicted on a sarcophagus of Arles (Garr., pi. 400/8) and on a sarcophagus of St. Maximin (Garr., pi. 353/2). 12 Table VIII, Group 3. The same type also occurs on the fifth century (?) silver casket of Brivio, now in the Louvre (Lauer, Mon. Plot, 1906, XIII, pi. XIX). Unlike the scene on any of the sarcophagi, Christ on the casket wears a nimbus and the tomb, while it has a lintel, is surmounted by a semi-circular pediment. 13 Table VIII, no. 92. 14 Table VIII, nos. 66, 67, 71, 89, 90. EARLY CHRISTIAN ICONOGRAPHY 113 who follows Christ as a witness of the miracle. This occurrence of the same scene on the Gallic sarcophagi sustains the connection of the covers with Provence and it is obvious that the use on the ivory of the distinctively Western method in depicting the scene precludes an Oriental origin. Group IV, wherein Martha stands at the side of the Saviour, contains Fig. ioo. Aix: sarcophagus. The Rais- ing of Lazarus. very few examples (Fig. 101). From the statement in the Gospels which says that Martha when she heard that Jesus was coming “went and met him,” this standing figure may be interpreted to represent the older sister of Lazarus. The assumption is in part borne out by the fact that the standing figure, which I would interpret as Martha, occurs on four sar- cophagi whereon the kneeling figure of Mary is also represented. Two of these sarcophagi are at Rome, 15 one is in Arles, 16 and the other from Auch. 17 The standing figure of Martha appears also on a sarcophagus of Clermont-Ferrand 18 and on an ivory pyxis of Darmstadt, 19 where in a suspiciously curious scene Lazarus, dressed in a tunic, kneels in the open doorway of a tomb with a conch roof. Group V presents an odd method of representing the scene on the so- called “Roman glass.” 20 Instead of being depicted in an aedicula, Lazarus on these bits of glass is shown wrapped like a mummy and with his head hare, leaning against a kind of foliate growth whose calligraphic scrolls, arranged in a row symmetrically on either side, appear to frame his body. Christ, like a magician, touches his body with a conjuror’s wand (see 15 Table VIII, nos. 97-98. ir ’ Table VIII, no. 95. 17 Table VIII, no. 96. J8 Table VIII, no. 93. 19 Table VIII, no. 94- 29 Table VIII, Group 5. EARLY CHRISTIAN ICONOGRAPHY 114 Fig. 74). We cannot say whether or not this was an early manner of representing the scene before any type had been established, for the date and even the provenience of the “Roman glass” are matters of speculation. The feature of Lazarus with his head bare, which is in direct contradiction of the Gospel account and the accepted method of depicting him, will appear later on a group of Coptic ivories and on the Brescia casket. The Fig. ioi. Rome: sarcophagus. The Rais- ing of Lazarus. casket is supposed by Strzygowski to have come from Asia Minor and the Coptic ivories represent the native Egyptian method, seen on the Fayum portrait mummies, of painting or carving an uncovered head of the deceased on the top of the mummy. In view of this it may be possible that the Roman glass was imported from an Eastern centre. The East The Oriental-Hellemstic group does not represent any new type, but only includes under three heads the variations of the Hellenistic method of representing the Raising of Lazarus occurring on monuments of un- questionably Oriental origin. The Hellenistic types of the early Christian scenes, as I have already said, must many of them have come from Hel- lenistic centres of the East. It is, therefore, not surprising to find the early types continued on the later monuments of the East. The Alexan- drian origin, which Ainaloff suggests 21 for the Hellenistic rendering of the scene, is confirmed by the appearance of the catacomb form on a monu- ment which in other scenes shows Alexandrian-Coptic iconography, namely, the Bologna ivory. 21 Ainaloff, op. cit. EARLY CHRISTIAN ICONOGRAPHY ii5 Group YI in the Hellenistic-Oriental division includes a single monu- ment, the Brescia casket, 22 which dates from the fifth century and which Strzygowski has attributed to Asia Minor. 23 The only unique feature is the bare head of Lazarus. Otherwise the scene is the same as the regular catacomb type. In Group VII of this same class, which can be broadly assigned to the Syr 0- Anatolian region, we find, instead of the regular gabled aedicula seen in the frescoes and on sarcophagi, a tomb of which only the fagade, which is an arch supported on two columns, is represented in strict frontal- ity. This type of tomb first occurs on a relief recently discovered in Phrygia (Lig. 102) and dated by Ebersolt in the fourth century. 24 The same type also appears on two sarcophagi 25 of S. Vitale at Ravenna which Diitschke dates in the fourth century (Pig. 103). It is to be noted Fig. 102. Constantinople: Imp. Ottoman Museum, relief from Phrygia. The Rais- ing of Lazarus. that the tomb of Lazarus on these three examples resembles the end of the sarcophagi with semi-cylindrical covers at Ravenna. Although the Phrygian relief would suggest that the tomb of Lazarus with a round arched entrance was primarily an Asia Minor method of representing the scene, the fact remains that it was in northern Syria rather than in Asia Minor that the arched entrance was frequently used for the fagades of tombs and sepulchres. In many places of northern Syria there are remains of Christian tombs which are either actually barrel vaulted 20 or have barrel vaulted porticoes entered by arched doorways, 27 22 Table \ III, no. 106. 23 Strzygowski, Kleinasien, p. 213. 24 Ebersolt, Revue archeologique, IV, XXI, 1913, pp. 333-339, fig. 3. 2j Table \ III, nos. 108-109; Diitschke, Ravennatische Studien, Leipzig, 1909, p. 228. 25 Catalogue of Photographs taken by the American Archaeological Expedition to Syria in 1899-1900, Princeton, X. J., nos. 365, 323, 238, 244 243. -• H. C. Butler, Architecture and Other Arts, pp. 106, 109, no, 158, 243, 300. ii6 EARLY CHRISTIAN ICONOGRAPHY while the use of the arch is one of the noticeable features of the Christian architecture. The arch and the barrel vault, if not so common in the free standing tombs of Anatolia, were constantly employed in the general archi- tecture of the country 28 and in Phrygia there have been found many re- liefs 29 which show the arcuated faqade rendered frontally in much the same manner as in the scene of the Raising of Lazarus on the Phrygian Fig. 103. Ravenna: S. Vitale, sarcopha- gus of the Exarch Isaac. The Raising of Lazarus. relief. The type of tomb, then, on the Anatolian relief and the two Ravenna sarcophagi, while it seems to be an attempt to picture the faqade of the free standing Christian tomb of northern Syria with its arcuated entrance, can not surely be connected with any definite region of the East. The Eastern decorative elements on so many sarcophagi of Ravenna and Provence which have the barrel cover instead of the gabled lid, and the generally admitted relations of these centres with the East, sustain the connection of this type with the general region of Syro-Anatolia. The connection with the East is borne out by some fragments of a white marble vase found at Sbeitla (Tunisie). The Raising of Lazarus on one of these fragments is practically identical with that of the Phrygian relief above mentioned. Both reliefs belong to a class of objects evidently de- signed for liturgical use, examples of which have been found in Bulgaria, Greece, Egypt, and Cyprus. Their community of style, ornament, and 28 Ramsay and Bell, Thousand and One Churches, p. 435 sq. 29 Ramsay, The History and Art of the Eastern Provinces, figs. 6-e, 20, 21, 22-a, 22-b, 29, 3i, 34- EARLY CHRISTIAN ICONOGRAPHY ii 7 iconography shows that all came from the same manufactory, and from the fact that they have been found mostly in the East, and from other indications, it is now held that they are of Oriental origin . 30 Group VIII of the Oriental-Hellenistic examples includes only the encolpium in the Museum of Constantinople 31 wherein Christ, wearing a nimbus, is represented touching with His wand the figure of Lazarus who, wrapped like a mummy and with his head covered by a napkin, stands in the entrance of a tomb which looks like a miniature apse. Another Oriental-Hellenistic example of the scene occurs on a relief at Constantinople , 32 which is dated as late as the beginning of the seventh century. Here the swathed Lazarus with his head covered is represented standing in strict frontality in the entrance of a gabled (?) faqade. The custom of representing either buildings or figures in a restricted frontality was an Oriental characteristic in direct contrast to the freer Hellenistic method of depicting them in perspective. Thus in the case of this relief as in that of the examples of Group ATI of the Oriental-Hellenistic Fig. 104. Paris: Bibl. Nat., ivory book covers. The Raising of Lazarus. division, the Eastern feature in what would otherwise be a regular Hel- lenistic rendering is the frontality of Lazarus and the tomb. Group IX, the Alexandrian-C optic, starts as usual with the purely Hel- lenistic form of the scene. It changes in the later Coptic examples, as the indigenous characteristics of Coptic art affect it, only to the extent 0 Merlin in Bulletin archeologique, 1913, p. clxxix, sq., pi. 46. This group of objects has been given detailed treatment by Michon, Bull, des antiquaires de France 1908, p. 268 sq. 31 Table VIII, no. hi. 32 Munoz, Nuovo Bull, di arch, crist., XII, 1906, p. 113, fig. 2. 1 18 EARLY CHRISTIAN ICONOGRAPHY that Christ carries the Coptic cross and the tomb becomes a nondescript sepulchre rendered in frontality. The first example occurs on the Alex- andrian pyxis at Bologna. 33 On the ivory book covers in the Bibliotheque Nationale at Paris, 34 which in iconography, as well as style, follow the Maximianus chair into the Alexandrian-Coptic group, essentially the same scene appears. While obviously derived from the Hellenistic form, it at the same time shows Christ with a cross and the gabled tomb fagade de- picted frontally (big. 104). In this last feature the scene is transitional to the Coptic type, as we find to be the case regarding later members of the Alexandrian-Coptic group in other scenes. Aside from the fact that Fig. 105. Cairo: Museum, ivory COMB FROM ANTINOE. The RAISING of Lazarus. Fig. 106. Paris : Cluny Museum, ivory pyxis. The Raising of Lazarus. the tomb has changed to a simple sepulchre of no particular character and represented frontally, the scenes on the Coptic pyxis at Bonn 35 and the odd ivory in the Micheli Collection at Paris 36 are the same as the scene on the covers. Group X, a Coptic type, is distinguished by Lazarus, his head bare, standing in the entrance of a frontal tomb fagade with a broken lintel and by Christ carrying a cross instead of a wand. On a sixth century ivory comb 37 from the excavations at Antinoe (Fig. 105), the tomb, of which only the fagade is represented, is shown as two spiral fluted columns bear- ing a broken, denticulated lintel, which gives the impression that the tomb had a mansard roof ; within the doorway stands the swathed figure of Lazarus, his head bare, and with curly wiglike hair; to the right of the tomb is Christ with the same wiglike hair, carrying a cross in His right 33 Table VIII, no. 112. 34 Table VIII, no. 113. 35 Table VIII, no. 114. 30 Table VIII, no. 115. 37 Table VIII, no. 117. EARLY CHRISTIAN ICONOGRAPHY 119 hand and a roll in His left. Four other Coptic ivories are connected in tvpe with this comb through their striking similarity in nearly every detail of the composition; a pyxis in the Cluny Museum at Paris 38 (Fig. 106) , a pyxis in the Basilewsky Collection at Petrograd 39 (Fig. 107), a pyxis in Fig. 107. Petrograd : Basilewsky Collec- tion, ivory pyxis. The Raising of Lazarus. the Vatican 40 (Fig. 108), and the Murano book cover at Ravenna 41 (Fig. 109). On all these examples is the same odd shaped tomb, Lazarus with his head bare, and both Christ and Lazarus with the same curly wiglike hair and drill holes for eyes; with the exception of the Basilewsky pyxis, where Christ carries only a roll, He is given the cross as the symbol of power. The similarity of the scene on the Cluny pyxis and the Antinoe comb is so close that even the wrappings of Lazarus make the same con- ventional design. There occurs on a Coptic textile from Akmim 42 a repre- sentation in which Lazarus is also depicted with his head bare but the tomb and Christ are not characterized, and Lazarus oddly enough appears to wear a nimbus. Thus we have a clear-cut type of the Raising of Lazarus in Egypt, which is to be characterized by the mansard roofed aedicula, Lazarus with his head bare, the cross in Christ’s hand, the drill holes in the eyes, and the curly, wiglike hair. Besides the absolute uniformity of the scene on these admittedly Coptic ivories, the feature of Lazarus with his head bare and his hair short and 38 Table VIII, no. 120. 39 Table VIII, no. 119. 40 Table VIII, no. 121. 41 Table VIII, no. 118. 42 Table VIII, no. 122. 120 EARLY CHRISTIAN ICONOGRAPHY curly may be associated most definitely with Egypt through the Fayum mummy portraits. While the custom throughout the history of Egypt was to depict the uncovered head of the deceased at the top of the mummy case, during the late Roman occupation of the country it became the cus- Fig. 108. Rome: Vatican, ivory Fig. 109. Ravenna: Museum, ivory book pyxis. The Raising of Lazarus. cover from Murano. The Raising of La- zarus. tom to paint a portrait head with short curly hair on the face of the mummy cloths . 43 The resemblance of Lazarus on the ivories, even to the conventional designs of the mummy wrappings, to these portrait mummies, is convincing . 44 The odd shape of the tomb on the ivories seems to have no apparent connection with Egyptian customs of burial, although it recalls vaguely an Egyptian mummy case set upon end. Group XI-A, the Byzantine type in the characteristic examples, follows accurately the account in John and introduces in a narrative composition all the characters of whom the Evangelist speaks. The type first occurs in the Codex Rossanensis 45 (Fig. no) which was executed somewhere in Asia Minor during the sixth century. Within a rock hewn tomb is seen the swathed body of Lazarus supported on the right by a servant; Christ, bearded and clothed with a purple tunic, over which hangs a gold mantle, wears a large cruciform nimbus; at the feet of the Saviour the two sisters 43 W. Flinders Petrie, Portrait Mummies of Kahun and Hawara, publication of the British School of Archaeology in Egypt, 1911. A bare headed mummy, represented by a figurine of Christian workmanship from a grave at Akmim, shows beyond question that it was the custom in Coptic times to depict such mummies as Lazarus with the head bare (Forrer, Die friihehristliehen Altertliiimcr aus dem Grdberfelde von Achmim- Panopolis, pi. XI II/ 17 ) . 44 Petrie, Roman Portraits and Memphis (IV), 1911, pis. XI, XIII, XIV. 45 Table VIII, no. 123. EARLY CHRISTIAN ICONOGRAPHY 121 of Lazarus. Mary and Martha, are prostrate; and behind Him, to right and left, are groups of disciples and spectators. This same type appeared among the lost mosaics of the Church of the Holy Apostles at Constantinople, which are dated by Heisenberg in the sixth century and by others in the ninth . 46 While certain of the details are left out in some later examples of this type (Fig. hi), this rendering of the Rossano Gospels was in its essentials faithfully preserved down to the fifteenth century in the East , 47 and imitated in the Ottoman 48 and Italo-Byzantine art 46 of the West where Byzantine iconography exerted influence. The presence of this type, with the two sisters omitted from the composition, on the ciborium columns of San Marco is rather hard to explain in view of the date in the sixth cen- tury which is usually given them . 50 Group XI-B, which is composed of Italo-Byzantine monuments and two examples from Mt. Athos dating from the twelfth to the fifteenth centuries, shows Lazarus with a nimbus . 51 XIV THE ENTRY INTO JERUSALEM A scene of far less popularity than many others, the Entry into Jerusa- lem is not only frequent on the monuments of early Christian art but also **’ See p. 31, note 74 47 Table VIII, nos. 127, 129, 130, 132, and Group 11-B. ♦-Table VIII, no. 131. 4ft Table VIII, nos. 126, 133-136. 50 Table V III, no. 125. 51 Table VIII, Group 11-B. 122 EARLY CHRISTIAN ICONOGRAPHY offers few marked features which aid in dividing the differing representa- tions into particular types peculiar to definite regions and centuries. The large division of East and West may be distinguished by the manner in which Christ rides; on the monuments of the West the Saviour is always depicted riding astride, while on those of the East He is invariably shown Fig. hi. Paris: Bibl. Nat., Ms. gr. 510. The Raising of Lazarus. seated sidewise on the ass. The actual types which emerge from the vari- ous representations of the Entry into Jerusalem are the Hellenistic or Western type, the Alexandrian-Coptic, the Palestinian-Coptic, and the Byzantine type. THE HELLENISTIC TYPE The dominant and distinguishing feature of the Hellenistic type is the representation of Christ riding astride. While all the monuments of this type are Western and this one detail of the scene distinguishes 1 the representations of the East from those of the West, the actual fashion of riding astride was not so much Western as it was Hellen- istic. The Greek rode astride his horse and this Hellenistic method of riding, followed by the Romans, was permanently preserved in the West. In the East, however, the native method of riding sidewise had by the sixth century so far supplanted in Asia Minor, Syria, and Egypt the older 1 Strzygowski, Bys. Dcnk., I, p. 38; Heisenberg, Grabeskirche und Apostelkirche , Leip- zig, 1908, II, p. 247; Dobbert, Jb. Preuss. Kunsts., XV, 1894, p. 149; Dalton, Byzantine Art and Archaeology, 1911, p. 656; Haseloff, Codex Purpnreus Rossanensis, p. 91 sq. EARLY CHRISTIAN ICONOGRAPHY 123 fashion that when the Entry into Jerusalem began to figure on the monu- ments of the Orient, the sidewise position was used for the figure of Christ . 2 Like the other Hellenistic scenes which originated in the cities of the East, were transmitted to Rome, and became Western types, the Fig. 1 12. Rome: sarcophagus. Christ en- tering Jerusalem. Hellenistc Entry into Jerusalem is a mere symbolical abbreviation of the actual event. Only the actual essentials of the scene are figured : as a rule the disciples who accompanied Christ are reduced to one; the city, and the multitude who came forth bearing palms, are omitted ; and the whole acclamation of Christ as King of the Jews is symbolized by the single figure of a youth plucking the branches in the top of a tree and by a man who spreads his mantle beneath the feet of the ass. Group I. The Hellenistic Type with Christ riding astride. On the monuments 3 of this group, which are all sarcophagi and with one exception all come from Rome, Christ, without the nimbus or any other symbol of His divinity, rides in the Hellenistic fashion (Fig. 112). Before Him a man spreads his mantle, and behind there usually follows a single disciple, although in three 4 representations the disciple is omitted. 2 The Hellenistic method of riding was preserved in scenes of actual horsemanship: fresco at Bawit (Diehl, Manuel, fig. 23), wood carving from Alexandria (Diehl, op. cit., fig- 25), Barberini diptych (Diehl, fig. 141), a sixth century silver dish from Kertch fDiehl. op. cit., fig. 150) and a Coptic textile in the South Kensington Museum (Dalton, op. cit., fig. 46). 3 Table IX, nos. 1-12. 4 Table IX, nos. 9, 10, 11. 124 EARLY CHRISTIAN ICONOGRAPHY All the scenes show the youth picking branches in the tree; in two 5 a few figures come forward with palm branches to meet him; and on two sarcophagi 6 the city of Jerusalem is represented. Group II. Provenqal Group of the Hellenistic Type. The scene on the Milan covers 7 (Fig. 113) connects itself with the representations of the Roman sarcophagi in that the Saviour rides astride, but is even more abbreviated and symbolical. Its essential difference from the scenes on the Roman sarcophagi is the omission of the youth in the tree. Group III. Manuscript Continuation of the Type in the West. While the scene in the Cambridge Gospels, 8 the Sacramentarium of Drogo, 9 the Codex Egberti, 10 and the Gospels of Bernward 11 is primarily a continuation of the Hellenistic type in that the Saviour is represented riding astride, it is also a derivative of what we shall see is the more amplified and historical Byzantine type. From Constantinople in the ninth, tenth, and eleventh centuries there was a steady infiltration into the West, and especially into the Ottoman Empire, of Byzantine decorative motifs and iconography. The more narrative representation of the scene in these manuscripts, in which the disciples are increased to three and even 5 Table IX, nos. 11, 12. 6 Table IX, nos. 8, 12. 7 Table IX, no. 15. 8 Table IX, no. 16. « Table IX, no. 17. 10 Table IX, no. 18. 11 Table IX, no. 19. Fig. 1 13. Milan: Cathedral, ivory book covers. Christ entering Jerusalem. EARLY CHRISTIAN ICONOGRAPHY I2 5 Christ entering Jerusalem. to six (Codex Egberti), and the number of people with palm branches to two, four, and six, and in which the city is always represented, is ex- plained by the spread of the Byzantine type. The East The Eastern types, like the Hellenistic, appear to have had only the canonical accounts in Matthew, Mark, Luke, and John 12 for textual sources. As the apparent sources were the same and as the slight popu- larity of the Entry into Jerusalem in the Orient did not tend to create distinct indigenous forms, the Eastern types, save for the Oriental feature of Christ riding sidewise, preserve quite closely the features of the older Hellenistic rendering. 12 Matt, xxi; Mark xi ; Luke xix; John xii. 126 EARLY CHRISTIAN ICONOGRAPHY Group IV. The Alexandrian-C optic Type where a carpet instead of a mantle is spread beneath the feet of the ass. The scenes on the Maximianus chair 13 from Alexandria (Fig. 114), the Etschmiadzin book covers, 14 a sixth century ivory cover in the Biblio- theque Nationale at Paris 15 (Fig. 115), and a carved lintel at el Fig. 1 15. Paris: Bibl. Nat., ivory book covers. Christ entering Jerusalem. Mu'allaka near Cairo 16 (Fig. 116) are bound into a group and segre- gated from the other representations by the curious feature of the long carpet which is unrolled, instead of a mantle, in the pathway of the Saviour. These representations are further bound together by other Fig. 1 16. Cairo: el Mu’allaka, carved wooden lintel. Christ entering Jerusalem. similarities in composition, by the odd, high stepping movement of the ass, and by Christ’s posture on the animal. In distinction from all other examples of the scene, the Christ on the Maximianus and the Etschmiadzin ivories carries in His hand a cross. This cross has already been pointed out as the emblem of divinity which the Egyptian artists regularly gave to Christ. The Etschmiadzin and Paris examples are more closely related by the introduction of a Tyche, wearing (Etschmiadzin) a crenellated crown, who comes forth as the personification of the city to meet the Saviour. This same representation of Tyche, with her turreted crown, appears among the miniatures of the Joshua Rotulus, 17 the original of which, according to Strzygowski, was executed in the third or fourth century by a Greek artist living in Alexandria. 13 Table IX, no. 20. 14 Table IX, no. 21. 15 Table IX, no. 22. 10 Table IX, no. 23. 17 Garr, op. cit., Ill, pi. 163. EARLY CHRISTIAN ICONOGRAPHY 127 The connection of this type with Egypt is sustained, not only by its appearance on the Maximianus chair, and by the presence of the cross which Christ usually carries on Egyptian monuments, but also by its occur- rence on the carved wooden lintel of the church of el Mu'allaka near Cairo Group V. Palestinian-C optic Type. There is little to distinguish this type from the regular Hellenistic type save the Saviour riding sidewise on the ass. I11 the Rabula Gospels 18 of Syria (Fig. 117) the Saviour, mounted sidewise, is followed by a single Fig. 117. Florence: Laurentiana, Gospels of Rabula. Christ entering Jerusalem. disciple, and before Him a man spreads his mantle while three other men come forward with palm branches. It is interesting to note that this scene is essentially duplicated on the ciborium columns of San Marco. 19 With slight variations this abbreviated Eastern type occurs in a Coptic textile from Antinoe 2 " and on a Coptic seal in the British Museum; 21 and was also continued in a mosaic of the chapel of John the Seventh at Rome. 22 Group VI. The Byzantine Type. This is the historical type which faithfully follows the Gospel accounts 18 Table IX, no. 24. Table IX, no. 25. In W ulff’s Altcliristliche Bildwerke (Cat. Kaiser-Friedrich Mu- seum). I, no. 72, p. 33, is published a stone relief from near Sohag (Egypt) ; on it is figured .'hat the author calls an Entry 7 into Jerusalem, where Christ wearing a cruciform nimbtu rides sidewise while one nimbed angel leads the ass and another follows. 20 Table IX, no. 26. 21 Table IX, no. 27. 22 Table IX, no. 28. 128 EARLY CHRISTIAN ICONOGRAPHY and translates into pictorial form the event with all its figures as related in the New Testament. Its earliest and perhaps first appearance is in the Codex Rossanensis from Asia Minor 23 (Fig. 118). In this minia- ture, Christ, wearing a cruciform nimbus and followed by two disciples, is depicted riding sidewise; two boys spread a mantle beneath the feet of the ass, while two other boys climb a tree, or are perched in its branches; a large multitude issues forth from the gates of the city to meet the Saviour. From the windows of the houses within the walls people look out upon the scene. This historical type of the Rossano Gospels existed in a mosaic on the walls of the Church of the Holy Apostles at Constantinople. 24 In the ninth century we see the Byzantine type fully developed, with all twelve apostles represented, in the Homilies 25 of Gregory of Nazianzus at Paris (Bibl. Nat., gr. 510), (Fig. 119). It continues on the Byzantine monuments, 26 with slight variations in the number of figures included in the scene, as the traditional type. As evidence of the spreading of the Byzantine type into the West, we find it in the eleventh century in the frescoes of S. Angelo in Formis. 27 Its influence is also apparent in the Ottonian manuscripts (Group III) which, although they adhere to the Hellenistic tradition in representing Christ riding astride, show, neverthe- less, an amplified and historical type like the Byzantine. 23 Table IX, no. 29. 24 See note 74, page 31. 25 Table IX, no. 30. 20 Table IX, nos. 31, 32, 33, 34. 27 Table IX, no. 35. EARLY CHRISTIAN ICONOGRAPHY I2Q Fig. ng. Paris: Bibl. Nat., Ms. gr. 510. Christ entering Jerusalem. XV THE LAST SUPPER The actual representation of the Last Supper as an historical event does not appear in Christian art until a late date. While it is necessary not to confuse this historical Last Supper with the Eucharistic repast , 1 so common in the catacombs, there also developed, besides the his- torical type, a liturgical type which was derived from the Eucharistic symbolism of the Hellenistic period. This liturgical type, which repre- sents the Last Supper as a First Communion, shows Christ standing and breaking bread, which, with the wine, He distributes to His apostles. The historical type is more narrative and regularly figures Christ and His disciples seated on the bolster of a sigma shaped couch about a semi- circular table, or in the later representations at a long straight table, on which is set the bread and wine. As a rule it represents the particular moment when Judas reaches forward to dip or to grasp the food and John affectionately reclines upon the breast of the Saviour. Besides these two types there was a transitional type, formulated in the West about the beginning of the sixth century, in which an actual Last Supper is represented with Christ and the apostles seated about a table, but the 1 Leclercq in Cabrol’s Dictionnaire d’archeologie chretienne ct de liturgie, s. v. “Cene,” col. 3046. For a discussion of the Last Supper in art see: Dobbert, “Das Abendmahl in der bildenden Kunst bis gegen den Scbluss des 14. Jahrhunderts,” Repertoriurn fiir Kunstivissenschaft, XIII (1890), pp. 281-292, 423-442; XIV (1891), pp. 175-203, 451-462; X\ fi892), pp. 357-384, 506-527; XVIII (1895), pp. 336-379; Heisenberg, Grabeskirche uiid A postelkirche, p. 175; Dalton, Byzantine Art and Archaeology, p. 658. EARLY CHRISTIAN ICONOGRAPHY 130 loaves and fishes, after the early catacomb conception of their Eucharistic significance, are substituted in place of the bread and wine of the historical event. It is this transitional type which appears on the Milan book covers and is important in showing the Western origin of the ivories. There- fore, in order that the transitional character and especially the Western relation of the type may be understood, it is necessary to give a brief out- line of the facts concerning the history of the Eucharist in early Christian art. The different theories which scholars have advanced concerning the Eucharist and the readiness with which the subject lends itself to de- ductions make it dangerous to do more than state the generally accepted steps in the formation and development of Eucharistic symbolism. From the very first the Multiplication of the Loaves and Fishes, after the inter- pretation in the Gospel of St. John , 2 was considered the principal symbol of the Eucharist. This is demonstrated by the fact that the baskets of FRESCO. loaves, the characteristic feature of the miracle, are included in the only representation of the Lord's Supper which occurs in the catacombs (Pris- cilla. second century ). 3 The miracle itself is represented by a series of catacomb paintings of the second century and the early third century in which seven people reclining on a couch of sigma shape partake of a meal at which are served loaves and fishes, while in the foreground of the scenes are usually figured seven baskets of bread . 4 In none of these representations is Christ depicted. By the first half of the second century the idea of feasting was no longer necessary to the symbolism, for in the crypts of Lucina in the catacombs of Callixtus 5 are two pictures (Fig. 120) consisting of a fish and a single basket filled with loaves while elsewhere 2 Lamberton, St. John’s Gospel in Roman Catacomb Painting, p. 79. 3 Lamberton, op. cit., p. 79. 4 Wilpert, Fractio Panis, pis. XIII-XIV. 5 Wilpert, Le pittnre dcllc catacoinbc romanc. pis. 27/1, 28. EARLY CHRISTIAN ICONOGRAPHY CM in the catacomb of Callixtus is a fresco in which a man extends his hands in blessing over a tripod on which are a fish and loaves ; before him stands a veiled orant raising her hands in prayer 6 (Fig. 121). There are other representations in which the scene is reduced only to the loaves and fishes. By the end of the third century the symbolism had been even more Fig. 121. Rome: Catacomb of Callixtus, fresco. abbreviated and the multiplication of the loaves alone had come to stand for the Eucharist. Thus the symbolism was narrowed down to the simple scene of Christ touching with His wand the seven baskets of bread. This scene while occurring twenty-eight times in the catacombs 7 during the third and fourth centuries was also equally common on the Western sarcophagi, although in these sarcophagi representations the number of baskets is usually reduced to six, three, or even one (Fig. 122). In a fresco of Callixtus the Eucharistic symbolism is extended by the addition of two disciples who present the bread and fish to Christ for His blessing. 8 This more historical, and probably less symbolical, rendering of the multiplica- tion where the disciples are about to distribute the bread and fish to the multitude, who are not figured, was rare in the catacombs but was a most common scene on the Roman sarcophagi of the fourth century (Fig. 123). An ample and interesting rendering of this Eucharistic Multiplication of the Loaves and Fishes is seen in the third century frescoes of the cata- combs of Alexandria in Egypt (Fig. 124). Christ is enthroned between the Apostles Peter and Andrew, who advance to receive His blessing on 6 Wilpert, op. cit., pi. 41/1. 7 For all the catacomb examples see Wilpert, op. cit., p. 269 sq. 9 Wilpert, op. cit., pi. 237/1, p. 276. 132 EARLY CHRISTIAN ICONOGRAPHY Fig. 122. Algiers : Museum, sarco- phagus. The Miracle of Loaves and Fishes. the food which they carry, while at His feet twelve baskets of bread are divided into two groups of six each. On either side of the central group are scenes of feasting which probably represent the Feeding of the Multi- tude, although the inscriptions IC over the standing figure at the left and H APIA MAPIA over another obliterated figure in the same group, have led to the suggestion that the scene represents the Miracle of Cana Fig. 123. Rome: sarcophagus. The Miracle of Loaves and Fishes. with both Jesus and the Virgin present. These inscriptions, however, over the group at the left, are in black while the authentic inscriptions over the two other groups are in red, and probably represent a late and popular emendation in the interpretation of the scene . 9 The same combination of 9 A curious rendering of the Miracle with Christ, three angels, twelve baskets, and the EARLY CHRISTIAN ICONOGRAPHY 1 33 Fig. 124. Alexandria: Catacombs, fresco. The Miracle of Loaves and Fishes. the Blessing of the Elements with the Feeding of the Multitude that ap- pears in the Alexandrian fresco occurs on the chair of Maximianus (Fig. 125) whose Alexandrian origin becomes more and more obvious as the different types on the famous cathedra are studied. Thus, the type of the Maximianus chair finds a natural continuation in the group of Christ seated and blessing the bread and fish offered by two disciples, on a pyxis of the sixth century in S. Pedro de la Rua at Estella, Spain (Photo. Lau- rent, Madrid, no. 916) ; the ivory is Alexandrian-Coptic in style and Christ carries a Coptic cross. Since in the third century the multiplication of the loaves without either the fish or the wine had become the symbol of the Eucharist, it is not surprising to find the creation of a new symbol for the missing ele- ment. the wine; the Cana Wedding, therefore, represented by Christ touch- ing with His wand the jars of wine, was introduced as a new and additional type of the Eucharist. 10 By the fourth century the natural relationship of the two themes of the Multiplication of the Loaves and the Miracle of the Wine is seen in the catacombs of SS. Pietro e Marcellino. 11 Here the two scenes occupy opposite end spaces in the vault of an arcosolium and in the lunette of the same arcosolium Wilpert restores a scene of feasting in which seven persons at a sigma shaped table are about to par- take of a fish on a tripod before them and a servant presents to one a cup of wine 12 (Fig. 126). In all these scenes of feasting where the Eucharistic four Evangelists is to be found on a Coptic relief in the Metropolitan Museum, New \ork (Morey: East Christian Paintings in the Freer Collection, p. 66). 10 W ilpert, op. cit., p. 277 ; Lamberton, op. cit., p. 89, quotes C. R. Morey who writes, “The breaking up of the Eucharistic symbolism through the isolation of the fish as a Christ symbol is, it seems to me, the reason for the introduction of the Cana Wedding as a new and distinctive type of the Eucharist.” 11 Wilpert, op. cit., pi. 186/1. 12 Lamberton, op. cit., p. 92; Wilpert, op. cit., p. 279, fig. 25. 134 EARLY CHRISTIAN ICONOGRAPHY meaning is apparent it is impossible to suppose that any of them symbolizes the Last Supper, for there was no significance in the Last Supper to the early Christians save as it symbolized the Eucharistic Transubstantiation. “A symbol itself is but an outward form by the understanding of which one appreciates a hidden significance. Hence it would be useless to give a form to a form, to symbolize a symbol. - ’ 13 The transitional type of the Last Supper is primarily a representation of the Last Supper as a symbol of the Eucharist. It is historical, however, Fig. 125. Ravenna : Museum, ivory panels from chair of Maximianus. The Miracle of Loaves and Fishes. to the extent that it depicts the actual repast that Jesus held with His disciples after His entry into Jerusalem. It is frankly symbolic in that it retains the catacomb symbolism of the Multiplication of the Loaves and Fishes and instead of depicting the actual repast of bread and wine it sub- stitutes, or rather retains, the fish in place of the wine. The most ap- parent example of this type is found at the beginning of the sixth century in the mosaics of Sant - Apollinare Nuovo (Fig. 127) where Christ and His disciples recline on a raised couch of sigma shape about a semi- 13 Lamberton, op. cit., p. 88. EARLY CHRISTIAN ICONOGRAPHY 135 circular table. Seated at the left end of the table is the Saviour who wears a cruciform nimbus and stretches forth His hand to bless the six loaves and the two fish which lie upon the table. The preservation of the old Eucharistic symbolism in an historical setting is clearly intentional on the part of the artist in this mosaic, and intended to give to the Last Supper a symbolic interpretation. While the highly developed symbolism of the Hellenistic period of Christian art was breaking down at the beginning of the sixth century there were still theologians at least as late as the fifth century who were expounding the Eucharistic significance of the Fish. 14 The transitional character of this scene, however, lies in the fact that the Fig. 126. Rome : Catacomb of SS. Pietro e Marcellino, fresco. Eucharistic Banquet. old Eucharistic repast is figured in the guise of an historical event. There is a striking similarity in the mosaic, especally in the form of couch and the table, whose ends -ftJrm three rectangular panels across the bottom of the composition, with the historical scene of the Last Supper in the Rossano Gospels (see Fig. 1 3 1 ) . This method of representing a banquet scene appears to have been more or less peculiar to Asia Minor in the sixth century, for the same sigma shaped couch about a semi-circular table occurs in the Feast of Herod in the Sinope Fragment. 1 ' 1 The subsequent Byzantine type as it first appears in the sixth century in the proto-Byzantine Gospels of Rossano is primarily an historical and narrative representation of the actual event without particular reference to its Eucharistic significance. This narrative tendency in Christian art was an Oriental reaction against 4 Augustine Chrysologus, the Pseudo-Prosper, C. R. Morey, The Princeton Theological Review, 1 ' Diehl, Manuel, fig. 125. and the author of the Narraiio; 1911, PP- 421, 424, 4 2C_4ju, see 1 36 EARLY CHRISTIAN ICONOGRAPHY the restrained, simplified and highly symbolic methods of rendering the Gospel scenes during the first five centuries when art, as well as thought, was dominated by Neoplatonic mysticism. While the realistic tendency was the result of indigenous Eastern influences, it naturally also appears in such centres as Ravenna where Oriental forms and methods met and gradually supplanted the Hellenistic symbolic types current in the West; 16 so that for a time there existed in the art of Ravenna an eclectic mingling of the two tendencies which resulted in transitional types wherein the older symbolism was presented in historical settings. From this fusion of the Hellenistic symbolism and the Oriental realism was formed a transitional type of the Last Supper which is most clearly expressed in the mosaic of Sant' Apollinare Nuovo. Although the mosaic scene is the clearest representation of this type, there exist other examples in the West where the Last Supper is repre- sented in a more limited historical manner. Among these other scenes which in the retention of the catacomb symbolism of the loaves and the fishes show a Western origin is the scene on the Milan book covers 17 (Fig. 128). Here are represented four figures reclining on a raised and sigma shaped couch about a semi-circular table on which are set seven pieces of bread and a dish containing a fish; at the left of the composition sits the Saviour blessing the food with His right hand; next to Him is a disciple who drinks while a second disciple watches ; at the end of the table opposite to the Saviour reclines a third disciple who turns his head to follow the Master’s gesture. 1,1 See Strzygowski : Ravenna als Vorort aramaischer Kunst , ( Orieus Christ., V, 1915, Fig. 127. Ravenna: S. Apollinare Nuovo, mosaic. The Last Supper. 7 Garr., op. cit., VI, pi. 455. EARLY CHRISTIAN ICONOGRAPHY 137 That the scene is a Last Supper, as it is called by Molinier , 18 Stuhl- fauth , 10 and Dobbert , 20 and neither the Supper at Emmaus as Venturi interprets it 21 nor Christ dining with Lazarus and his sisters as Westwood suggests , 22 is indicated by the gesture which Christ makes toward the viands on the table. The limited number of disciples is explained by the transitional character of the composition. Other representations of the Last Supper during the first eight centuries of Christian art in the West show an abbreviation of the number of the disciples who sat at table with the Saviour. In the scene on the columns of the ciborium of San Fig. 128. Milan: Cathedral, ivory book covers. The Last Supper. Marco 23 there are only three figures seated at the table, in the eighth century mosaic in the Chapel of John VII 24 the number is five, and in the seventh century Codex Cambricensis 25 there are eight disciples. V hile the narrative rendering of the scene on the Milan book cover is not so accurate and realistic as in the mosaic of Sant' Apollinare Nuovo, it is none the less a transitional representation of the Last Supper and is to be classified with the Ravenna example as a Western type. In the first place it preserves the Eucharistic symbolism of the loaves and fishes which was characteristic of the Ravenna representation. In the second place, although the number of disciples is abbreviated, it is none the less a narrative rendering of the Last Supper as an event and is similar to ' Molinier, Hist. gen. dcs arts appliques, I, Ivoires, p. 61. Stuhlfauth, Elfenbeinplastik, p. 68. 20 Dobbert, Repcrtoriuui, XIV, p. 183. \ enturi, Storia, I, p. 510. 22 V estwood, Fictile Ivories, p. 41. Garr., op. cit., VI, pi. 496/3. 24 Garr., op. cit., IV, pi. 280/7. 25 Garr., op. cit., Ill, pi. 141. 1 38 EARLY CHRISTIAN ICONOGRAPHY the scene of the mosaic in that the ends of the raised couch with the end of the table form three contiguous panels across the bottom of the com- position. This feature is important, for it recalls the Eastern banqueting scenes rather than the Western feasts in the catacombs where the people lie on a semi-circular bolster placed upon the ground. The semi-circular table at Ravenna with the seven loaves and the two fishes, which reappears on the column of San Marco with six loaves, was, according to Fleury, 26 one of the most ancient forms given to the sacred table itself ; at Vienne he Fig. 129. Florence: Laurentiana, Gospels of Rabula. The Communion of the Apos- tles. cites an altar of the seventh century which not only has the same shape but also has on its top the places for six loaves of bread. The scene on the Milan ivory is, like the mosaic scene, a rendering of the Last Supper in its Eucharistic, symbolic sense. Thus the scene on the Milan ivory, while it does not aid in localizing the covers in Provence, sustains their Western origin and points to their origin in a community similar to Ravenna, as was Marseilles, where Western symbolism was breaking down under Eastern influences. The Western character of the transitional type is mainly shown by the retention of the old catacomb symbolism of the loaves and fishes and the resultant symbolic significance of the type which contrasts with the more 20 Fleury, La Messc, I, p. 164, pi. LIT. EARLY CHRISTIAN ICONOGRAPHY 139 purely narrative rendering of the historical type in the East. This symbolic character is more manifest on the Milan covers where the apostles are reduced to three. The parallels for this abbreviation of the number of the apostles are all on monuments of the West which show in other scenes an odd eclectic mixture of Eastern and Western forms. The liturgical type scarcely enters into a discussion of the Last Supper as a scene in early Christian art. 27 Tt is hardly correct to say that it repre- sents the Last Supper at all, for Christ is represented as a priest in the First Communion with His apostles, and the scene is only related to the historical event by reason of the fact that the Last Supper was the occasion of the First Communion. As a type, then, it was developed in a purely Fig. 130. Rossano: Gospels. The Communion of the Apostles. dogmatic sense and was related solely to the Eucharist. As the Multiplica- tion of the Loaves and Fishes was the symbolic method of depicting the Eucharist followed in the West, this more realistic form where Christ, as the celebrant, is shown administering the bread and wine to His disciples, was the Oriental manner of representing the sacrament. The liturgical type appears in the sixth century in the Rabula Gos- pels 2 ’* (I ig. 129). In a miniature of this manuscript the Saviour stands on a small mound holding in His left hand the chalice and in His right hand the wafer, both of which He offers to the eleven apostles who stand before Him. In the same century it occurs along with the historical repre- E. Dobbert, l, her die Darstellung dcs Abciidmahls durch die bysantiuische Kunst, is, 2, p. 22\ Baumstark, Rom. Quart., 1905, pp. 206-207: Heisenberg, Graheskirche und Apostetkirche, p. 175; Leclercq, op. cit.; Dalton, op. cit., p. 666. 25 Garr., op. cit., Ill, pi. 137/2. 140 EARLY CHRISTIAN ICONOGRAPHY sentation of the Last Supper in the Codex Rossanensis 29 (Lig. 130). After Byzantine art became fully formulated this subject was a favorite theme for the decoration of apses and occasionally of domes, as in the monasteries of Vatopedi and Chilandari at Mount Athos. 30 The fully developed form of the type is found in the eleventh century in the mosaics of the cathedral of Kieff. 31 In the centre is an altar standing before a ciborium on each side of which is an angel serving as sub-deacon to our Lord. Christ is twice represented, on the right offering a cup to six apostles and on the left presenting the bread to six more apostles. Much the same rendering occurs in the eleventh century in the mosaic of the Cathe- dral of Serres 32 in Macedonia and in the frescoes of the churches of Cappa- docia and Lycia. 33 Later in southern Italy, as a result of Byzantine influence in the West, the type appears in the frescoes of S. Angelo in Lormis and in the chapel of S. Angelo at Monte Raparo (Basilicata). 34 The historical type, which is an actual representation of the Last Supper as an historical event without any manifest importance at- tached to the symbolic significance of the scene, was created in the East and developed as a Byzantine scene which soon after its inception in the sixth century passed into the art of the West. The various changes that the type went through in its development have been most carefully studied and clearly presented by Dobbert and it suffices for our purpose to note that it appears to have originated in the sixth century on proto- Byzantine monuments of Asia Minor. In a miniature of the Rossano Gospels 35 (Lig. 1 3 1 ) Christ with His twelve disciples is represented re- clining at the left hand end of a couch of semi-circular shape with the chord toward the spectator; near the middle of the disciples, who are arranged about on the bolster, lies Judas stretching forth his hand to grasp the food on the table. The fact that the liturgical type occurs in the Rossano Gospels along with this historical representation shows that there was no particular importance laid on the actual Last Supper as a 29 Haseloff, Codex Rossanensis, pis. VI-VII. 30 Dalton, op. cit., p. 667. 31 Ainaloff and Riedin, Ancient Monuments of Art in Kieff, figs. 8-11; Dalton, op. cit., pp. 395, 666. 32 Dalton, op. cit., fig. 423. 33 Rott, Kleinasiatische Denkmdler, pp. 138, 144, 338. 34 E. Bertaux, L’ art dans I’ltalie mcridionale, p. 123. 35 Gebhardt und Harnack, op. cit., pi. VIII. EARLY CHRISTIAN ICONOGRAPHY 141 symbol, at least as a symbol of the Eucharist. The scene of the Last Supper among the Biblical representations on the walls of St. Sergius at Gaza, built by Justinian, as it is described by the chronicler Chorikios, was very like the miniature scene of the Rossano Gospels. 30 With few changes this rendering of the Rossano Gospels was adopted as the Byzan- tine type; by the ninth century the added feature of John leaning on the Master's breast made the scene even more narrative in character. 37 XVI THE PRESENTATION OF THE CROWNS The scene on the Milan ivory (Fig. 132) of Christ, young and beard- less. seated upon a starry globe and blessing with outstretched and veiled hands the crowns, which an apostle, saint, or martyr presents to Him on either side, appears to have been an Italian motif used only in the West. 1 It is impossible to interpret the scene as the Multiplication of the Loaves and Fishes, as Dobbert suggests, 2 from the fact that the figures on either side of Christ present neither bread nor fish for His blessing, but instead carry wreaths. These two figures presenting their wreaths to Christ are not Peter and Paul, as many writers consider them to be, 3 but are unknown Chris- 39 Dobbert, Rcpertorium (1891) XIV, p. 198; Choricii Gazaci Orationes, fragmenta, ed. Boissonade, pp. 91-98.- 37 The first representation of John leaning upon the breast of the Master is said to occur in the eighth century at Ferentillo (Dobbert, op. cit., XIV, p. 182). 3 Garr., op. cit., VI, pi. 455. 2 Dobbert, Rcpertorium fiir Kunstwisscnschaft, VIII, 1885, p. 172. 3 Bugati, Mem. di S. Celso, p. 275; Westwood, Fictile Ivories, p. 41; Molinier, Hist, gen des arts appliques, I, Ivoires, p. 61 ; Stuhlfauth, Elfenbeinplastik, p. 68, calls the scene “die Huldigung an den triumphierenden Christus.’’ EARLY CHRISTIAN ICONOGRAPHY 142 tian martyrs. The two figures on the ivory are both beardless, and at the time the Milan covers were executed, c. 500, the types of the two apostles had been firmly established in the Western art. Paul was repre- sented with a long, pointed beard and a bald forehead and Peter was always depicted with a short, stubby beard and the Roman tonsure. This Fig. 132. Milan: Catheidral, ivory book covers. Presentation of the Crowns. convention regarding the two Saints in the “Traditio Legis,” which the scene on the ivory would represent if we were to consider the figures to be Peter and Paul, was generally preserved on all the monuments of Rome, 4 Ravenna, 5 and Gaul. 6 While it is impossible to interpret the figures as Peter and Paul, it is certain from the wreaths which they carry that they are Christian martyrs as Garrucci 7 and Venturi 8 assert. There is no reason, however, to assume with Garrucci that the martyrs are the Milanese Saints Gervasius and Protasius. The fact that it was the custom throughout the West to represent local martyrs, as well as apostles, offer- ing crowns or wreaths to the Saviour makes any specific interpretation of their names a pure speculation. 9 Christ enthroned upon a globe, which is the unique feature of the composition, was, as far as existing monuments give any evidence, an 4 Sarcophagi: Garr., op. cit., V, pis. 327/2, 33o/S, 334/V 335/4, 341/2. Mosaics: S. Andrea (Garr., op. cit., IV, pi. 240/1) and S. Agata in Subura (op. cit., pi. 240/2). 5 Sarcophagi : Garr., op. cit., V, pis. 332/2, 336/4, 345/i, 346/2. 0 Sarcophagi; S. Maximin (Garr. 334/3), Arles (335/2), Arles (342/3), Marseilles (342/1). 7 Garr., op. cit., VI, p. 82. 8 Venturi, Storia, I, p. 510. 0 Sarcophagi ; Lerins (Garr., op. cit., V, pi. 329/2), Ravenna (pi. 346/2, 349/2, 3). EARLY CHRISTIAN ICONOGRAPHY 143 Italian motif. While the globe on the Milan covers is its only occurrence on the early Christian ivories, it was frequently used in the West, before the iconoclastic controversies, for the scene of the “Traditio Legis” and of the blessing of the crowns of Saints. Christ seated upon a globe has not remained to us on any early monuments of the East but, as may be seen from the following list, it occurs in Rome as early as the middle of the Fig. 133. Rome: S. Costanza, mosaic. Christ as Logos, giving the Law to Moses. fourth century (Fig. 133) and in many Western monuments of the fifth, sixth, and seventh centuries : 1) Santa Costanza, Rome, 10 IY century; mosaic. 2) S. Agata in Subura, Rome, 11 V century ; mosaic. 3) Milan book cover, Provence, cir. 500. 4) San Yitale, Ravenna. 12 c. 530-547; mosaic. 5) Sts. Felix and Adauctus, Rome, 13 VI century (middle) ; fresco. 6) San Lorenzo, Rome, 14 578-590; mosaic. 7) Cathedral, Parenzo. 15 VI century; mosaic. 8) San Teodoro, Rome, 10 YII century; mosaic. It is possible that the motif of Christ seated on a globe originated in the East, for Christ standing on the globe occurs not only in a fourth century fresco in the catacomb of Priscilla at Rome, 17 but also in the fifth century in the mosaics of the Baptistery of Naples which were executed under 10 Garr., IV, pi. 207; Ainaloff in the Journal of Public Instruction, St. Petersburg, 1895. p. 267. sees in this scene a strong resemblance to the scene in San Vitale. 11 Garr., op. cit., IV, pi. 240/2. 12 Garr., op. cit., IV, pi. 258. 13 Marucchi. Nuovo Bullettino di arch, crist., X, 1904, p. 135, pi. V. 14 De Rossi, Musaici cristiani di Roma; Diehl, op. cit., p. 322; Alinari, no. 21249. 15 Dalton, Bye. Art and Archaeology, p. 373. 15 Garr., op. cit., V, pi. 252/3. 17 De Rossi, Bullettino, 1887, p. 25, pi. VII. EARLY CHRISTIAN ICONOGRAPHY 144 Syrian influence. 18 The “Traditio Legis” in which it first appears is prob- ably also of Syrian origin. 19 Nevertheless, the monuments listed above show that by the fifth century the type was well known in Rome and in the north of Italy. In much the same rendering as that on the Milan ivory the mosaic of San Vitale shows Christ offering a crown to San Vitale. The scene in San Lorenzo at Rome represents St. Hippolytus offering his crown to the Saviour, and the mosaic of San Teodoro depicts two saints with crowns being presented to Christ by Peter and Paul. As martyrs offering their wreaths to Christ were customarily represented on the monuments of Rome, Ravenna, and even of Narbonne, 20 and Christ seated on a globe and either blessing or presenting crowns to saints was figured on the mosaics of Rome and Ravenna before the Milan covers were executed, the appearance of this scene on the ivory is indicative of Western origin. XVII THE WIDOW'S MITE A highly problematic scene on the Milan ivory shows Christ seated upon a globe and stretching forth His hand to a woman who appears to be dropping something on a table which is between them (I’ig. 134) ; be- hind the woman are figured two men who express by their gestures great surprise at the event. 1 Of all the various subjects which have been suggested, the most plausi- ble is Christ and the Widow’s Mite. That it does not represent Christ and the Samaritan Woman, as Molinier asserts, 2 is manifest from the absence of the well and the presence of a table between Christ and the woman. If anything, it is less likely that it represents the Woman Taken in Adultery, for not only was the scene very uncommon in the whole history of Christian art, but also the appearance of the table on the Milan 18 Bertaux, L’art dans I’ltalie meridionale, fig. 71 Garr., IV, pi. 269; Diehl, Manuel, p. 1 17. 19 Baumstark, Oriens Christianas, III, 1903, pp. 173-200; Byz. Zeit., XIII, 1904, p. 661. 20 Le Blant, Les sarcophages chretiens de la Gaule, pi. XLV/2. 1 Garr., op. cit., VI, pi. 455 - 2 Molinier, Hist. gen. dcs arts appliques, I, Ivoires, p. 61. EARLY CHRISTIAN ICONOGRAPHY 145 ivory can not be explained or paralleled in any other examples. 3 Both of these interpretations are based on the fact that in the eleventh century there appear in the frescoes of S. Angelo in Formis representations Fig. 134. Milan : Cathedral, ivory book covers. The Widow’s Mite. of these two scenes, in both of which Christ, most oddly, sits upon a globe. 4 As the Widow’s Hite the scene on the Milan book covers 1 is similar, save for the globe, to the rendering of the same scene in a mosaic of S. Apollinare Nuovo at Ravenna (Fig. 135). In this sixth century mosaic the widow stands at the left of the composition, her hand extended over a table, which resembles the table on the ivory, and faces Christ, who is accompanied by a single disciple. 6 With the exception of the globe, both representations are noticeably alike and accurately illustrate the Gospel account. 7 It is also interesting to note that about the middle of the fifth century Perpetuus recorded that among other Biblical scenes the Widow’s Mite was figured on the walls of the church of St. Martin at Tours. 8 Whatever the correct interpretation of the scene may be, we have :: Examples of this scene occur at S. Apollinare Nuovo (Fleury, L’£vangile, pi. LYIII/2), and in the Homilies of Gregory of Nazianzus, gr. 510, Bibl. Nat., Paris (Fleury, op. cif., pi. LYIII/3). * Bertaux, L’art dans J’ltaUe mcridionale, fig. 99. This interpretation is that of Westwood (Fictile Ivories, p. 41), Stuhlfauth, Elfen- beinplastik, and Dobbert, Jb. Preuss. Kunsts., 1894, p. 136. c Garr., op. cit., IY, pi. 248/5. 'Mark xii, 42; Luke xxi, 1-4. ' Kraus, Gcscliichte dcr christlichen Kunst, I, p. 602 sq ; Perpetuus, “Tituli der Basilica des Heil. Martin zu Tours” in Von Schlosser, Quellenbuch cur Kunstgeschichte (Neue Folge, 4), p. 32. 146 EARLY CHRISTIAN ICONOGRAPHY Fig. 135. Ravenna: S. Apollinare Nuovo, mosaic. The Widow’s Mite. already indicated the Italian character of the globe as a seat for Christ . 0 Although the only other examples of its use in miracle scenes in the eleventh century are the frescoes of S. Angelo in Formis, its frequent use in other connections from the fourth to the eighth century in Italy and its continuation in Carolingian art leaves little doubt as to the West- ern origin of the scene on the Milan book covers. XVIII CHRIST TRIUMPHANT The representation of Christ trampling under foot the lion, the dragon, the asp and the basilisk as depicted on the Bodleian book cover, the ivory so frequently cited in connection with the preceding scenes, presents a composition which not only may be connected with early Caro- lingian art in France but also may be traced back to a prototype of Egyptian origin. This transmission of Egyptian types into early French art is neither surprising nor peculiar to this particular scene. In the first place it was the natural continuation, if not the actual result, of that Eastern influence which entered Provence through Marseilles and spread through Gaul in Carolingian times. The importance of the role that Egyp- tian models played in the development of Carolingian art at that time when the art of the North was emerging under Christian tutelage from barbar- ism has been shown in various aspects. The Syro-Egyptian influence on Merovingian and Carolingian miniatures , 1 the presence of Egyptian art D See p. 142. 1 The influence of Egypt and Syria on Merovingian art has already been established EARLY CHRISTIAN ICONOGRAPHY 147 objects, like the Barberini diptych , 2 in Frankish cities at an early date, and the dependence of certain reliefs at Aachen on Egyptian or Alexandrian models make this relation manifest. The most interesting parallel for the present problem is the connection of the scenes, and their technique, on the ivory pulpit of Charlemagne at Aachen with Coptic and Alexan- Fig. 136. Alexandria : Catacombs, fresco. Christ Triumphant. drian prototypes . 3 Among these early Frankish scenes, the representation of a mounted horseman lancing an animal upon the ground has been shown, by Strzygowski , 4 to have been taken from a Christian type of Egypt which in its turn was derived from the mounted Horus in Egyptian and a brief summary, with a good bibliography', of the subject may be found in Michel's Histoire de I’art chretien, vol. I. pp. 395-400. For the dependence of Carolingian and Ottoman illumination on Syro-Egyptian models see : Strzygowski, Hcllenistische und Koptische Kunst, and Byzantinische Denkvtaler, I, p. 58, Michel’s Histoire de I'art chretien, I, pp. 340, 400 sq., HaselofT, Der Psalter Erzbiscliof Egberts von Trier, p. 133, cf. Byz. Zeit., XI, p. 566, and Janitschek, Die Trierer Ada-Handsclirift, p. 85. 2 Strzygowski, Hell, und Kopt. Kunst, p. 68. 3 Strzygowski, op. cit., p. 21 sq. 4 Strzygowski, op. cit., fig. 16. 148 EARLY CHRISTIAN ICONOGRAPHY art. By the same stages the scene of Christ standing triumphant on the evil beasts came to Carolingian art directly from Egypt where it was also developed from the representations of the youthful Horus trampling under foot the beast which, in Nilotic mythology, symbolized the evil and malignant powers. This relation between the Christian idea of Christ Triumphant over evil and the old Egyptian conception of Horus was presented by Neroutsos- Bey when he published an account of the Christian frescoes remaining in a funeral chapel of Alexandria. 5 The usual type of Horus on the Egyptian reliefs, as this author points out, shows the young god standing on the heads of two crocodiles while in his hands he carries serpents, scorpions, a lizard, and a lion. 6 Often there are figured on either side of him other divinities who aid in mastering the evil beasts and above his head is always portrayed the masque of the great protector, Bes. With this common Egyptian representation of Horus is to be compared the scene in the catacombs of Alexandria (Fig. 136) where the young Christ, beardless and nimbed, tramples under His bare feet the lion and dragon while at His side are the asp and the basilisk. 7 The two apostles (omitted in Fig. 136) on either side of Him take the place of the Egyptian divinities who frequently accompany Horus, while the nimbed head of God the Father (omitted in my illustration) recalls the masque of Bes, the pro- tector. Beside the apostles, two angels, one on either side, decorate the jambs which terminate the picture. The analogy between the Christian and the pagan scenes is more than that of a similarity in iconography. It is be- tween almost identical conceptions: the Horus relief (Fig. 137) depicts the formula, carved upon it, '‘trample under foot the crocodiles and master without effort the lion, serpents, etc.” ; and the Christian scene, as the in- scription below it shows, illustrates the words of the Psalmist, “Thou shalt tread upon the asp and the basilisk: the lion and the dragon shalt thou trample under feet.” 8 Other writers have pointed out this relation between the two repre- sentations and have admitted the Egyptian origin of the Christian scene. 9 5 Neroutsos-Bey, L’ancienne Alexandrie, Paris, 1888, p. 45 sq. 6 Op. cit., p. 46. 7 Op. cit., p. 49. 8 Op. cit., p. 48. Super aspideni et basiliscum ambulabis, ct conculcabis leoneni et draconcm (Ps. XCI, 13. Vulgate). 9 Lmile Male, Comptes rendus du Congres international d’archeologie classique, Cairo* 1909, p. 270 ; Leclercq in Cabrol’s Dictionnaire d’archeologie chreiienne et de liturgie, s. v. “Basilic”; Laurent, Les ivoires pregothiques, p. 40 sq. EARLY CHRISTIAN ICONOGRAPHY 149 In fact, before the Alexandrian fresco had been published, Reveillout had noted the striking resemblance which existed between Christ in Coptic literature and Horus in the Egyptian tradition . 10 The Egyptian origin of the scene would be more than certain if the third century date to which Neroutsos-Bey ascribed the Alexandrian frescoes seemed possible . 11 Un- luckily the general character of the frescoes, the inscriptions, and the iconography that includes the Virgin, impresses one with the idea that the frescoes could not have been as early as the third century or, possibly, even the fifth century which marks the limit set by Neroutsos-Bey for their execution and repainting. The connection of the type with Egypt, however, does not depend solely either upon the date of this single fresco or upon the Horus analogy. The same iconography of the scene is pre- served on a large number of various objects, of which the majority 10 Reveillout, Rez ue Egyptologique, 1881, II, p. 65. 11 Xeroutsos-Bey, op. cit., p. 51. EARLY CHRISTIAN ICONOGRAPHY 150 are known to have come from Africa and the rest may be shown to have the same origin. The scene occurs most frequently on the early Christian lamps from Egypt. On one lamp, found on the Palatine and therefore attributed by De Rossi to Rome , 12 the Saviour, beardless and with a cruciform nimbus, is represented treading under foot a serpent on whose head He rests the end of a long handled cross ; on either side of the Saviour’s head is an Fig. 138. Strassburg, Forrer Collection : Christian lamp from Akmim. Christ Tri- umphant. angel, at His left another serpent, to His right a dragon, and beneath His feet a lion. This same type of scene with only the most minor variations occurs on a lamp in the Brills Collection at Rome 13 and on another at Madrid which is said to have come from Athens . 14 Besides these examples of apparently doubtful origin the same composition which appears on the ] -Dc Rossi, Bull, di arch, crist., 1867, p. 12. 13 Garr., Storia, VI, pi. 473/4. 14 Garrucci, /. c. p. 109. EARLY CHRISTIAN ICONOGRAPHY 151 Palatine lamp occurs also on three lamps from Africa, viz., a fragment from Carthage, 15 a lamp from Akmim, 16 and another from Bagai'. 17 The like- ness between the Palatine example and the one from Bagai, in Upper Egypt, is extremely close. Further evidence for an Egyptian origin of the three lamps which are not known to have come from Africa is afforded by the Greek inscription on the Brills lamp, and the ornament and material of the Palatine lamp, which show, as Grisar says, 18 that it could have been executed only in Egypt or the Holy Land. With an almost certain Egyp- tian origin for all these lamps, the relation of the scene of Christ Tri- umphant with the same scene in the Alexandrian catacombs is obviously intimate, for the similarity extends even to the angels who fly toward Christ on either side (Fig. 138). This developed type where Christ tramples on the four beasts was abbreviated in Egypt into a second type, seen also on the lamps, where Fig. 139. Strassburg, Forrer Collec- tion: Christian lamp from Akmim. Christ Triumphant. the Saviour stands upon a single serpent. On a lamp from Akmim (big. 139) Christ is depicted standing on a huge serpent on whose body He rests the end of a long cross that He carries in His right hand. 19 The same rendering, with only a few slight variations, occurs on three lamps from Carthage. 20 and on the cover of an earthen vessel in the Cairo 15 Revue de l’art chretien, 1893, p. 37, no. 903. lf; Forrer, Die friihchrisllichen Alterthiimcr aus deni Gr'dberfclde von Achmim-Panopo- lis, pi. IV, no. 3. 17 A. Heron de Villefosse, Lampcs chretiennes inedites in Le Musce archcologique, 1876. I, pp. 113-117; Leclercq in Cabrol’s Dictionnaire, s. v. “Basilic,” col. 511, note 1. 1 - Grisar, Roma alia fine del mondo antico, 1908, p. 619, fig. 181. 19 Forrer, op. cit., pi. IV/2. 20 Dclattre, Musees de I’Algcric ct dc la Tunisia, III, pis. VIII/i, IX/2; Dalton, Cata- logue of the Early Christian Antiquities in the British Museum, 1901, no. 721. EARLY CHRISTIAN ICONOGRAPHY 152 Museum.- 1 The scene on a textile from Akmim is of the same order, although in this case the Saviour carries a small cross in His left hand and with His right hand drives a lance into the mouth of the dragon at His feet. 22 This second abbreviated type passed from Egypt to Ravenna and occurs there for the first time among the wall mosaics of Sant’ Apollinare Nuovo. 23 In the lunette above the palace portal which is represented in mosaic on the walls of the nave, Christ, carrying a cross in His right hand and a book in His left hand, stands on the head of a long serpent; on either side of Him is a single apostle. This early sixth century representation is remarkably close to the Coptic examples. It is to be noted that the two apostles who appear in the Ravenna mosaic also occur in the Alexandrian fresco. In Merovingian times the type of Christ standing on the single serpent appears to have been introduced, not necessarily through Ravenna, into Gaul. In a cemetery of Orleans (Fig. 140) there was discovered Fig. 140. Fragment of terra cotta vase DISCOVERED AT ORLEANS. CHRIST TRIUMPHANT. a fragment of vase on which is represented the figure of Christ carrying a cross and standing on the head of a serpent. 24 The composition of the scene is curiously analogous to the examples from Akmim and Carthage. 21 Strzygowski, Koptische Kunst, p. 248, no. 7142. 22 Forrer, op. cit., pi. XVIII/i. 23 Alinari, no. 18238; Julius Kurth, Die Mosaiken der christlichen Ara, p. 177 , men- tions the scene but apparently fails to understand its meaning; he writes that three figures are represented having an adventure with a snake, a reptile which was frequently seen in the woods around Ravenna. 24 Garr., op. cit., VI, pi. 466/2; Revue archeologique, XV, 1867, p. 376. EARLY CHRISTIAN ICONOGRAPHY 153 Christ on the vase wears an Egyptian coiffure and a Gallic sagum 25 and carries a kind of cross similar to those on the African lamps, while the serpent coils around the feet of the Saviour in a manner identical with that of the Coptic representations. Many of the monks, hermits, and ecclesiastics of Gaul were trained in the monasteries of Upper Egypt and the monastic system of the country was at first based on the Egyptian rule, so that it is not in the least surprising to find this Egyptian type in a Gallic town. In fact, it offers a parallel for the later introduction into Gaul of the developed type of the scene such as was seen in the Alexandrian fresco. Before turning to the introduction of the developed type into Caro- lingian art another variation in the manner of representing Christ Tri- umphant over the beasts should be considered. Instead of depicting the Saviour standing on the four beasts or upon the single serpent, this method, which is first seen in Ravenna, shows Him, cross in hand, trampling upon the necks of a lion and a dragon. This type, if it was not brought to Ravenna from Egypt as the other form of the scene must have been, was Fig. 141. Ravenna: S. Francesco, Pignatta SARCOPHAGUS, DETAIL. CHRIST TRIUMPHANT. surely developed there under direct Egyptian influence for, after all, there is little difference between it and the example of the scene in the Alexan- drian fresco. Christ trampling on the lion and the dragon occurs in the fifth century at Ravenna in a stucco relief of the Catholic Baptistery where Christ curiously enough is represented wearing an Egyptian coiffure. 20 Of the same order is the scene on the Pignatta sarcophagus (Fig. 141) 2 ' Leclercq, of. cit., col. 513. 26 Garr., op. cit., VI, pi. 406/4. From Dr. Shapley, of Brown University, who is making a study of the monuments of Ravenna, I learn that Christ carries a cross in Flis right hand. *54 EARLY CHRISTIAN ICONOGRAPHY at Ravenna where Christ, with an apostle on either side, is seated upon a raised throne resting His feet upon the heads of a lion and a dragon. 27 In Merovingian times the type is found on a wooden pail from a sepulchre at Miannay, near Abbeville. 28 So far, the review of the early examples of Christ trampling on the beasts has shown that the monuments on which the scene occurs were of Egyptian or Western origin. In other words, the scene, having originated in Egypt, did not pass into Byzantine art, but at an early date was trans- mitted to Ravenna and later to Gaul. When the representations of the scene which were executed after the seventh century are considered it will be apparent that the idea of depicting Christ treading on the evil beasts was also transmitted to Carolingian art and was characteristic of Frankish monuments. It is in regard to the use of the scene in Carolingian art that the question as to the date and execution of the same scene on the Bodleian book cover arises. A summary, however, of the late examples of the scene, which are all on either Carolingian, Frankish, or Ottoman monuments, prepares the way for showing the connection between the scene on the Bodleian cover and the early Carolingian representations. Also when the Carolingian renderings are compared with the representation in the catacombs of Alexandria it becomes more evident that the scene, as a type, must have been introduced directly from Egypt. In the eighth or ninth century the scene of Christ trampling on the four beasts occurs on a Carolingian ivory book cover in the Musee du Cinquantenaire at Brussels 29 and on the cover of the Lorsch Gospels (Fig. 142) in the Vatican. 30 Although this last example has been variously attributed to Byzantine and Italian as well as to Carolingian art, it is now admitted to be of Frankish origin. 31 Besides these two ivory examples, a representation of Christ flanked by two angels and treading on the lion and the dragon appears in a minia- ture of the Gospels of the Duke of Arenberg (Brussels), a tenth century Anglo-Saxon manuscript. 32 In the eleventh century the same scene occurs 27 Garr., op. cit., V, pi. 344; cf. Diitschke, Ravennatische Studicn, no. 68, p. 246. 28 Garr., op. cit., VI, pi. 461/4. 29 Laurent, Les ivoircs pregothiques, pi. II. 30 Garr., op. cit., VI, pi. 457. 31 The latest commentator on the ivory, Goldschmidt, places it in a group connected with the atelier of the Ada Gospels ( Jb . Preuss. Kunsts., XXVI, 1905, 64, no. 31). 32 Beissel, Geschichte dcr Evangelienbiicher, Freiburg i/n, 1906, p. 133, fig. 35. EARLY CHRISTIAN ICONOGRAPHY : 55 Fig. 142. Rome: Vatican, ivory book COVER FROM LoRSCH. CHRIST TRIUMPHANT. on an ivory from the lower Rhine , 33 on the metal cover of the Gospels of Poussay , 34 and on the reliquary of Saint Hadelin at Vise . 35 In the twelfth century there is a representation in a missal of Hildesheim 36 and a French ivory in the South Kensington Museum depicts Christ trampling on the four beasts . 37 Although it is not possible that this list of ex- amples of the scene is complete, it is singular that not a single late representation has been found either on an unquestioned Byzantine or an Italian monument and that the scene should have been so faithfully pre- served in those regions whose art emanated from the Carolingian schools of miniature painters, who, as we know, drew largely from Oriental sources. Turning to the representation on the Bodleian book cover (see Fig. 3 Graeven, Elfenbeinzuerke, series I, no. 30, in the Bargello at Florence. 34 Laurent, op. cit., p. 44: Sauerland and Haseloff, Der Psalter Erzbiscltof Egberts von Trier, 1901, pi. 32. 3 ~ Laurent, op. cit., pi. V. 36 Beissel, op. cit., p. 322, note 1 ; Loubier, Der Bucheinband, fig. 51. 37 Westwood, Fictile Ivories, p. 165, no. 260 ’67. 156 EARLY CHRISTIAN ICONOGRAPHY 169) we see within the central panel the figure of Christ depicted before a loggia. The Saviour is young, beardless, with long hair falling on His shoulders, and wears a cruciform nimbus. In His right hand He holds a cross which rests on His shoulder; His left hand bears an open book. With the right foot He treads upon the lion and with the Fig. 143. Brussels: Musee du Cinquan- TENAIRE, IVORY BOOK COVER FROM GeNOELS- Elderen. Christ Triumphant. left upon the dragon while beneath the lion is the asp and beneath the dragon is the basilisk. On one side of the open book is inscribed the monogram IHS XPS and on the other sheet are the first two words of the thirteenth verse of the ninety-first Psalm, , SUP (er j ASP (idem) If this rendering is compared with the same scene on the Carolingian diptych of Genoels-Elderen at Brussels (Fig. 143) the similarity is ap- parent. In the first place both scenes are set within columns and the Saviour in each case shows the same traits and characteristics. More- EARLY CHRISTIAN ICONOGRAPHY 157 over, the Saviour on both ivories carries in identical fashion the same kind of a cross while He treads on the four beasts which are arranged in the same positions. In fact, the renderings are remarkably alike with the exception that Christ on the Brussels cover is flanked by two angels and that there is far greater animation, not only in the attitude but also in the draperies, of the Christ on the Bodleian cover. This particular type of youthful, long haired Christ, seen on the two ivories, was peculiar to Carolingian art and primarily to the Godescalc school of miniaturists. It is the resemblance of the Christ type on the Genoels-Elderen diptych to the traditional rendering of Christ in the Godescalc and Ada manuscripts which, with other points of semblance, led Laurent to associate so successfully the Brussels ivory with that artistic centre in northern Gaul where the miniatures of the Godescalc school were painted. 3S The ivory, he points out, preserves in the scene of Christ Triumphant a type which originated in Egypt. This dependence on Eastern types is also characteristic of the Godescalc miniatures, since these, in much of their ornament and iconography, were based upon Eastern models similar to the Rabula and Etschmiadzin Gospels which Oriental traders, monks, and ecclesiastics, as well as pilgrims to the Holy Land, had been bringing into Gaul ever since the fifth century. 39 There- fore if there exists so intimate a relation between the Genoels-Elderen diptych and the Godescalc school as there is an apparent relation between the Genoels-Elderen diptych and the Bodleian cover, there should be mani- fest certain analogies between the Bodleian cover and the Godescalc school. Comparing the Christ on the Bodleian cover with the Enthroned Christ on the Godescalc Gospels, 40 which were finished shortly after 781, it is evident that the type of Christ is the same. Notice should be taken of the fact that in the Godescalc miniature, behind the head of Christ, ap- pears the same monogram IHS NPS which is inscribed on the book which Christ holds on the Bodleian cover. Carrying the comparison with the Godescalc school still further, the Christ on the ivory is very much like the same figure in the Soissons Gospels. 41 In this early ninth century 3S Laurent, op. cit., p. 39. .see Orientalizing of Gaul, Part II, section II, of this volume. 4 Michel, Histoirc dc I’art chrctien, I, fig. 166. 4 Laurent, op. cit , pi. I\ . This similarity of the Christ on the Bodleian ivory with the type of the Godescalc manuscripts is shown by another miniature in the Soissons Gospels CR. de Lasteyrie, L’architccture religieuse en France a I’epoque romane, 1911, fig. 181). The similarity extends even to the cross which Christ carries in His right hand, the open hook in His left hand, and to the agitated folds of His draperies. EARLY CHRISTIAN ICONOGRAPHY 158 manuscript the drapery has that nervous, sharp edged character which is apparent on the Bodleian cover. Nervous drapery, which appears to have been ironed stiff at the moment when it was fluttering most violently, was one of the most char- acteristic features of many of the Carolingian and Ottonian schools of art. As the artists sought to give greater life and vigor to their figures, the animated drapery became more and more pronounced until in some of the late schools, as the Anglo-Saxon, the exaggeration became the principal characteristic of the style. This pronounced animation ap- pears in the scene of Christ Triumphant in the tenth century Anglo- Saxon Gospels of Arenberg. Although more animated than the drapery in the Godescalc Gospels themselves, the drapery of Christ on the Bodleian ivory is not as nervous as in this Anglo-Saxon example and is closest to the type of drapery seen in miniatures of the Godescalc and Ada manu- scripts. Besides the marked similarity with the Christ of the Soissons Gospels, the animated folds of Christ’s costume on the ivory recall the drapery of the Evangelist in the Codex Aureus (Harleian, 2788) of the British Museum, 42 which is another manuscript of the Godescalc group. It is significant that the same technique, though more fully developed and more beautifully executed, occurs on the Carolingian ivory in the Vatican which also represents Christ Triumphant. This diptych leaf came from the monastery of Lorsch to the Vatican as the cover of the Gospels of Lorsch, which also belongs to the Godescalc group of manuscripts. 43 It is remarkable that the three similar representations of Christ tram- pling on the lion and the dragon, the asp and the basilisk should all present the same type and, at the same time, should each one bear some resem- blance or relation to the miniatures of the Godescalc school. The scene on the Bodleian cover, irrespective of the other scenes, from this analogy with the Godescalc miniatures and its similarity to the Genoels-Elderen diptych, would appear to date the ivory at the end of the eighth century or at the beginning of the ninth century and would place its execution some- where in the north of France. Whether the cover was actually done at this time or was a prototype for the Carolingian scenes, as those who date the cover early must assume, is a question answered elsewhere in my discussion of the iconography and style of the rest of the Bodleian ivory (p. 254). 42 Humphreys and Jones, The Illuminated Books of the Middle Ages, pi. II. 43 Leprieur in Michel’s Histoire de I’ art, I, p. 338. EARLY CHRISTIAN ICONOGRAPHY *59 XIX SUMMARY OF THE SCHOOLS OF EARLY CHRISTIAN ART Although unable to assert that early Christian iconography was an exact science, governed by consistent rules of composition, even in the sense that Byzantine iconography of the eleventh century approached the exactitude of a science in its formulae of representation, I feel that the preceding pages have justified an iconographic classification of the early Christian schools of art. In the case of each Biblical scene that has been studied the same monuments, as the varying methods of representation on them were allowed to divide into groups, fell within approximately the same classification and logically grouped themselves around a few monuments of known provenience in the separate schools. Before closing, then, this study of early Christian methods of Biblical representation, it may prove valuable to gather together the results and in a brief summary to view the iconography of each individual school as a whole. The Hellenistic, the Oriental-Hellenistic, the Provencal, the Alexan- drian-Coptic, the Palestinian-Coptic, the Syrian and the proto-Byzantine schools, into which I have divided the Christian art of the first seven centuries of this era, are from necessity rather broadly inclusive and inca- pable of being exactly bounded and limited. They are in most cases not so much schools as sets of types which were popular in or emanated from cer- tain regions during certain centuries. Yet in such cases as the Provencal and the Alexandrian-Coptic they represent very specific schools and as schools, in which the area of artistic activity was somewhat limited and consistent, it is perhaps permissible to treat them all. The Hellenistic type, as characteristic of what we may loosely call a Hellenistic school, is by far the most difficult to define. Since it is ex- clusively the Hellenistic type that appears in the catacomb frescoes and on the sarcophagi of Rome, it is commonly recognized as a Western type and as evidence of a Western school. That, however, does not prove that it was of Western origin. When Christianity was born, an Eastern cult in an Eastern land, the civilized world was under classical domina- tion ; a neo-Greek veneer of manners, customs, literature, and art over- spread the whole Mediterranean world and the new faith, seeking a means of expression in literature and art. found at hand only Hellenistic 160 EARLY CHRISTIAN ICONOGRAPHY models and types. At once it found that the rational, material character of classical art was hardly suitable for the expression of its new ideals. The artist could not render in stone or paint a comprehensible expression of the omnipotent power, the everlasting love, and the infinite grace of the Father; neither could he so much as sketch the soul, which was fundamental to the Christian faith, nor adequately express the neces- sarily indefinite blessings of the hereafter where the weary soul would find peace. Even the spirituality of the Saviour evaded the concise and material limitations of the sculptor's chisel, while to portray Eternal Salva- tion, which above all others was the appealing promise that awoke the fettered imagination of a pagan world to new hope, the artist had at his command only the expressive forms of flesh and blood, of a rational and material world, from which his dominant idea was to escape. Realizing the value and the necessity of the pictorial image to stir the imagination of the uneducated and simple masses, the theologians turned to symbolism. As the theological writers, especially of the Neoplatonic school, adopted the Hellenistic figures of speech and even the classical philosophy and, by allegory and analogy, sought to harmonize them with the new faith, so the artists, under the direction of the theologians, accepted the models of Hellenistic art and, through a symbolism that was carefully simplified, explained, and stylized, portrayed the otherwise inexpressible ideals and yearnings of the new religion. Thus the young Orpheus, seated, harp in hand, amid his flock, was first taken as the type of Christ, the Good Shepherd; the more common Hellenistic forms of decoration, as putti playing in the branches of the grapevine, were given a concise Christian significance and used to decorate the Christian edifices; and the classical banquet scene was chosen to symbolize the joys of the Christian Paradise. In the great Hellenistic cities of the Mediter- ranean, largely in Alexandria, Antioch, and Ephesus, which were the earliest theological centres of the new faith, a highly symbolical art was evolved, based in its forms on classical models. As Christianity rapidly permeated the West, taking with it its newly acquired means of expression, the West, primarily concerned at the time with preserving and spreading the faith in the face of imperial opposi- tion, readily accepted the new symbols, which by their simplicity and classical character could attract little attention from antagonistic pagan- EARLY CHRISTIAN ICONOGRAPHY 161 ism and could silently preach the Christian sermons to the masses of uneducated but devout converts. The Hellenistic type developed in the West; while it gradually abandoned the more obvious Hellenistic models and was constantly enriched, nevertheless, when Christianity at the beginning of the fourth century emerged from the catacombs to be- come a state religion, the Hellenistic form retained its highly symbolical character, remaining essentially simple in its composition, so as not to obscure the spiritual lesson by a too realistic rendering of the material event, and preserving also in a crude way the feeling for classical forms and drapery. That time, wars, fires, and earthquakes, ravaging the great Hellenistic cities of the Orient, have completely destroyed nearly all the monuments of this early phase of Christian art in the East and that only in the Eternal City have the Hellenistic types been luckily preserved in frescoes and mosaics and on sarcophagi, has led to the old assumption that Hel- lenistic Christian art originated in Rome. Today the balance of judg- ment, regardless of the meagre evidence that remains, is inclining strongly toward the Orient. In reality Rome loses little ; she adopted the Hellen- istic types of Christian art and made them hers; she changed, enriched, and formalized them and hence made of them the products of a Roman school. At Rome the Hellenistic school flourished long after new fashions, Oriental rather than Hellenistic, had grown up in the East and were work- ing westward again to supplant in time the school that for our purposes we may now call Western. Hellenistic Types The symbolic, dogmatic, and simple character of this school will appear in a review of the salient features of the iconographic types that have been studied in this volume and, of those scenes that appear in the frescoes and on the sarcophagi of Rome, the following are among the most common and significant : The Nativity . 1 In some ways a misnomer, the Nativity in the Hellenistic type was not so much a representation of the Birth as it was a pictorial sermon confirming the universal worship that was accorded to the Saviour by the Jews, Gentiles, and natural world. Five groups appear within this 1 See p. 14. EARLY CHRISTIAN ICONOGRAPHY 162 type: (1) the Adoration of the Ox and the Ass occurs, without any other elements in the composition, as symbolic of the worship of the natural world; (2) the shepherds are introduced into the previous composition and their theological significance, as typifying the Adoration of the Jews, is shown by the fact that the Child lies either upon the ground or on an altar; (3) the manger or crib is added and the scene is figured in the traditional Western shed with open sides and front; (4) the Virgin, sitting apart from the Child who lies in His crib, is introduced; (5) the three Magi, adoring the Child as He lies in His manger, or crib, are figured in the scene. The Adoration of the Magi. 2 The Magi in this type are beardless and, with the exception of a few early examples in the catacombs where they are arranged symmetrically on either side of the Virgin, they advance with their gifts in a line; the young Jesus, a child of two years, sits in the lap of the Virgin who is figured either in a three-quarters posture or in profile; as a rule Joseph is omitted. The Massacre of the Innocents , 3 The only example of this scene in the Hellenistic school occurs in the arch mosaics of Santa Maria Maggiore. The classical restraint of this period of art, when symbolism rather than realism governed the repre- sentation of scenes and prohibited the rendering of passion, is manifest in this representation ; instead of depicting the actual massacre, as was the custom in the East, the Hellenistic artist represented the children as hopeful martyrs awaiting execution in their mothers' arms. The Baptism . 4 Essentially symbolic with the same significance in all communities, the Baptism shows a marked uniformity in all schools. In the Hellenistic type, however, Jesus is always beardless, is represented as a small child, and is never immersed in the water more than to His knees; John the Baptist is dressed in an exomis or perizoma in the catacombs and on the sarcophagi generally wears the pallium instead of the shepherd’s garment; while not always depicted as such, the Jordan as a waterfall behind the Saviour is characteristic of this type. 2 See p. 38. 3 See p. 60. 4 See p. 73 - EARLY CHRISTIAN ICONOGRAPHY 163 The Miracle of Cana . 5 Christ, never accompanied by more than one disciple, touches with His wand the jars of water which vary in number from one to seven. During the fourth and fifth centuries, this scene gained great theological import- ance in the West and, in combination with the Miracle of the Loaves and Pishes, became a symbol of the Eucharistic transubstantiation. The Healing of the Blind . 6 While not common enough in either the East or the West to show any strikingly distinct characteristics, the Hellenistic type depicts a single blind man, small of stature, as compared to Christ, and sometimes carry- ing a staff. It represents the Saviour either touching the blind man on the eyes or laying a hand on his head. The Healing of the Paralytic . 7 This scene of deliverance symbolized the Baptism and although com- mon in the Hellenistic school it manifests no single and distinct type. In the catacomb frescoes the paralytic usually carries his bed upside down on his shoulders and Christ is omitted from the composition ; on the sarcophagi the paralytic is always small, and stands in profile at the feet of Christ holding his bed with the legs hanging down his back. The Raising of Lazarus . 8 The Hellenistic group, appearing on Western monuments, subdivides into five variants of the general type. With the exception of the first and fifth variants, the uniform characteristics of the Hellenistic type show Christ touching with His wand the figure of Lazarus who, wrapped like a mummy, his head covered with a cloth, stands in the doorway of a little gabled aedicula set upon a podium. The sub-groups are characterized as follows: (1) in a few of the earliest examples Lazarus sometimes lies on the steps of the aedicula and is seldom represented as a mummy; (2) this group includes the majority of examples and shows only the uniform characteristics of the Hellenistic type as a whole; (3) one of the sisters of Lazarus, kneeling at the feet of Christ, is added to the composition; 1 4 ) the second sister, standing at the side of the Saviour, is introduced ; (51 in a unique and highly symbolical type appearing only on the 5 See p. 85. e See p. 95 See p. 103. 8 See p. 109. 1 64 EARLY CHRISTIAN ICONOGRAPHY “Roman glass,” Lazarus, wrapped like a mummy with his head bare, leans against a kind of foliate growth. The Entry into Jerusalem . 9 The distinguishing feature of the Hellenistic type is the manner of representing Christ riding astride instead of sidewise. As in all the West- ern types, only the actual essentials of the scene are figured : as a rule the disciples who accompanied Christ are reduced to one; the city, and the multitude who came forth, bearing palms to greet the Saviour, are omitted; and the whole acclamation of Christ as King is symbolized by the single figure of a youth plucking palm branches in the top of a tree and by a man who spreads his mantle beneath the feet of the ass. The Last Supper . 19 It does not exist in the Hellenistic school. As there was no significance in the Last Supper to the early Christians save as it symbolized the Eucharistic Transubstantiation, which was commonly represented by the Multiplication of the Loaves and the Fishes, the scene does not occur in the West until Eastern influence gave it an historical value. ^ ^ ^ The Oriental-Hellenistic scenes do not mark the existence of a school. They represent only the first breaking down of the symbolic Hellenistic types, due to the gradual dissolution of Hellenistic influence in the East and the resultant tendency of the Orient, from the end of the fourth century, to assert in art its indigenous and racial characteristics. While this assertion of Oriental tendencies commenced at the close of the fourth century, it was not until the beginning of the sixth century that clearly distinguishable Eastern schools appeared in the Eastern provinces of the Roman Empire. In the interim between the end of the fourth century and the opening of the sixth century, the transition was gradual; no sharp lines of demarcation are apparent, for art still continued to follow Hellen- istic models. The first suggestion of approaching change appears in the Oriental cast that is given to what would otherwise be traditional Hellen- istic types. These types, slightly modified by compositional details, which are later to be characteristic of Eastern types, make up the group that I have called Oriental-Hellenistic. As the process of Orientalizing the Hellenistic Christian art in the East 9 See p. 123. in See p. 129. EARLY CHRISTIAN ICONOGRAPHY 165 developed more slowly in some localities and artistic forms than in others where the Hellenistic traditions persisted with less tenacity, the presence of Oriental-Hellenistic types offers no criterion for date or even for pro- venience. They only signify a development that was going on in Christian art under Eastern influences. In fact, certain monuments preserve these types as late as the seventh century, although as a rule the types are char- acteristic of the fifth century. In the fifth century the doors of Santa Sabina at Rome show in certain Oriental-Hellenistic types the change that was taking place, while in the sixth or seventh century these renderings are almost consistently continued on a gold encolpium in the Museum at Constantinople. The noticeable iconographic isolation of many of the marginal miniatures of the Rabula Gospels denotes this same attempt to Orientalize Hellenistic types. Unlike the full page miniatures of these Gospels, which are known to have been written in 586 A. D. at the Syrian monastery of Zagba. the iconography of the marginal miniatures does not adhere to any one type ; while the full page scenes are obviously Eastern in iconography and show certain affinities with Byzantine art, 11 the marginal scenes present both Eastern and Oriental-Hellenistic types. In the Healing of the Paralytic the rendering reverts to the old catacomb method and depicts the paralytic carrying his bed upside down on his back; the scene of the Nativity, while different from the recognized Hellenistic type in the West, especially in the absence from the composi- tion of the ox and the ass, is not wholly Oriental, since the Virgin does not recline in Eastern fashion on a mattress; and yet the representations of the Baptism, the Miracle of Cana, the Samaritan Woman at the Well, and the Massacre of the Innocents are Eastern types, while the Healing of the Blind and the liturgical rendering of the Communion are related to proto-Byzantine types in the Rossano Gospels. This mingling of types and preservation of Hellenistic details of iconography in so famous a monument as the Rabula Gospels, shows how gradual and almost organic was the growth of Eastern iconography. Oriental-Hellenistic Types The types which I have classified as Oriental-Hellenistic are the fol- lowing : 11 It is possible that the full page miniatures are later than the marginal miniatures in the text (Reil, BUdzyklcn dcs Lcbnis Jesu, p. 78). EARLY CHRISTIAN ICONOGRAPHY 1 66 The Nativity . 12 All Eastern representations of the Nativity are distinguished from the Hellenistic type in the West by the absence of the manger shed. The earliest and briefest example of the Oriental-Hellenistic Nativity is found on a marble relief of Naxos of the third or fourth century. On two late encolpia, one in the Museum at Constantinople and the other in Reggio, and a gold medallion from Akmim the scene takes place in the open ac- cording to the Eastern manner, although, after one of the common Hellen- istic methods of representation, the Virgin and Joseph are still omitted and only the Adoration of the animals and the shepherds is figured. The most developed type is found on the Strzygowski medallion (p. 23) with its addition of the figure of Joseph. The Adoration of the Magi . 13 The main distinction between the Adoration scenes of the East and the West is the aspect of the Magi, who in the West are beardless, while on Eastern monuments they are either all bearded or are differentiated in age, the first Magus being heavily bearded, the second lightly, and the last beardless. On several encolpia and ivories, including the encolpium at Constantinople, the Adoration, with the exception of the three bearded Magi, follows the Hellenistic type in that it does not include Joseph in the composition and represents the Magi advancing in a line. The Miracle of Cana . 11 In the case of this scene the Alexandrian-Coptic type, where one servant, pouring the water into the jars, is added to the Hellenistic type, is really an Oriental-Hellenistic type although it became one of the tradi- tional types in Coptic art. The Healing of the Blind . 15 Without any changes the Hellenistic type, where Christ usually touches the eyes of the blind man, who is small in stature and sometimes carries a staff, occurs on the encolpium in Constantinople, and also on the Brescia casket and a pyxis at Bologna, both of which are attributed to Eastern sources. 12 See p. 22 . 13 See p. 47. 14 See p. 88. 13 See p. 97. EARLY CHRISTIAN ICONOGRAPHY 167 The Healing of the Paralytic A 3 Again the Hellenistic type common on the catacomb walls and sar- cophagi of Rome occurs in a few Eastern examples, as the Rabula Gospels and the Constantinople encolpium, and on the Bologna pyxis to which has been attributed an Alexandrian origin. The Raising of Lazarus The Oriental-Hellenistic type of this scene is only distinguished from the Hellenistic by minor changes of detail which differ in the various examples. The one feature that differentiates the scene on the Brescia casket from the Hellenistic type is the bare head of Lazarus; on reliefs found in Phrygia and Sbeitla (Tunisie), and on two sarcophagi of the fourth century at Ravenna we find, instead of the regular Hellenistic aedi- cula, a tomb of which the fagade, with an arched lintel, is represented in strict frontality; on a relief of Constantinople the only variation from the regular type is the fact that the Hellenistic aedicula is depicted in frontality after the Eastern manner of representation. The resume at this point should include a summary of the iconographic features of the Provengal school which in its blending of Hellenistic types and Oriental influences represents the same transitional tendency in Chris- tian art which formulated the Oriental-Hellenistic types. As the second part of this volume is devoted to a study of the Provengal school, to save unnecessary repetition I will pass directly to an Oriental school, the Alex- andrian-Coptic. The Alexandrian-Coptic school, in its beginnings, represents a localized and specific manifestation of a transitional phase again analogous to, though more pronouncedly Oriental than, the Oriental-Hellenistic. The designation of the school at once announces its Alexandrian origin and its subsequent continuation in the indigenous art of Upper Egypt. I have already referred to the theory, first advanced by Ainaloff, that Alex- andria was the great centre of the East where the Hellenistic types and motives of Christian art were formulated, given their specific sym- bolic meaning, and transmitted to Rome to eventually become the basis of a Western school of Christian art. \\ hile not everywhere accepted, such a theory is reasonable and very 15 See p. 104. 1' See p. 1 14. EARLY CHRISTIAN ICONOGRAPHY 1 68 probable. Rich in a flourishing commerce which brought her tribute from all the shores of the Mediterranean and even from the distant lands of the East, Alexandria, even in Christian times, was famous for the magnificence of her edifices, the profusion and picturescjueness of her art, and the wealth of her learning; to her unsurpassed library she owed a great school of erudition and a prodigious literary activity; and up to the time that the Arabs captured the city she was a veritable centre of Hellenistic cul- ture. With the advent of Christianity and from the end of the second century, with Clement, to the third century, with Origen and the Neo- platonists, she was a formative source of Patristic theology. There the Christian theologians, still inspired by Greek models and classic philosophy, sought to harmonize the spiritual infinitudes of Christianity with the rational philosophy of the ancients and to seek through allegory a revela- tion of the New Testament in the prophecies of the Old. Hence in Christian literature and philosophy there grew up in Alexandria a school of theology founded on mysticism, allegory, and symbolism which boldly adopted Hellenistic models only to imbue them with Christian meaning. Certainly no better soil could be found to which to attribute the incubation of Hellenistic Christian art, which in its turn is also an adaptation of Hellenistic models made serviceable to the needs of Chris- tianity by the intricate symbolism read into their classic forms. Through the first part of the fourth century the role of Alexandria continued to become more celebrated. There Arianism was born; thence, from the reaction against Nestorianism, the Monophysite heresy spread throughout the Oriental world. By the beginning of the fifth century, how- ever, Christianity developed as a manifestation of national spirit and the heresies were largely a form of separation — a violent reaction against Roman domination and hence against Hellenism, as the imposed culture of another civilization. From time immemorial tradition was the dominant factor in the de- velopment, or rather sterilized preservation, of Egyptian art. While Alexandria could flourish, for a time, as the capital of Hellenism and could even spread the culture and art of the Greeks across the soil of the Pharaohs, the veneer was thin and the heart of the country remained inherently Egyptian, unalterably imbued with the ancient traditions of the land. When Christianity triumphed and spread up the valley of the Nile, it was accepted as a sign of the renaissance of national tendencies EARLY CHRISTIAN ICONOGRAPHY 169 and aspirations and spread as a reaction, and finally almost as a revolt, against Hellenistic domination. During the closing years of the fourth and the beginning of the fifth century this reaction actually broke forth; the destruction of the Serapeum, the death of Hypatia, and the violent leadership of the nationalists under Schnoudi, bear witness to the hostility of the new faith to Hellenism. With this awakening of national feeling, the centre of Christian art passed from Alexandria to the monastic communities of Upper Egypt and the old traditions of Pharaonic art reasserted themselves in the pictorial representations of Christianity. Originality, however, was never a char- acteristic of Egyptian art. The creative faculty had long since succumbed to the tyranny of immutable tradition. The native art offered few accept- able forms when pressed by the necessity of finding expression for Christian ideals. It was only the spirit of the old art that was revived in the new. Above all else, it was the inherent, non-artistic attitude of mind of the Egyptian, still continuing to view pictorial form not as art but as cold and lifeless hieroglyphs — satisfactory if expressive — that gave to the native Coptic art of Upper Egypt its character. Under this non-artistic tradition all classical beauty of form was abandoned and all pride in execution was lost Coptic art became coarse, crude, and sketchy in technique; it forsook the freedom and naturalness of the Hellenistic rendering of the human fig- ure and reverted to the old habits of frontality and the static rigidity of Nilotic prototypes. Like Pharaonic art before it, Coptic art became con- ventional and stylized, a mere backwater in the history of artistic de- velopment. Sterile and more hieroglyphic than artistic, this Coptic art could originate little and find less in its native monuments to use for Christian expression. It therefore had many ties with the classic Christian art of Alexandria. The ornament and style of Christian art in the Hellenistic capital, although it became more and more debased and Oriental under the introduction of Syrian motives and ideas, remained neo-classic until the sixth century, when the artistic activity passed up the Nile to such monastic centres as Bawit, el Bagawat, and Sakkara. During the transi- tion. especially in the sixth century, Coptic art began by adopting the Hellenistic ornamental forms and ignorantly copying the style. Both ornament and style, however, were soon coarsened and conventionalized and the ornament enriched bv native motives. In the seventh centurv. I/O EARLY CHRISTIAN ICONOGRAPHY when the last vestige of Hellenistic style had almost disappeared and the ornament had been stylized beyond recognition, the iconography originally emanating from Alexandria persisted still. Although Coptic art origi- nated a few types and drew others from Palestinian sources, as I will show in the discussion of the Palestinian-Coptic school, by far the greater part of the Alexandrian iconographic types were continued and, with native veneration for the hieroglyphic form, were conventionalized and so preserved in Coptic art. The most important and the most prominent monument in this dis- semination of Alexandrian types was the Maximianus chair. The Egyp- tian origin of the cathedra, which has been generally accepted and the iconographic proofs of which I have already published , 18 is chiefly demon- strated by the fact that practically every iconographic type occurring on the chair appears at a later date on assured Coptic monuments. It is further established by the nature of these types which in many cases may be shown to have been of Egyptian origin or peculiar to Egyptian art. Moreover, in the costumes of the soldiers, in the treatment of the hair, in decora- tion, and in style, the cathedra is analogous to frescoes at Bawit, to the Barberini diptych, the St. Menas pyxis in the British Museum, and numerous ivory and wood carvings from Egypt in the Cairo, Berlin, and Paris Museums. Its Alexandrian origin, once its con- nection with Egypt is established, becomes certain. The beauty and superior character of its execution, the reminiscences in many of its scenes of classical motives and types, the similarity that its ornament, style, and technique bear to ivory fragments from Alexandria, and the Hellenistic, transitional character of its iconography, all point to its exe- cution at Alexandria, the mother of Hellenistic Christian art and the great centre of the ivory trade of the Orient. Accepting the sixth century as the customary date assigned to the cathedra, although it might have been carved as early as the end of the fifth century, the Maximianus chair is to be considered the earliest and the most pretentious monument of a group of ivories which are intimately related to it in style and iconography. While a large number of ivories may be said to be of the same school as the chair , 19 the ivory book covers of the Etschmiadzin Gospels, two covers in the Bibliotheque Rationale at Paris, and a carved wooden lintel from the church of el Mu’allaka near 18 In American Journal of Archaeology, 1917, pp. 22-37. 19 Dalton, Byzantine Art and Archaeology, pp. 207-209. EARLY CHRISTIAN ICONOGRAPHY 171 Cairo are most closely connected with it. 20 Clearly based upon the icono- graphy of the cathedra, these examples of sixth century art are more transitional in style, less Hellenistic in execution, and hence a little later in date. From this group the iconography, soon to lose all Hellenistic style and technique, passed to Upper Egypt and was continued in the frescoes of Antinoe and Bawit, on the Coptic ivories of the Murano book Fig. 144. Ravenna : Museum, ivory panel OF CHAIR OF MAXIMIANUS. The ANNUNCIA- TION. cover group, and on the medallions, fabrics, vestments, lamps, and other minor objects from the Thebai'd. Alexaxdriax-Coptic Types The salient features of these iconographic types of the Alexandrian- Coptic school are the following: The Annunciation. The Annunciation (Fig. 144), as it occurs on the chair of Max- Leclercq in Cabrol’s Dictionnaire d’aicheologie cliretienne et dc liturgie, s. v. “Caire,” fig- 1833. EARLY CHRISTIAN ICONOGRAPHY 172 imianus, the ivory covers of the Etschmiadzin Gospels and the book covers in the Bibliotheque Nationale, depicts an angel accosting the Virgin, who sits at the left of the composition in a wicker chair spinning the purple veil for the temple. All these examples are nearly identical and the Virgin, with her spinning, seated in a high backed, round topped, wicker chair is the distinguishing feature of the type. This type, with the ex- ception of the examples already cited, occurs only on monuments whose Coptic origin is assured. These Coptic monuments are a fragment of the Murano book covers in the Stroganoff Collection 21 (Fig. 145), an Fig. 145. Rome: Stroganoff Collection, ivory panel from Murano book covers. The Annunciation. ivory plaque in the Uwaroff Collection, 22 a medallion from Egypt in the Gans Collection at Berlin 23 (Fig. 146), a Coptic textile in the Vic- toria and Albert Museum, 24 and a fresco at Bawit. 25 On all, save the Fig. 146. Berlin : Antiouarium, gold medallion from Egypt. The Annun- ciation. Fig. 147. Monza: ampulla. The Annunciation. 21 Graeven, Elfenbeinwerke ( Italien ), no. 64. 22 Strzygowski, Byzantinische Denkmaler, I, p. 43. 23 Amtliche Bcriclite aus den Kgl. Kunstsammlungen, XXXV, 1913, no. 3, fig. 46; Den- nison: Gold Treasure from Egypt ( Univ . of Mich. Studies, Hum. Ser. XII), p. 127. 24 Dalton, op. cit., fig. 381. 23 Cledat, Comptcs Rendus, 1904, p. 525. EARLY CHRISTIAN ICONOGRAPHY !73 Stroganoff fragment, the angel carries a cross instead of a flowering wand. The presence of the cross, either borne by angels or by the Saviour in miracle scenes, is an Egyptian feature. Comparing the Annunciation in Syrian and Palestinian examples, one finds on the Monza phials (Fig. 147), in the miniatures of the Etschmiadzin Gospels, 26 in the Rabula Gospels, 27 and in a Syrian miniature of the Bibliotheque Nationale at Paris (syr. 33) that the Virgin stands as the angel of the Lord accosts her. that the odd wicker chair seen on the Maximianus chair does not occur, and that the Virgin and the angel wear each a nimbus. The Test of the Virgin by Water. This scene, which occurs on the chair (Fig. 148) and the related book covers, is in itself an Egyptian type, for it was a rare scene in early Chris- Fig. 148. Ravenna : Museum, ivory panel of chair of Maximianus. The Testing of the Virgin. tian art and, save for these examples, occurs only on the Uwaroff ivory and the fragment of the Murano covers in the Stroganoff Collection (Fig. 149). Joseph Assured By An Angel. While occurring on a sarcophagus of Le Puy (see Fig. 154) and on the V erden casket in the Victoria and Albert Museum (see Fig. 157) 29 Strzygowski, Byz. Denk., I, pi. V/2. Garr., op. cit., Ill, pi. 130/1. 174 EARLY CHRISTIAN ICONOGRAPHY which I have attributed to Provence, 2S a region whose relations with Egypt will be pointed out, 29 this scene was even rarer than the Test by Water and may also be called Alexandrian-Coptic. The only other examples known to me, on the Maximianus Chair (Fig. 150) and in the frescoes at Antinoe 30 (Fig. 151), are strikingly analogous. The Journey to Bethlehem. This scene, besides appearing on the cathedra, the book covers, and a pyxis from Minden in the Berlin Museum (Fig. 152), that belongs to the Maximianus group, also occurs on the Stroganoff fragment of the Murano covers and in the frescoes of the church of De'ir Abou Hennvs at Antinoe (Fig. 1 5 1 ) . The similarity of these last two Coptic examples, in which Joseph supports the pregnant Virgin on the animal that an angel leads, is too close to the rendering on the chair and too rare a scene in Christian art not to suggest that it originated in Egypt, preferably in Alexandria, whence it passed to become a characteristic type in Coptic art. The Nativity ? 1 The Alexandrian-Coptic type of the Nativity which occurs on the chair is characterized by the introduction of the doubting midwife Salome into what would otherwise be a general Oriental type. The Christ Child, wrapped like a mummy, lies on a stone or brick crib as in the Provengal Nativities, and the ox and the ass adore Him; on one side stands Joseph and on the other Mary reclines on a mattress; either in front of the crib 28 See p. 221 sq. 29 See p. 192 sq. 30 Leclercq, op. cit., s. v. “Ane,” col. 2058, fig. 599. 31 See p. 23. Fig. 149. Rome: Stroganoff Collection, IVORY PANEL FROM MURANO BOOK COVER. THE Testing of the Virgin. EARLY CHRISTIAN ICONOGRAPHY i/5 or before Mary, kneels Salome, extending her withered arm in suppli- cation to the Yirgin. This distinctive rendering, whose Egyptian origin has been shown from the Coptic emphasis on Salome in the apocryphal literature of the country, occurs not only on several Coptic pyxides but also in the frescoes at Bawit and on the Manchester fragment of the Murano book covers. The Adoration of the Magi . 32 The salient feature of the type is the introduction of an angel, carry- ing a wand, who points out the Christ Child to the Magi. These adoring Magi advance in a line and, after the method on all Eastern representa- tions. are differentiated one from another by their beards. Their costume. Fig. 150. Ravenna: Museum, ivory panel OF CHAIR OF MAXIMIANUS. The AnGEL AP- PEARING to Joseph and the Journey to Bethlehem. with heavy bands of embroidery running down the sleeves of their blouses and their long trousers, is characteristic of the type and peculiar to Egypt. The Yirgin sits in a three-quarter pose in a high backed, round topped, wicker chair, similar to the chair seen in the Annunciation. While this type, which appears on the Maximianus Chair and the Etschmiadzin covers, occurs on two Coptic textiles and a medallion from Egypt, it was not common in the native art. as the more Oriental type of the Adoration from Palestine was adopted on most of the Coptic monuments. 32 See p. 48. 176 EARLY CHRISTIAN ICONOGRAPHY The Massacre of the Innocents , 33 This scene as it appeared on the Maximianus chair and later in the frescoes at Antinoe and Bawit was the characteristic “sword type” of the Orient. The Baptism. 3 * The distinguishing feature of the type is the introduction of the per- sonified Jordan as a fleeing river god, although the presence of one or more angels as attendants is also characteristic. On the Maximianus Chair and an Alexandrian ivory in the British Museum, Christ is small and beardless but in the representation in the Bawit frescoes he is bearded after the Syro-Palestinian type. The Miracle of Cana . 35 This is a “transitional type” where a single servant, pouring the water into the six jars, is added to the old Hellenistic type in which Christ, Fig. i 5 i . Antinoe: fresco. The Angel Appearing to Joseph and the Journey to Bethlehem. as a magician, touches the jars with His wand. Besides occurring on the Maximianus chair, this semi-historical type appears on the Gans medallion from Egypt. The Healing of the Blind . 36 This is simply a continuation of the Hellenistic type, with the added feature of the Coptic cross which Christ carries in miracle scenes; it ap- pears in both the group which imitated the style of the Maximianus chair, and that which is connected with the book cover from Murano. 33 See p. 62. 35 See p. 88. 34 See p. 79. 36 See p. 98. EARLY CHRISTIAN ICONOGRAPHY i/7 The Raising of Lazarus ? 7 Again the rendering does not differ materially from the Hellenistic type save in the added feature of the cross that Christ carries. While this type occurs on a few Egyptian examples besides the Maximianus chair and the book covers, there was another and more clearly Coptic type that was popular in Upper Egypt and will be described under the Palestinian- Coptic school. The Entry into Jerusalem , 3S The type differs from all other renderings in the long carpet which is unrolled, instead of a mantle, in the pathway of the Saviour. The East- ern character of the scene is shown in the sidewise manner in which Christ rides on the ass. while the representations of the Maximianus group and the carved lintel from the church of el Mu'allaka near Cairo, are Fig. 152. Berlin: Museum, ivory PYXIS FROM MlXDEN. The JOURNEY to Bethlehem. bound together by the odd. high stepping movement of the ass. In the scenes on the Maximianus chair and the Etschmiadzin covers Christ carries a cross, the Egyptian symbol of divinity. ijc # In the general reassertion of national tendencies and the awakening of Oriental ideas that followed the triumph of Christianity and the gradual dissolution of Hellenism, Syria and especially Palestine were centres as influential in the formation of Orientalized Christian art as they were in the evolution of the Christian faith, its forms, liturgies, and literature. As the Syrian Church separated more and more from the Western Church, as new feasts were observed, and, after the Council of Ephesus in 431, 3 - See p. 126. See p. 1 1 7. i/8 EARLY CHRISTIAN ICONOGRAPHY as the worship of the Virgin became authorized and universal throughout the eastern shores of the Mediterranean, Christianity in Syria rapidly accpiired an indigenous character. Heresies sprang up and the Oriental aversion to the impersonality of the Hellenized forms of Christianity gave rise to a large body of literature that afforded a narrative and personal ac- count of the miracles and lives of the Virgin and Christ. The formation of new feasts and liturgies, the creation of cjuasi-historical stories about the Biblical characters, and, above all, the reaction against the cold impersonality of Hellenistic symbolism, developed new types in art. New scenes arose around the feasts and liturgies, from the apocryphal gospels came a more narrative art wherein the miracles and events in the life of the Saviour were less symbolically and more picturescpiely repre- sented and the Virgin received a more important role. In place of the impersonal Christ, the young and unbearded Hellenistic type, was substituted an historical and mature representation of the Saviour, with bearded face in the Passion scenes or those symbolic of divinity; the apostles were more carefully characterized; and the Virgin was universally conceived as a woman of sorrowful maturity with a long face and a heavy veil. The new spirit animating Christianity not only manifested itself in the gospel miniatures, which are exemplified for us in the Etschmiadzin Gospels, but also found monumental expression in the mosaics and frescoes that decorated the celebrated sanctuaries erected by Constantine in Pales- tine. Christian piety drew the devout in throngs from all parts of the Christian world to these holy places so intimately associated with the memory of the Saviour. At the hands of these returning pilgrims and by means of reliquaries, ampullae filled with holy oil, and other minor ob- jects of the cult which were sold at the shrines and apparently decorated with more or less faithful copies of the scenes in the sanctuaries, the composi- tions of the mosaics and frescoes of the Holy Land were spread through- out the East and subsequently into the West. Thus the themes of the Nativity and the Adoration in the basilica at Bethlehem, the Ascension in the church on the Mount of Olives, the Pentecost in the church of Zion, and the Anastasis in the apse of the Martyrion, all of which are now de- stroyed and known to us only by the copies on the Monza ampullae and other minor objects of Palestinian origin, were developed into sacred and traditional types. EARLY CHRISTIAN ICONOGRAPHY 179 While the tendency of Oriental Christian art was narrative, historical rather than highly symbolic, and personal rather than impersonal, this Palestinian iconography, partly due to its dependence on the large mosaic compositions and in part to the reawakened Eastern tendencies in art, soon became stylized and took on a highly symmetrical, static, and strictly frontal character that was often monumental and hieratic. It was this monumental style which passed with the iconography into other localities. The relations between Palestine and Egypt were intimate. From Syria devout Christians penetrated to Upper Egypt where they sought refuge in the monasteries and studied the cenobite system before spreading into remote countries to found new monasteries. From Egypt came an endless procession of pilgrims to the shrines of the Holy Land where they did reverence before the monuments and then returned to their monastic refuges in the Thebaid. Thus there was a constant inter- change between Palestinian art and the monastic art of Upper Egypt, which we call Coptic, and many of the monumental types of iconography seen in the mosaics of Bethlehem and the more narrative and apocryphal types seen in Syrian miniatures were introduced into Coptic art. Whether the frontal, symmetrical, and monumental character of many of these scenes, which were common both to Palestine and Egypt, was drawn entirely from the monumental types of the churches of the Holy Land or whether there was also a reversion to native, Nilotic methods of repre- sentation in Egypt which influenced the Palestinian compositions, we can not say. It is interesting, however, to note that the scenes in Coptic art which revert to strict frontality, rigid and symmetrical composition, and a monumental style, are those scenes which I call Palestinian-Coptic, and which in general show the influence of the Holy Land. Palestinian-Coptic Types The following types may be characterized as Palestinian-Coptic : The Nativity , 39 1 he Birth, with the adoring ox and ass and the star hovering above the Child in His crib, is figured in the open; the Virgin lies, in the Eastern manner, on a mattress, and Joseph is usually figured seated at her side. This type, which is supposed to have decorated the walls of the Church 39 See p. 29. i8o EARLY CHRISTIAN ICONOGRAPHY of the Nativity at Bethlehem, appears on the Monza phials of Palestinian origin and on the Golenisheff panel and a Coptic tapestry from Egypt. The Adoration of the Magi , 40 The dominant features of this type are the monumental character of the composition, its symmetry, and the frontality of the figures. The Virgin, instead of being represented in profile or in a three-quarters pose, is in strict frontality and the Magi are arranged symmetrically on either side, the two Magi on one side being balanced by a Magus and an angel on the other or the scene of the Adoration by the calling of the shepherds. The Magi are invariably differentiated by their beards. The Baptism , 41 Aside from the absence in the composition of the personified Jordan, the only feature that distinguishes this type from the Alexandrian-Coptic is the consistent use of the nimbus on all the figures, a characteristic detail in all Palestinian scenes. In the Rabula Gospels from Syria and the Golenischeff panel from Egypt the Saviour is bearded, which is a Syrian feature. The Miracle of Cana . 42 This type manifests the narrative tendency in Eastern art. The Vir- gin and two servants pouring the water are introduced, the number of jars is always six, and the disciple is omitted. The type appears in Syria in the miniatures of the Rabula Gospels and in Egypt in the frescoes at Antinoe. The Healing of the Paralytic , 43 Inasmuch as there is an Alexandrian-Coptic type of this scene and yet the Murano book cover group from Egypt presents another type, wherein the paralytic is represented in profile looking back at Christ who carries the Coptic cross, I have classified the second rendering as Pales- tinian-Coptic, although there are no certain Palestinian examples of the scene. The Raising of Lazarus. 4 * The same is true of the Raising of Lazarus. There are no Palestinian examples of the scene and yet on the Murano book cover group there is 40 See p. Si. 41 See p. 82. 42 See p. 92. 43 See p. 106. 44 See p. 118. EARLY CHRISTIAN ICONOGRAPHY 181 a marked type, strikingly different from the Alexandrian-Coptic. This rendering is distinguished from all others by Lazarus, his head hare, stand- ing in the entrance of a frontal tomb faqade with a broken lintel and by Christ carrying a cross instead of a wand. Entry into Jerusalem , 45 There is little to distinguish this type from the regular Hellenistic type save that the Saviour rides sidewise, instead of astride, on the ass. ^ ^ ^ With the Palestinian-Coptic one would expect to discover a clearly distinguishable and influential Syrian school. The commercial activities of the Syrians, the wealth and originality of their architecture, and the importance of their theologians, justify such an expectation. The school no doubt existed, but in iconography it is not distinguishable ; the monu- ments are too few in number and those that remain are too eclectic in their types to disclose its character. I have already pointed out the icono- graphic isolation of many of the types in the marginal miniatures of the Rabula Gospels, as well as their Oriental-Hellenistic characteristics and, in some cases, their affinities with proto-Byzantine iconography. The miniatures of the Etschmiadzin Gospels, more Oriental in style than those of the Gospels from Zagba, present so many points of iconography analo- gous to Palestinian-Coptic types, that they also fail to suggest a separate tradition of Biblical representation in northern Syria. Lor a few scenes it is possible to recognize a Syrian type and in the development of Byzantine iconography, as in the evolution of Byzantine architecture, it is possible to trace certain methods of representation to Syrian sources as well as to Anatolia where it has generally been supposed that the prototypes of Byzantine iconography were formulated. The iconographic development of Byzantine art brings this review to the final phase of early Christian art. The custom of including as “Byzantine” the entire development of east Christian art from the fifth century in Egypt, Syria, and Asia Minor, as well as in Constantinople, is not sufficiently accurate. The actual Byzantine, with its consistent f •rmulae of iconography, its fully developed methods of representation, and its traditional forms, did not appear until the ninth century as a homogeneous style, although as early as the sixth century, and far earlier 43 See p. 127. EARLY CHRISTIAN ICONOGRAPHY 182 in the case of its architectural and ornamental forms, its component char- acteristics appear in the arts of the East. Prior to the reign of Justinian, the new capital which Constantine had established in the East was artistically subservient not only to the Hellen- istic art of Rome and Greece, but also to the new arts which were origi- nating in the old, established centres of the Orient. Springing suddenly from an obscure seaport on the Bosporus into the prominence of an im- perial capital, it lacked an artistic background and was dependent on the traditions of the Hellenistic art of Asia Minor and of the developing arts of the Orient which could be appropriated, by imperial patronage, for its adornment. For centuries Constantinople was the clearing house of the Orient and, while all the elements of the new styles from Egypt, Syria, and Asia Minor found their way to the capital, and the ateliers of Constantinople became centres of artistic production, it was an eclectic and exotic art that was practiced there. By the sixth century the brilliancy of Constantinople reached a kind of zenith under Justinian. Cinder this ambitious, art loving, and lavish Emperor who adorned not only the capital but even his whole realm with great churches, resplendent with mosaics, rich carvings and sumptuous inlays of costly marble, the artistic centre of gravity shifted to Con- stantinople. This “first Golden Age of Byzantine Art,” as it has been frequently called, did not mark the final formation of a Byzantine style. Rather it was the culmination of the early Christian period, a magnificent compilation of the new artistic elements of the East, exe- cuted on a scale hitherto undreamed of, which were now to be assimilated into Byzantine art. “The systems of architecture,” wrote Strzygowski, “peculiar to the most important regions of the Hellenistic Orient (Egypt, Syria, and Asia Minor) combined to evolve the new forms of Hellenistic- Oriental art, the Byzantine, and Hagia Sofia must be considered as the incomparable monument of this reciprocal penetration.” Magnificent and unapproached as was this edifice, rather than an example of a Byzantine type, it was the unexcelled perfection of Hellenistic and Oriental forms brought into perfect and harmonious accord for the first time. While exemplifying the elements of Byzantine architecture, it was not, as a whole, characteristic and typical of the Byzantine type of church which appeared in the ninth, and was universal in the eleventh century. EARLY CHRISTIAN ICONOGRAPHY 183 In iconography this is even more obviously true. Although any of the forms of Byzantine iconography may be recognized in the sixth cen- tury as they occur in the native arts of Asia Minor, Syria, and Egypt (and were brought to Constantinople), it was not until after the Iconoclastic controversies and the invasions of the Arabs, Persians, and barbarians from the north had stilled the national arts in the provinces, that Byzantine iconography appeared in fixed forms and universally adopted types which approached a science in their consistency. Byzantine iconography, which is itself characteristic and specific, was eclectic in origin. As in the architecture, features and types were adopted from Egypt and Syria, but by far the greater number of prototypes emanated from Anatolia. The influence of Syria and Palestine on this formation has already been pointed out by Heisenberg in his study of the mosaics of the Church of the Holy Apostles at Constantinople, and the influence of the Rossano Gospels and the Sinope fragment, from Anatolia, has been generally recognized. The character of proto-Byzantine iconography, as it occurred in these Anatolian miniatures and became characteristic of the developed types, was both monumental and narrative. The proto-Byzantine types, which were to be formalized into an almost adamant tradition, a kind of pictorial script for Byzantine painters, were highly historical in conception as they in- cluded all the characters in any way connected with the actual event to be depicted. This amplification of the types, in spite of their dogmatic, liturgi- cal. and symbolical forms and the stereotyped and hieratic rendering which gave them a monumental style, made the iconography more narrative than that of other Oriental schools. Proto-Byzantine Types The following summary of the scenes which have been discussed in the previous pages briefly illustrates the character of proto-Byzantine types and their relation to Byzantine iconography of the ninth century. In many cases, where the prototypes do not exist, it is only the salient features of the Byzantine type which are stated. Nativity.*® There is a small and uncommon sub-type, which includes an eighth century censer from Kertch, in which the midwives are omitted and the ^ See p. 30. See also note 78, p. 32. i 84 EARLY CHRISTIAN ICONOGRAPHY Adoration of the Angels introduced. The regular type, which occurs on no early Christian examples and was not universal until the ninth century, represents Mary lying on a mattress with Joseph sitting at one side, depicts at first the event in the open and later in a cave, and besides the two mid- wives washing the Child, includes the adoring angels, animals, shepherds, and often the Magi. The Adoration of the Magi , 4T In the ninth century the famous Homilies of Gregory of Nazianzus presents a type, in many ways a continuation of the Alexandrian-Coptic, in which an angel points out the Child, seated in His mother's lap, to the three Magi, who, differentiated one from another by their beards, advance in a line. There appeared, however, in the eleventh century a rigid liturgical type, based on the combination feast celebrated at Con- stantinople, which combines the Epiphany with the Nativity and with the Adoration of the Shepherds, and it was this type which passed into Italo- Byzantine art. The Miracle of Cana .* 8 This is an amplification of the Syrian and Palestinian-Coptic type in which the wedding feast is added to the representation, with the bride, groom, and guests sitting around the table as the Saviour, at the suggestion of the Yirgin, performs the miracle. The Healing of the Blind , 49 There are two types based on two different miracles wherein Christ healed the blind. One is a continuation of the Syrian type, as it occurs in the Rabula Gospels, in which two blind men seek the mercy of the Saviour. This type occurs in the sixth century in the proto-Byzantine miniatures of the Sinope fragment and in the ninth century in the minia- tures of the Homilies of Gregory. The second type comes from a proto- Byzantine source in the miniatures of the Rossano Gospels and also occurs ,in the miniatures of the Homilies. Here Christ, accompanied by His disciples, touches the eyes of a single blind man and to the right of the composition is the pool of Siloam where the blind man washes the spittle from his eyes before a throng of onlookers. 47 See p. 58. 48 See p. 93. 49 See p. 101. EARLY CHRISTIAN ICONOGRAPHY 185 The Healing of the Paralytic . 50 This again is the amplified, historical rendering of the scene wherein the paralytic, after the Biblical account, is being lowered on his bed to Christ through the roof of the house. While there remain no early Christian prototypes of this scene, it appears in the eighth century in the Byzantine frescoes of San Saba at Rome and in the ninth century in the Homilies of Gregory. The Raising of Lazarus . 51 The scene that appears in the sixth century in the miniatures of the Rossano Gospels became the traditional Byzantine type. In the rock hewn tomb is seen the swathed body of Lazarus supported on the right by a servant, while another servant is often figured either undoing the wrap- pings or holding back the slab of the tomb. Christ is bearded and clothed in a purple tunic and wears a cruciform nimbus; at His feet the two sisters of Lazarus lie prostrate and behind Him are groups of disciples and spec- tators ; a characteristic feature is a man holding his nose. The Entry into Jerusalem . 52 Again the type which appears in the sixth century in the Rossano Gospels was fully developed by the ninth century in the Homilies. The Saviour, followed by His disciples, rides sidewise, while boys spread a mantle beneath the feet of the ass and others look down from treetops on the scene. A large multitude issues from the gates of the city and people look out from the windows. The Last Supper , 33 Two types of this scene exist, one liturgical and the other historical. The Liturgical type, which appears in both the Rabula and Rossano Gospels, is a representation of Christ as Priest in the Lirst Communion with His apostles, and only relates to the Last Supper as that event was the occasion of the Lirst Communion. This type became a favorite theme for the decoration of Byzantine apses. The historical type, which also occurs in the Rossano Gospels and became traditional in Byzantine art. depicts the Saviour seated with His twelve disciples, among whom Judas is distinguished by the way he reaches forward to grasp the bread and wine of which they are to partake. By the ninth century the beloved •*° See p. 106. 52 See p. 127. r ' 3 See p. 139. 51 See p. 120. EARLY CHRISTIAN ICONOGRAPHY 1 86 disciple, John, is characterized as he leans on the Master's breast and Judas is isolated from the rest of the Twelve. JfC JjC The most interesting result of the foregoing classification of types is the emergence, — out of the hitherto confusing mass of sixth century monu- ments,' — of the Alexandrian-Coptic and Palestinian-Coptic groups of Egypt. To the former must now be assigned the much discussed Chair of Maximi- anus and the ivories related thereto ; the latter includes the numerous works that have been gathered about the Murano book cover, and adds materially to the archaeological data bearing on the relations of the Holy Land with Upper Egypt in the sixth century. These groups, and the Provenqal which is treated at length in Part II, are the only ones that can be localized into “schools,” and therefore afford criteria by which to judge the specific provenience of works presenting their characteristic types. The eclectic group of Provence presents a few types of its own, a number of traditional Hellenistic forms, and several others of Oriental origin which will be dis- cussed in Part II . 54 54 See p. 201. PART II A SCHOOL OF IVORY CARVERS IN PROVENCE I INTRODUCTION The first part of this volume has been devoted to a study of iconography with the idea of isolating and defining the artistic centres of early Chris- tian art which flourished around the shores of the Mediterranean during the first seven centuries of Christianity. With the exception of Christ Triumphant, the Biblical scenes which have been chosen to show the icono- graphic independence and distinguishable character of these “schools” of Christian art are the New Testament scenes that appear on a group of ivories that I would attribute to Provence. Since, in my classification of the iconography of the early Christian period, the types that occur on the ivories of this group have stood out in marked isolation from the other types, I have assumed the existence of a new school of art, a school of ivory carvers in Provence, whose connection with Marseilles and pre- sumably with the famous convent of St. Victor I hope to sustain in the following pages. The group of ivories, which consists of a diptych in Milan cathedral, a casket from Werden that is now in the Victoria and Albert Museum at London, a fragment in the Nevers Museum, another piece possibly be- longing to the same ivory in the Berlin Museum, and a diptych in the Library at Rouen, has been generally recognized by the authorities as a unit, with the exception of the Rouen diptych . 1 The eclectic nature of 1 Since this book has gone to press I have come to the conclusion that the four ivory plaques in the British Museum (Dalton, Catalogue of the Early Christian Antiquities in the British Museum, no. 291) belong to this Provencal group. The iconography for the most part is Hellenistic and is similar to the types seen on the sarcophagi of Gaul. Among the scenes on the plaques is an Oriental-Hellenistic type of the Crucifixion which was known in Provence (Gregory of Tours, De Glor. Martyr., XXIII, see p. 205) and is 187 A SCHOOL OF IVORY CARVERS IN PROVENCE 1 88 the group, and the confusion that its mixture of Eastern and Western characteristics has caused in attribution, are most clearly brought out by a cursory review of the various artistic centres to which it has been assigned. The Eastern character of the ivories was first recognized by Labarte 2 who called the Milan and Werden ivories the products of early Byzantine art executed under Justinian at Constantinople. This early Byzantine origin was later reasserted by Dobbert 3 on the ground of the enamel work occurring on the Milan covers. Strzygowski. having first said that the group was Italian (Milanese?) done under Eastern influences, 4 asserted at a later date that the Milan covers may have been brought to Milan, but that they must have been executed nearer Asia Minor. 5 While the ivories of the group have been attributed to Italy by De Rossi, 6 Venturi, 7 Schmid, 8 Stuhlfauth, 9 and Westwood, 10 Haseloff was the first to show their connection with Rome. 11 Of these same authorities who sustained an Italian origin, De Rossi and Schmid have suggested Milan as the possible centre where the work was done and Stuhlfauth has attempted to point out the actual connection of the group with that city. Except for a few analogies with certain sarcophagi, his ostensible reason for such an attribution lay in the fact that Milan was the principal seat of govern- ment in Italy from the reign of Diocletian to 404, and that as early as the middle of the fourth century it was already the great ecclesiastical consistent with the Oriental-Hellenistic character of the region. This representation of the Crucifixion, however, manifests an odd feature in the centurion who thrusts his lance into the left side, instead of the right, of the Saviour. All Eastern examples, from early Christian times throughout the history of Byzantine iconography, represent the lance thrust on the right. While uncommon, this rendering on the ivory appears later on Merovingian and still later on Irish monuments, and leads me to believe that it was a type originating in Provence and passing thence to the later art of Gaul and so to Ireland. Other iconographic peculiarities on these ivories were continued, as I have already shown was frequently the case with Provenqal types, on Carolingian and Otto- man monuments. The technique of the ivories is, moreover, similar to the workmanship and style of the Berlin fragment, which belongs to the Provencal group, and to figures on the sarcophagi of Provence. 2 Labarte, Histoire des arts industrials, I, pp. 43, 211. 3 Dobbert, Repertorium fiir Kunstwissenschaft, VIII, 1885, p. 172. 4 Strzygowski, Bysantinisclic Denkmdler, I, 1891, pp. 45, 49. 5 Strzygowski, Kleinasien, p. 198. c De Rossi, Bullettino, 1865, p. 26 sq. 7 Venturi, Storia deli’ arte Italiana, I, p. 509. 8 Schmid, Die Darstellung der Geburt Christi, p. 109. 9 Stuhlfauth, Die altchristliche Elfenbeinplastilc, p. 84. 10 Westwood, Fictile Ivories, pp. 38-43. 11 Haseloff, Jb. Preuss. Kunsts., XXIV, 1903, p. 53. A SCHOOL OF IVORY CARVERS IN PROVENCE 189 centre of the West. Molinier, who saw the eclectic nature of the Milan covers, did not assign them, but simply remarked upon their Greek char- acter qualified by Italian affinities. 12 Finally Dalton, who with Stuhlfauth viewed the group as a unit, said that it was “less obviously Oriental” than some of the other groups and yet in certain respects suggested a doubt as to its Western origin. 13 Keeping in mind this strange mixture of Eastern and Western features, which nearly all the students of the ivories have recognized, but which none has fully explained, I wish to show in the following chapters how in Provence the East and the West mingled, how the local traditions were Roman while the commercial and religious relations were nearly all with the Orient, and how the art of the sarcophagi manifests the same eclectic character. Before considering, however, the Orientalizing of Gaul during the first six centuries of the Christian era, I wish to emphasize the importance of the fact, already brought out in the previous pages, that the rendering of many scenes on the ivories which I would call Provencal, follows the account in the apocryphal Gospel of Pseudo-Matthew. This interesting gospel purports to be a Latin translation made by St. Jerome from a Hebrew Gospel of St. Matthew, 14 but is really a compilation from the old Jewish-Christian Gospel attributed to Tames, and a Gnostic recension thereof. 15 It certainly dates no earlier than the end of the fourth century and according to Michel it might have been written as late as the sixth. 16 Lipsius inclines to the end of the fifth century as the most likely date. 17 Its date is thus only an approximation while its odd text, primarily Ori- ental in origin, written in Latin, and pervaded with interpolations which are foreign to the other apocryphal gospels of the East, has been inade- quately explained. While this text differs in many respects from other gospels, it is in the description of the Nativity and the Adoration of the Magi that it be- comes evident, from a study of Christian art, that the translator interpo- lated his original with the conscious purpose of bringing the account into j2 Molinier, Hist, des arts appliques, I, Ivoires, p. 61. lZ Dalton, Byzantine Art and Archaeology, p. 203. • * Michel, PLvangiles apocryphcs, page xix. 1 ' Lipsius, in Diet. Christ. Biog., II, p. 703. 16 Op. cit., p. xxi. 1 Lipsius in Diet. Christ. Biog., II, p. 702. 190 A SCHOOL OF IVORY CARVERS IN PROVENCE accord with the established art in the West. At the outset, in the de- scription of the Nativity, Pseudo-Matthew differs from all the canonical texts authorized in the West and agrees with the purely Oriental versions in relating that the time for the delivery came when Mary and Joseph were on the way to Bethlehem and that Mary went into a cavern. While the writer, following Eastern tradition, introduces the doubting midwife into the Nativity, he also inserts a description of the Adoration of the Ox and the Ass which appears in no other gospel account, either apocryphal or canonical. We read there that “on the third day after the birth of our Lord Jesus Christ the Blessed Mary went forth out of the cavern, and entering a stable, placed the Child in the manger, and the ox and the ass adored Him . . . and Joseph and Mary lived in that place with the Child three days.” I have already pointed out how slight was the consideration of the early theologians for the actual Birth 18 and how in the fifth century, when Pseudo-Matthew was being translated from earlier accounts, the presence of the ox and the ass as symbols of the adoration of the natural world was already traditional. 19 We have also seen that the manger-shed was a fixed feature in Western Nativities. As so frequently hap- pens, the established art influenced the literature of the period ; into the Eastern narrative of the Birth, whose Eastern motives of midwives and the cave contrast so noticeably with the canonical accounts, the Latin translator inserted the only description that we have of the Adoration of the Ox and the Ass, and takes obvious pains to reconcile the cavern of his Eastern source with the shed of Latin tradition. In the case of the Adoration of the Magi, an artistic type had already grown up in the West before the fifth century which was in direct contra- diction to an early tradition that the Birth and the Adoration were syn- chronous. 20 Inasmuch as the canonical versions are vague as to the age of the Child at the time of the Adoration, the translator sought to recon- cile his text with the prevalent custom in the West of representing Jesus as a child of two years. It reads: "Two years having passed, some Magi came from the Orient to Jerusalem, and entering into the house, they found the Child Jesus resting on His mother's breast." Thus as in the case of the Adoration of the animals the translator brought his source 18 See p. 19. 19 See p. 20. 20 See p. 45- A SCHOOL OF IVORY CARVERS IN PROVENCE 191 into accord with an already established artistic tradition, although the East commemorated Birth and Adoration on the same night. At the same time the interpolator, in his effort to popularize the text, makes specific refer- ence to the gifts that the Magi brought; besides referring to the customary triple gift of gold, myrrh, and incense, he tells of “great gifts” given to the Virgin and to Joseph and of the pieces of gold which each of the Magi gave to Jesus. Again he was seeking to conform his text to the frequent presence of Joseph in the scene, and to the variety of objects which in Western art the Magi present to the Saviour. That the Gospel of Pseudo-Matthew was peculiarly adapted to the re- ligious character of Provence will become evident from the following chap- ter on the Orientalizing of Gaul. The country was primarily Gallo-Roman but permeated with an Oriental element in the population which dominated the trade, introduced many of its own motifs of art, and brought the Christianity of Provence more into accord with the Christianity of the Orient than with that of Rome. In the fifth century the West, wishing a more detailed account of the birth of the Virgin, her marriage, and the circumstances connected with the birth of the Saviour, eagerly appropri- ated, regardless of the prohibitory mandates of the orthodox church, the various apocryphal narratives which had been written in the Orient. Of these apocryphal gospels the Pseudo-Matthew, as a Latin translation brought by interpolation into accord with Western notions, was the most popular. If not actually provided, it was admirably adapted for just such a community as Provence where there was an active mixture of Oriental peoples and the indigeneous Latin population. The relation of the ivories to the gospel, as demonstrated in Part I, is obviously an argument for an earlier date than the latter part of the fifth century, for one of the ivories of our group, the Nevers fragment, held by competent authority to be no later than 450, appears to base its scenes on Pseudo-Matthew. In closing this introduction a word of explanation is perhaps necessary concerning the addition to this volume of a long chapter on an ivory that is not included in the Provencal group. The Bodleian book cover, a Caro- lingian ivory of about 800, not only preserves many of the Provencal iconographic types, but is also to a large extent a copy of a diptych of which the Berlin ivory (and possibly that of Nevers) was a part. So striking is the affinity with these ivories that a few authorities have I 9 2 A SCHOOL OF IVORY CARVERS IN PROVENCE even gone so far as to attribute it to the early Christian period. The very fact that a Carolingian ivory preserves the iconography of a group of ivories that may possibly be attributed to Provence helps to confirm the Provenqal origin of the ivories; for as Marseilles and the other cities at the mouth of the Rhone were the main channels through which Eastern ideas, products, and art entered and spread through Charlemagne’s king- dom, so is it more than likely that some of the iconography which passed up that highway into Carolingian art was Provencal. II THE ORIENTALIZING OF GAUL A country's art springs from the deepest strata of its life and activ- ities. Its roots lie not only in the racial characteristics but also in the more dominant exotic influences which are always tending to reshape a race. They are interwoven with and to a large extent dependent upon the social, religious, and economic factors which continually change the aspects of society and alter, stimulate, or hinder the development of its ideals and artistic efforts. I hope to make plain that the early Christian art of Provence was largely inspired by Eastern importations. If so, its many artistic parallels with the East should appear as the manifestations of strong commercial, religious, and artistic currents which were flowing from the Orient into Provence and should make intelligible the curious mixture of Eastern and Western features in the eclectic art of southern Gaul I wish to sketch briefly the commercial, religious, and artistic fac- tors from the East which were working in and on Provence to qualify in favor of Oriental ideas and customs the Hellenistic Roman traditions al- ready established in the region. The commercial relations of Provence with the East reach back to the times when the Phoenicians had a direct route between the Orient and Provence and spread their civilization through the region . 1 During Chris- tian times, when Syria continued the traditional commerce of the Phoeni- cians and spread her products over all the Mediterranean, it was along this route that Eastern merchants with their wares, and later Eastern mis- 1 Courajod, Leqons professees a I’ecole du Louvre, I (1899), p. 48. A SCHOOL OF IVORY CARVERS IN PROVENCE 193 sionaries and Eastern art, travelled from the Orient into Provence. At the beginning of the Middle Ages Syria was one of the most fertile and industrious countries of the world 2 and from the fifth to the eighth century Syrian merchants were the principal navigators of the Mediterranean . 3 Of all the countries of the West which they entered, Gaul was the territory where their influence became most widespread and powerful. The points of departure for this Oriental expansion into Gaul were Mar- seilles and in a less degree the other old Greek cities of Provence . 4 Arles especially had for a long time been a depot of Oriental commerce and a centre of immigration. The relation of Arles with the Orient is attested by a funeral inscription 5 found in that city and by a rescript 6 of Claudius Julianus that was found in Deir in Lebanon, and gives evidence of the relations of the mariners of Arles with the Orient. The forms of merchandise which these Eastern traders imported into the ports of Provence and then dispersed over the whole Gallic region were many and varied. Commodities of all sorts, as the contemporary writers tell us , 7 came from centres in Syria famous for their production. The Syrians, St. Jerome writes , 8 did a great commerce with the West in ivories, stuffs, precious stones, wines, and many other things. Besides the commodities and luxuries destined for the tables of the Gallican nobles there were introduced great quantities of silks 9 to be used for the vest- ments and ornaments of the Christian cult. Syria was not the only country whose products were brought to the shores of southern Gaul. The com- merce with the Egyptians was very active . 10 Also, according to Gregory 2 Diehl, Justinien et la civilisation byzantine, see the chapter devoted to Antioch. 3 Louis Brehier, Byz. Zeit., XII, (1903), p. 37. 4 Brehier, op. cit., p. 12. That Greek had been spoken in Marseilles at a very late date is shown by the inscription on a votive tablet of Christian origin which dates from the fifth century (Le Blant, Inscriptions chrctiennes de la Gaulc, no. 547). 'Brehier, op. cit., p. 12; Le Blant, op. cit., no. 521. ' Brehier, op. cit., p. 12; Waltzing, Hist, des corporations industrielles sous VEmpire, III, no. 1961. ' Brehier, op. cit., pp. 20, 21, 22, summarizes all the evidence. 8 Courajod, op. cit., p. 326; St. Jerome, In Ezekiel, XVIII, 16-18. 9 Marignan, La societe Merovingienne (1899), P- 145 ; Heyd, Gescliichte des Levanthan- dels, I, p. 68; Schefter-Boichorst, Zur Gescliichte dcr Syrer in Abcndlande, Mitt, des Insfituts fiir bsterreich. Geschichtsf., VI, P- 545; Reil, Bildzyklcn des Lebens Jcsn, p. 21, note 2. 1,1 Sulpicius Severus, Dialog., I (ed. Halm, p. 152) ; Gregory of Tours, Hist. Franc., V. 5. See also below, p. 253, note 24. 194 A SCHOOL OF IVORY CARVERS IN PROVENCE of Tours, 11 there were constant relations between Marseilles and Con- stantinople. Not only did these Eastern carriers pour their wares into the ports of Gaul, but the Syrian merchants settled in all the principal towns and carried on most lucrative trades. 12 Corporations of Syrian merchants existed in the larger towns of Italy, Gaul, Spain, Africa, and Brittany. The exist- ence of these colonies is attested by Jonas, biographer of St. Colombanus. 13 St. Jerome says they were everywhere 14 and Salvianus 15 writes that they had taken possession of the greater part of the Gallic towns. In the north we see them at Worms, Cologne, Metz, and Poitiers. 10 The many funeral inscriptions, of which some are dated after the Syrian calendar, 17 show them established in Arles, Narbonne, 18 Tours, 19 Clermont, 20 Vienne, 21 Lyon, 22 and Paris. 23 An epitaph from Lyon, 24 half in Greek verse and half in Latin prose, reads : “Venturous and sweet child of Athele, decurion of the city of Canatha [Ivanawat in the Hauran] in Syria, who left his native country in order to come into this land to trade : he had at Lyon a booth, furnished with Aquitaine merchandise. Irresistible destiny has made him find death on a foreign soil." These large communities of Orientals, which kept together in the towns and often continued to speak their own language, became very powerful. 25 Many interesting stories of their riches and influence are related by Gregory of Tours : he tells of a Syrian merchant in Paris 26 who gained, at the price of silver, the episcopal seat and then introduced into the administra- tion of the diocese a crowd of Syrians from his own family; he also writes of a Syrian merchant at Bordeaux, 27 named Euphron, whose riches 11 Gregory of Tours, op. cit., IV, 39, 40; VI, 24. 12 Marignan, op. cit., p. 145. 13 Courajod, p. 328; Jonas, Vita S. Colombani abbatis, Acta Sanctorum S. Benedicti, II. 14 Courajod, op. cit., p. 325; St. Jerome, In Ezekiel, XXVII, 16; Epist., CXXX, 17. 15 Courajod, op. cit., p. 329; Salvianus, De Gubernatione Dei, IV, 4. 10 Marignan, Louis Courajod, p. 89; Brehier, op. cit., pp. 11-19. 17 Le Blant, Inscriptions chretiennes de la Gallic. 18 Le Blant, op. cit., no. 613-A. 19 Brehier, op. cit., p. 16. 20 Brehier, op. cit., p. 16; Bull, de la Soc. Ant. de France, 1861, p. 86. 21 Le Blant, op. cit., no. 415; Brehier, p. 14. 22 Brehier, op. cit., p. 14; and notes. 23 Brehier, op. cit., p. 17; Vita S. Genovefae (Scriptor. rer. mer., Ill, p. 226). 24 Allmer et Dissard, Mem. de la Soc. Ant. de France, 1865, pp. 1-19. 23 Brehier, op. cit., pp. 18, 19. 28 Brehier, op. cit., p. 17; Gregory of Tours, Hist. Franc., X, 26 27 Brehier, op. cit., p. 13; Gregory of Tours, Hist. Franc., VII, 31. A SCHOOL OF IVORY CARVERS IN PROVENCE 195 excited the jealousy of Berthramnus, the bishop of the town. As these Eastern peoples commenced to gain control of trade and commerce, they profited by it to enrich themselves, and their riches soon raised their social position. With this acquired power, not only did they elect many of the bishops in Gaul and Italy, but in Rome from the sixth to the seventh century they came near to possessing the exclusive privilege of choosing the Pope. 2S During all the time that their influence was augmenting, they tended to preserve their customs and language, and the manner of living of their Oriental lands. Therefore “it is not astonishing that the perpetual contact between the Occidentals of the barbarian epoch and these Orientals, more refined than they, had brought in the long run some changes of ideas and modified in a certain measure the Western culture of the Middle Ages." 29 In other words, the manners, the customs, and the ideas of the region, as we shall see in the case of its art, were bound to become eclectic, a mixture of Oriental notions with Occidental characteristics. Religious expansion has always tended to follow in the wake of the navigator and along the beaten path of the trader. It was in the third century that Christianity began to render itself master of the principal towns of Gaul, 30 and throughout its history the relations of the Gallic Church were largely with the East. The early Gallican liturgy was that of Ephesus, 31 of the primitive church of Asia, imported into Gaul by the founders of the church of Lyon. From this church it radiated over the whole Transalpine West. In the fourth century the leadership passed from Lyon to Trier, Vienne, and Arles, and in the fifth century Arles be- came the centre of ecclesiastical law. Nearly all the “Libri canonum” in usage in Gaul during the Merovingian times were derived from those of the church of Arles. This relation with the East was also preserved by the monasteries which were largely founded by Eastern monks and based on Eastern models. Both at Liguge, near Poitiers, and at Tours, St. Martin, 32 about 360 A. D., founded monasteries modelled after the rule of an Egyptian coenobium. The most famous and influential for its learning was the order of the 2 - Brehier, op. cit., p. 19. 2r ‘ Brehier, op. cit., p. 19. 30 Marignan, Louis Courajod, 1899, Paris, p. 89. 31 Duchesne, Origiues du cultc chretien, pp. 84-86. 32 Brehier, op. cit., p. 32. 196 A SCHOOL OF IVORY CARVERS IN PROVENCE Insulani founded in 410 by Honoratus 33 on the island of Lerins. This order was also partly based cn me cenobite rule. An even greater and more powerful community was the monastery of St. Victor at Marseilles founded by Cassianus, a man who had visited many parts of the East and had lived for years as a monk in the deserts of the Thebaid. 34 Many of the monks in Gaul had also emigrated from various places in the East : Sidonius Apollinaris tells 35 of the death of a Syrian monk from the Thebaid who had been living near Clermont ; St. Abraham, the founder of the monastery at Clermont, was born on the banks of the Euphrates; 36 and in the fourth century the celebrated Bishop Cassianus of Autun was born at Alexandria. 37 Thus from the fourth to the seventh century Gallic monasticism was “clearly Egyptian both in theory and practice” 38 and we may expect to see the influence of Egyptian, as well as Syrian, motifs and ideas on the art of Provence. As a result of this intimate relation of Christian Gaul with the Orient rather than with Rome Eastern feasts 39 were celebrated there which were not observed in Rome nor in that region of the West where the Roman church and liturgy prevailed. Furthermore, the Eastern mission- aries, monks, and ecclesiastics brought into Gaul the worship of Eastern martyrs, their relics, 40 and many art objects 41 of the Eastern cults, all of which had a great influence in shaping after Oriental models the forms, observances, and above all the art of the Gallican church. “French art, which was born Christian,” writes Courajod, 42 “had con- tact only with Judea, Syria, Greece, Byzantium, Ravenna, that is to say, with the Hellenistic Orient. When French art was born, Rome was simply a dependent province of the East; Rome was itself Byzantine and used the Neo-Greek art.” This dependence of Gallic art on the Orient was nothing new in Christian times. Gaul was in close intercourse with the 33 Montalembert, Monks of the West, I, p. 346. 34 Montalembert, op. cit., I, p. 355. 35 Dalton, Byzantine Art and Archaeology, p. 89; Brehier, op. cit., p. 34. 3C Brehier, op. cit., p. 33; Gregory of Tours, Vitae Patrum, III. 37 Brehier, op. cit., p. 15, note 6. 38 Dalton, Bys. Art and Arch., p. 89; Brehier, op. cit., p. 33. 39 Duchesne, op. cit., pp. 258, 264. 40 Brehier, op. cit., p. 35. 41 Dalton, op. cit., p. 88. 42 Courajod, Logons professees a I’ecole du Louvre, I, p. 109. A SCHOOL OF IVORY CARVERS IN PROVENCE 197 Greek world in Roman times and with Persian and Assyrian art before our era . 43 As early as the first century B. C. the Gallo-Roman art was largely influenced by Graeco-Egyptian ideas and motifs, and Marseilles and the cities of Provence were centres where Hellenistic traditions, ema- nating from Alexandria, played a most important role . 44 At Nimes even the municipal institutions were analogous to those of Egyptian metropoles and the cults of Isis and Anubis were popular there . 45 At this early date also the influence of Syria was only second to that of Alexandria, for relations were already established with the large Hellen- istic centres, like Antioch, in Syria . 46 The sarcophagi of Gaul are the only art products of the region which show directly the existence there of schools of craftsmen. Whatever artistic influences, then, were working in the various parts of Gaul would be apparent on these sarcophagi. While sarcophagi have been found in nearly all the old French towns, by far the largest number have come from the cemeteries of Arles, Toulouse, Marseilles, and other cities of Provence. The type of sarcophagus adopted in the valley of the Rhone proceeded from Roman models 47 and the centre of this Latin school was the atelier of Arles . 48 The school of southern Gaul, of the second Nar- bonnaise, extended as far west as Toulouse; far from participating in the style of the schools of Arles and Rome, it recalls, on the contrary, the Eastern style of Ravenna and attaches itself by contact, more or less direct, with the Christian art of the Orient. The Greek culture stream, coming from Alexandria and other Hellenistic centres in the East, which spread into Provence from Marseilles , 49 influenced even the school at Arles, and some of its sarcophagi are more Hellenistic 50 in character than those of Rome. A type such as Christ Enthroned among His Apostles , 51 rare on Roman sarcophagi and common on those of Gaul, recalls in its ren- 43 Courajod, op. cit., I, p. 5-6. 44 Salomon Reinach, Gas. des Beaux-Arts, XI, 1894, pp. 25-42. 43 Reinach, op. cit., p. 26. 46 Reinach, op. cit., p. 38; Lognon, Geographic dc la Gaule merovingienne , p. 177. 4 ' Le Blant, Les sarcophages chrctiens dc la ville d’Arles, introduction, p. vi. 4S Courajod, op. cit., pp. 105, 106. 49 Schonewolf, Die Darstellung dcr Aufcrstehung Christi, p. 38; Lentheric, La Grcce ct YOrient cn Provence, 1848; Louis Jalabert, Les colonics chretiennes d’Oricntaux du V e an VHP Siecle, Revue de I’Orient chretien, IX (1904), pp. 96, 104. 0 Schonewolf, Die Darstellung der Aufcrstehung Christi, p. 40. 51 Garr., op. cit., V, pis. 343/1, 2, 3. 198 A SCHOOL OF IVORY CARVERS IN PROVENCE dering the same scene on the Alexandrian pyxis at Berlin, 52 pointing thus to the Hellenistic art of Egypt. It is the Syrian influence, however, which is the strongest on the sarcophagi of Provence and is the most interesting for our purposes. While the number of actual Syrian tombs in Gaul is large, 53 the decoration of many others, not definitely connected with Syrians by inscriptions, is sufficient in itself to show their Eastern in- spiration. “The ornamental forms and decorative system employed by a people have more importance than ethnological facts or architectural details in determining the relations and migrations of a race.’’ 54 If this be so, the ornamental proof of the peaceful Syrian invasion in Gaul is convincing. The elements of the Oriental grammar of decoration which occur on the sarcophagi have already been compiled by Courajod 55 and Marignan; 56 they include the star with six rays, rosettes, interlacings, the braid, the heart shaped leaf, the helix, the palmette, the vase with birds or animals confronting it, and above all the grapevine. Among the many analogies in decoration between Gaul and Syria there are five motifs which most clearly show the ornamental intimacy between the two regions. The star with six rays was one of the most common elements of decoration on the buildings and tombs of the East and it was introduced directly into Provence from Syria. 57 It entered the country, as did the trade, through Narbonne, and from the region around Narbonne there remain several examples of it on the sarcophagi. 58 The running grapevine with or without birds interspersed among the leaves, common on the buildings of Syria from the earliest to the latest period, 59 appears on several Gallic reliefs of which one is in the crypt of the church of Saint Vic- tor at Marseilles. 60 The vase with a vine or palmette growing out of it was 52 Dalton, op. cit., fig. no. 33 Courajod, op. cit., p. 119. 34 Quoted from Ch. Dresser (Japan, its Architecture, Art, and Manufactures, 1882) by Pottier, Gaz. des Beaux-Arts, 1890, p. 129. 33 Courajod, op. cit., p. 310. 30 Marignan, Louis Courajod, p. 20 sq, 132 sq. 37 Marignan, op. cit., p. 134. 38 Sarcophagus of Venasque (Le Plant, Sarcophages chretiens de la Gaule, pi. LVI/2) ; others cited by Courajod (op. cit., pp. 121, 122; figs. 7, 8 , 9); and a sarcophagus from Poitiers (Le Plant, op. cit., pi. XX/2). 39 At Safa (De Vogue, Syrie centrale, pi. 24), at Si’ (op. cit., pi. 3), at Sabba (Pren- tice, Greek and Latin Inscriptions, Publ. of the Princeton University Archaeological Ex- pedition to Syria, P, I, p. 8), at Dana (De Vogue, op. cit., pi. 45). 00 Sarcophagus of Soissons (Le Plant, op. cit., pi. IV/i) ; sarcophagus of Angouleme A SCHOOL OF IVORY CARVERS IN PROVENCE 199 nearly as common an ornamental motif in Gaul 131 as it was in Syria . 02 The decorative border of heart shaped leaves springing from a running stem occurs on two sarcophagi of Narbonne 63 and on another at Le Mas- Saint- Antonin 04 which is near Narbonne. This motif was character- istically Oriental and one of the most common motifs of the Syrian architectural ornament . 63 The motif of the two peacocks, symmetrically arranged on either side of a vase from which grows a grapevine, appears on a sarcophagus of Provence 66 in a manner that recalls the same form on Syrian churches . 67 Furthermore, the arcade motif 68 so common on Gallic sarcophagi 69 as well as the animation of a flat wall by curtains, birds, and vine motives, and also the conch shell over the niche 70 have already been shown to be characteristic of Oriental ornament. This Eastern ornament was spread in Gaul by three agencies : first, the Oriental merchants and ecclesiastics transmitted for execution their own native ideas to the local schools of carving; second, they also intro- duced from the Orient many art objects which served as models for the Provencal work; and finally. Eastern craftsmen themselves entered Gaul, along with the general migration which was pouring into the country. Gregory of Tours states that the Syrian merchants possessed many art objects and were often venders of them 71 and that they frequently brought relics of Oriental saints , 72 which must have been encased in reliquaries (Le Blant, op. cit., pi. XXIV/3) ; sarcophagus of Bordeaux (Le Blant, op. cit., pi. XXXIII/i) ; sarcophagi of Toulouse (Le Blant, op. cit., pi. XXXVIII/i, 3) ; sarcophagus of Saulieu (Le Blant, op. cit., p. 3) ; altar of Marseilles (Le Blant, op. cit., pi. X/4). ^ Le Blant, op. cit., pis. XX/2, XXVIII/i, XXXIII/i; p. 3. 62 De Vogue, op. cit., pis. 45, 62, 129, 151. 63 Le Blant, op. cit., pis. XLIV/2, XLVI/i. 54 Le Blant, op. cit., pi. XLVITI/i. 65 Relief from Phrygia ( Revue archeologique, xxi, 1913, pp. 333-339, fig. 3) ; De Vogue, op. cit., pis. 3, 4, 43, 62; Prentice, op. cit., B, I, p. 40, B, II, pp. 61, 76; H. C. Butler, Ancient Architecture in Syria, Publ. of Princeton Univ. Exped. to Syria, B, II, figs. 88, 89, 94, 95; Architecture and Other Arts, pp. 296, 302. Le Blant, op. cit., pi. VI. e, ~ De Vogue, op. cit., pi. 45; Prentice, op. cit., B, I, p. 8; Butler, Ancient Architecture in Syria, B, II, figs. 68, 96. 68 Schdnewolf, op. cit., p. 40; O. Wulff, Bye. Zcit., XIII, 1904, p. 573; Strzygowski, Orient odcr Rom, p. 54. 159 Le Blant, op. cit., pis. II/i, XIV, XVII, XVIII/2, XLII/3, etc. 70 Schdnewolf, op. cit., p. 40; Strzygowski, Bye. Zcit., XII, 1903, p. 705, “Mschatta,’' Jahrbuch d. Kgl. Pr. Kunstsammlungcn, 1904, p. 260. 71 Marignan, Louis Courajod, p. 89, note 1 ; Gregory of Tours, Hist. Franc., IV, 35. 72 Gregory of Tours, Hist. Franc., VII, ch. 32. 200 A SCHOOL OF IVORY CARVERS IN PROVENCE of Eastern workmanship. Besides importing the foreign works of art the Syrians were sometimes the artisans, mosaic workers, and sculptors in Gaul. 73 An inscription of Lyon 74 mentions the presence there of Con- stantine of Germanicia (Syria, Commagene) who exercised the wholly Oriental art of gold and silver applique on metals which the Romans called ars barbaricaria. The special ground of distinction, however, in early Christian art does not lie so much in ornament and style as in iconography, the method of representing the scenes of the Old and the New Testaments. In the few active centres of Christianity in the East, as Alexandria and Antioch, where the new faith first took form as a religion to be defended and pro- mulgated and where there was first felt the necessity of enlisting the aid of art in its advancement, the more appealing and theologically important features of the new religion were symbolized in Hellenistic forms. When Christianity, working westward, permeated the Empire, carrying with it new forms of pictorial representation, these Hellenistic symbolic types found in the West a ready acceptance. In Rome and the West, as well as throughout the Empire until the fifth century, Christianity was occupied in its artistic creations chiefly with the expression of dogma, with the justification of the new faith in the eyes of paganism, and with the con- version and edification of the people. Therefore, with all their interest centred in the theological significance of their works and with no concern for fidelity to nature, the artists who painted the catacombs and carved the sarcophagi of the first five centuries gladly accepted these Hellenistic types as a solution of a distracting problem. While the West continued to follow in its art the meagre descriptions of Biblical scenes in the canonical Gospels, and with its increasing interest in symbolism, allegory, and analogy adhered to the established Hellen- istic types, the Christian art in the East was developing along different lines. In the first place, the liturgies of the East differed from those of the West and there were, therefore, other scenes to be figured in Oriental art. Furthermore, Oriental art was reverting to latent in- digenous characteristics. With the dissolution of the Roman Empire the classical traditions of the Hellenistic period were gradually submerged 1 73 Frohner, La verrcrie antique, p. 124. 74 Brehier, op. cit., p. 14. A SCHOOL OF IVORY CARVERS IN PROVENCE 201 in the rising tide of national feeling throughout Syria, Anatolia, and Egypt. The result of this Oriental assertion was that the East became dissatisfied with the impersonal, restrained, and somewhat abstract ac- counts of the life and teachings of the Saviour as they were related in the canonical books, which had been written under classical influence, and demanded more animate, personal, and appealing narrations of the life and miracles of Jesus and the Virgin Mary, giving rise in the East to a large body of apocryphal writings which at once re- ceived the credence of the people and consequently became popular in the art. In the East, therefore, from the liturgies which gave to the artists new feasts to be represented, from the apocryphal gospels giving more detailed and narrative versions to the episodes of the New Testament, and from the assertion of Oriental ideas of frontality and symmetry in art, there developed a cycle of Biblical scenes difitering radically in iconography from those of Rome. PROVENCAL TYPES OF EASTERN ORIGIN The first artistic impulse in Gaul, as in Rome, was Hellenistic. Again, as the sarcophagus style of Rome spread into Gaul and there was estab- lished at Arles a school of sculptors who adhered in the main to the Roman manner, the art of southern France tended to preserve the Hellenistic types even after they had begun to die out at Rome. At the same time there appear on these Gallic monuments several Eastern types which show that there was a steady infiltration of Oriental iconography from the fifth century on. It is the presence of these Eastern scenes, exe- cuted after Oriental methods, and the Eastern ornament along with the regular Hellenistic and Western types that gives to the art of Provence its peculiar eclectic character. The following scenes are examples. The Virgin in the Temple (Fig. 153), labelled by an inscription, occurs on a fifth century slab in the crypt of Saint Maximin, 75 and again on the Milan book covers and the Werden casket. From the fourth century Syria took an extremely active part in the development of Mariolatry 76 and a feast of the Virgin was introduced from the East into the Gallican church a century before the first feast of the Virgin was Le Blant, op. cit., p. 148, pi. LVII/i. 70 Lucius, Lcs origitics du culte dcs Saints, p. 603 202 A SCHOOL OF IVORY CARVERS IN PROVENCE solemnized by the church of Rome in the seventh century. 77 In fact the devotional attitude of the church in Rome toward the Virgin was very slight until after Eastern prelates, apocryphal books, and Eastern art com- menced to gain control in the sixth century. The first representation in Fig. 153. St. Maximin : inscribed slab. The Virgin in the Temple. Rome of the Virgin, alone and as an object of veneration, occurs in a sixth century fresco of Santa Maria Antiqua, 78 which was executed under Eastern influences. Therefore, while the possibilities of a Roman origin for the Virgin in the Temple are negligible, the importance which was given to her life by the Syrian church would account for the appearance of the scene at St. Maximin, a dependency of the monastery of Saint Victor founded by Cassianus, a monk who had spent much of his early life in the Orient. Joseph Assured by an Angel that he need have no fear to take Mary to wife is represented on a sarcophagus of Le Puy 79 (Fig. 154). The only other representations, with the exception of the scene on the Werden casket, which is also Provenqal, where Joseph is represented receiving an angel in his sleep, are on monuments of Egyptian origin. They are the ivory chair of Maximianus 80 (see Fig. 150), which was exe- cuted at Alexandria, and a fresco at Antinoe 81 (see Fig. 151). The Recon- 77 Duchesne, Origines du culte chretien, p. 258. 78 Griineisen, Sainte-Marie- Antique, pi. XLVI. 70 Le Blant, op. cit., pi. XVII/4. This event is related in the Gospel of Matthew (i, 20), in the Protevangelium of James (Michel, Evangilcs apocryphes, p. 29), and in the Gospel of Pseudo-Matthew (Michel, op. cit., p. 91). 80 Garrucci, Storia, VI, pi. 417/3. 81 Leclercq in Cabrol's Dictionnaire d’arclicologie chretienne et dc liturgie, s. v. “Ane,” col. 2060, fig. 599. . A SCHOOL OF IVORY CARVERS IN PROVENCE 203 ciliation of Joseph and the Virgin, which also occurs on this same sarcopha- gus, from its dependence solely on the apocryphal sources and its unique character in the West at this time, is also evidence of Oriental iconography in Gaul. The Appearance of the Star to the Three Magi, as a scene separate from the Nativity, is depicted on two sarcophagi of Arles 82 (see Fig. 26), on the Milan book covers and on the Werden casket. The Roman church did not celebrate the appearance of the star, but in Syria it played an im- portant part in the feast of the Epiphany on January 6. 83 The Massacre of the Innocents, which appears on a sarcophagus cover Fig. 154. Le Puy: sarcophagus. Joseph REASSURED BY THE ANGEL. RECONCILIATION OF Joseph and Mary. at St. Maximin 84 (see Fig. 56), as well as on the Berlin ivory and the Milan book covers, points conclusively to Eastern inspiration. There was no feast of the Holy Innocents in Rome until at least a century after their remains, brought from the Orient to Marseilles by St. Cassianus at the beginning of the fifth century, had been worshipped at Marseilles and in the neighboring towns. Cassianus deposited the relics of the Holy Innocents in the Church of St. Victor at Marseilles, which he founded after 414, but distributed portions of them in neighboring churches of Provence. 85 Among the churches favored in this way we may certainly include that of St. Maximin, which was itself a de- pendency of St. Victor. The style of the sarcophagus cover points to 82 Le Blant, Les sarcophages chreticns dc la ville d’Arles, pis. XXI, XVIII. i? - See Appearance of the Star to the Magi, p. 34. 84 Le Blant, op. cit., pi. LVI/i. 8 " See p. 67. 204 A SCHOOL OF IVORY CARVERS IN PROVENCE the first half of the fifth century, and it is likely that the scene of the Massacre of the Innocents which adorns it commemorated the advent of relics of the Holy Innocents at St. Maximin. The scene of Christ and the Samaritan Woman at the Well was frequently represented on the monuments both of the East and of the West. The Hellenistic and hence the Roman method was to depict the Saviour standing at the side of the well; while the Eastern method, the Syrian, Palestinian-Coptic, and Byzantine, figured Him seated . 86 On three sarcophagi of Provence 87 Christ is represented in the traditional Roman manner standing at the side of the well ; but on a sarcophagus of Narbonne 88 He is seated after the Eastern manner, showing the existence in Gaul of the two distinct iconographical traditions. Ananias Being Carried Out is represented on a sarcophagus of the Abbey Saint-Ruf . 89 Although only a fragment, the scene is practically the same as that on the Brescia casket which Strzygowski attributes to Asia Minor . 90 The fact that these are the only two representations of the scene in eSily Christian art makes the analogy more striking. 86 The scene of Christ and the Woman of Samaria adheres to the same iconographic classification as the other scenes treated in the first part of this volume. The Hellen- istic type, which was only concerned with symbolism, disregards the Biblical account which says that Jesus sat down at the well (John iv, 1-26) and persistently represents Jesus and the woman standing at either side of the well. In the Oriental-Hellenistic type the same method of representation is preserved on the Brescia Casket and the gold encolpium in the Museum at Constantinople. The Alexandrian-Coptic type, which was a native adaptation of the Hellenistic iconography of Alexandria, shows Christ standing but introduces the Eastern form of well house. The examples of this type occur on the Maximianus chair and pyxides in the Cluny Museum, the Basilewsky, Youlgrave, and Uwaroff Collections. The actual Eastern, or Syrian rendering which later became the Byzantine type, follows the Biblical account and depicts the Saviour seated at one side of the well ; this method appears in the Rabula Gospels, on an ivory cover in the Bibliotheque Nationale at Paris, and a pyxis in the Louvre. It occurs in the West not only on two sarcophagi of Gaul but also in the mosaics of the Baptistery at Naples and the church of Sant’ Apollinare Nuovo at Ravenna, and on the ciborium columns of San Marco. After the seventh century this type with the seated Christ appears to have become the customary method of representation in the East and occurs later in the Italo-Byzan- tine monuments. As a souvenir of the traditional Oriental iconography, there is pre- served today at Constantinople, as one of the rarest relics of Hagia Sofia, the rock upon which Christ sat when He conversed with the Samaritan woman (Le Blant, Scire, chret. de la Gaule, p. no). 87 Sarcophagus of Arles (Le Blant, Save, chrflt. de la villc d’ Arles, pi. XVIII/2), of Clermont (Le Blant, Sarc. de la Gaule, pi. X\ III/2), of Nimes (Le Blant, op. cit., pi. XXX/4). 88 Garr., op. cit., V, pi. 402/7; Le Blant, op. cit., pi. XLVII/2. 89 Le Blant, op. cit., pi. VIII/4. 90 Garr., op. cit., VI, pi. 444 ; Strzygowski, Kleinasien, p. 213. A SCHOOL OF IVORY CARVERS IN PROVENCE ^05 In further confirmation of the Oriental influence on Provencal iconog- raphy, Schonewolf points out the Asia Minor type of the young Christ with long hair falling in locks over His shoulders, on a sarcophagus of Nimes . 91 Furthermore, on a sarcophagus of Marseilles 92 and on one of Arles 93 occurs the Palestinian type of the seated Christ, bearded and with long hair. The means by which these Eastern types passed into Gallic iconography were doubtless the ivories and other objects of the minor arts which af- forded convenient articles of import to the Oriental traders. Vestments, ornamented with scenes of the Gospels, also played their part . 94 Syrian influence in iconography may perhaps be recognized in the description by Gregory of Tours of a Crucifixion in the Genesius church of Narbonne wherein Christ is raised naked on the cross after the Oriental-Hellenistic type 95 The apparent scandal that this exhibition of an unclad Saviour caused in Provence at the end of the sixth century would suggest that the painting at Narbonne antedated such an antipathy and was probably done in the early part of the fifth century when this rendering, as seen on the doors of Santa Sabina at Rome and on the ivory plaque in the British Mu- seum, which I would attribute to Provence, was the prevalent type . 96 At the same time, the marked aversion of the populace by the close of the sixth century to the naked Saviour also suggests that the colobium type of the Crucifixion, which depicts Christ clad in a long garment, as is seen in Oriental Crucifixions of that period, had been introduced into Provence and was the recognized and preferred manner of representation. To sum up: During the first seven centuries of our era there was in Provence an increasing invasion of Syrian merchants, an influx strong enough in some cases to control many of the principal towns. With these merchants came the missionaries and monks who founded churches and monasteries on Eastern models and kept alive a relation with the Christian East, quite as intimate as that with Rome. At the same time that the commercial and religious expansion was spreading over Gaul and Provence, 91 Schonewolf, op. cit., p. 40. 92 Garr., op. cit., V, pis. 341/4. 93 Garr., op. cit., V, pi. 342/3. 94 Brehier, op. cit., p. 28. 9 ’ Gregory of Tours, Dc Gloria Martyrum, XXIII. 99 See p. 187, note 1. 206 A SCHOOL OF IVORY CARVERS IN PROVENCE Oriental craftsmen and Eastern art objects, such as ornamented reliquaries, embroideries, and carved ivories, were entering the country through Mar- seilles, Narbonne, and the other ports of Provence. Therefore, the art of Provence, traditionally Gallo-Roman and Hellenistic, whose initial style was that of the sarcophagus carvers of Rome, gradually turned, even on the sarcophagi themselves, to Eastern ornament, peculiar to the churches of Syria, and Eastern methods of representing the scenes of the Gospels. This art, however, was an eclectic one and the Oriental characteristics ap- peared only as interpolations in a Latin setting. It is apparent that Provence was a region where the ideas and customs of the East and West met freely and where the mark of Oriental influence was clearly defined. The society in Provence, so largely Syrian, the East- ern character of the church, and the constant introduction of Eastern art objects, made it a centre peculiarly equipped to execute the group of ivory carvings to be discussed in the following pages. In the discussion of the early Christian scenes in Part I, I have already noted the peculiar and Provencal character of the iconography found on these ivories; I shall now assemble tbe data thus afforded, studying the ivories separately and considering the Provencal school as a whole. Ill IVORY BOOK COVERS IN MILAN CATHEDRAL Book covers Milan Date c. 500 A. D. Cathedral In Milan cathedral are two ivory book covers (Figs. 155, 156) which are the most interesting and important examples of this disputed group of ivory carvings. The scenes depicted on these panels are all drawn from either the canonical or apocryphal accounts of the life of Christ and offer an extensive range of iconography. At the centre of the first cover (Fig. 155) is a rectangular panel, formed by a delicate acanthus border, in which is carved a wreath rich with various fruits and flowers and bound at the bottom by a ribbon the ends of which flutter off, like scrolls, into the two lower corners of the panel. Within the wreath is the Lamb of God, crowned with a laurel nimbus and executed in cloisonne enamel. In the background, behind the lamb, appear the outlines of two pilaster strips. A SCHOOL OF IVORY CARVERS IN PROVENCE 207 Fir,. Milan : Cathedral, ivory hook cover. 208 A SCHOOL OF IVORY CARVERS IN PROVENCE capped with Corinthian capitals, supporting an unadorned lintel which, in its turn, bears a cornice carved with acanthus leaves. Across the top of this cover is a long panel with the Nativity in the centre and, at either corner, a symbol of an Evangelist framed by a wreath. Reading down, the scenes on the left are the Annunciation to the Virgin at the Spring, the Three Magi Gazing at the Star of Bethlehem, and the Baptism of Christ. The scenes on the right, following the same order, read : The Virgin in the Temple, Christ among the Doctors, and the Entry into Jerusalem. In a panel extending across the bottom of the cover is depicted the Massacre of the Innocents before Herod, and in the corners are busts, framed in wreaths, of the two Evangelists whose symbols appear in the two upper corners. The second cover (Fig. 156) has the same middle panel with the acan- thus border and pilaster strips; between the pilasters stands a cross in cloisonne and from the lintel hang curtains which are caught back to the pilasters. The cross is set upon Mt. Zion from the base of which flow the four rivers of life. Within the horizontal panel at the top is the Adoration of the Magi and a symbol of an Evangelist in each corner, the lion for Mark and the eagle for John. The scenes on the vertical panel at the left, reading down, are Christ Healing the Blind, the Paralytic, and the Raising of Lazarus. On the right, following the same order, are Christ (seated upon a globe) Presenting Wreaths to Two Saints or Apostles, the Last Supper, and the Widow’s Mite. The Miracle of Cana with the heads of two Evangelists in the corners Alls the bottom panel. The iconography of the scenes on the two covers, regarding them in sequence, shows a peculiar mingling of Eastern and Western methods of rendering the Gospel events. This eclectic character of the iconography is consistent to a marked degree with the eclectic character of the orna- ment and iconography on the sarcophagi of Provence and in a general way with the constant infusion of Oriental forms, customs, and religious rites into the more native Roman traditions of southern Gaul. It has already been shown how freely the ideas and customs of the East and West met and mingled in Provence and how the society in Provence, so largely Syrian, the Eastern character of the church, and the constant intro- duction of Eastern art objects, made it a centre peculiarly adapted to execute a group of ivories whose motifs and iconography present a curious A SCHOOL OF IVORY CARVERS IN PROVENCE 209 3*^ vW'WKWMi)4-\A y-W vV SwVivv ^ / ».' 1 ‘/try A, in K , A^77ynr^ Fig. 156. Mi ATHEDRAL, IVORY BOOK COVER. 210 A SCHOOL OF IVORY CARVERS IN PROVENCE mixture of Oriental and Occidental methods. 1 Moreover, several of the scenes on the covers are compositions peculiar to the monuments of Provence. As one would naturally imagine in a region where Syrian influence was dominant, the scenes dealing with the life of the Virgin and the infancy of Jesus are based on Eastern apocryphal gos- pels instead of on the canonical books, which were faithfully adhered to under the Roman church during the first five centuries of Christian art. While many of the scenes are based upon apocryphal sources, they do not appear to follow the purely Oriental books but conform more closely to the Gospel of Pseudo-Matthew. 2 This partial dependence of the artist of the Milan covers on a semi-Oriental apocryphal gospel not only aids in explaining the presence of Eastern iconography in the West but is also consistent with the heterogeneous racial and religious elements in Provence which would readily have received a gospel like Pseudo-Matthew. I shall here review in brief the iconographic features of the book covers, which have been developed more fully in Part I. The iconog- raphy of the scenes on the covers is principally Western and for the most part may be connected with Provence. Intermingled, however, with Occi- dental types are Oriental features which, as I pointed out in the Intro- duction, have given rise to the divergent opinions among the writers on these covers as to their place of origin. 3 If my classification of the types of the various scenes, which I have sought to sustain in the first part of this volume, is accepted, the Western origin of the Milan book covers is unquestionably demonstrated. Of the sixteen scenes appearing on the ivories seven, the so-called Nativity, the Adoration of the Magi, the Heal- ing of the Paralytic, the Raising of Lazarus, the Presentation of the Crowns, the Entry into Jerusalem, and the Last Supper, are types peculiar to the Christian art of the Occident at the end of the fifth century. In contrast to these Western types the rendering of the Healing of the Blind is a Syrian type, while the Miracle of Cana and the Last Supper present in their compositions transitional types marked by the introduction of Eastern elements into Western renderings of the scenes. At the end of the fifth century, moreover, when the Milan covers were carved, such scenes as the Virgin in the Temple, the Annunciation at the Spring, and the Appearance of the Star to the Magi could only have been executed 1 See p. 192 sq. 2 See pp. 8, 12, 18, 35, 70. 3 See p. 188. A SCHOOL OF IVORY CARVERS IN PROVENCE 21 I in a community where the influence of the Eastern church was strong and where Oriental forms were popular. Moreover, five of the scenes, the Virgin in the Temple, the Annunciation at the Spring, the Adoration of the Ox and the Ass, the Appearance of the Star, and Christ among the Doctors, are based on the Latin Gospel of Pseudo-Matthew which, up to the end of the fifth century, shared with other apocryphal gospels the disapproval of the church at Rome. Of the other scenes on the Milan covers, the Massacre of the Innocents and the Baptism are distinctly Pro- vencal. These last two scenes, together with the Virgin in the Temple, the Adoration of the Ox and the Ass, the Appearance of the Star, and the Adoration of the Magi, occur as types or show details peculiar to the group of ivories that I would attribute to Provence and are represented in the form which they assume on monuments either found in Provence or stylistically connected with that region. These six scenes, by their dis- tinctive composition and rendering, isolate the ivories as products of a particular school whose most important product remaining to us is the diptych in the Cathedral of Milan. The evidence for the Western origin of the covers is amply conclusive: the representation of what is generally recognized as the Nativity but is really a scene of the Adoration of the Ox and the Ass 4 is a Western type, since the event is depicted in a shed instead of in the open as on all East- ern examples ; 5 the Adoration of the Magi is to be classified as Western from the fact that the Magi are beardless ; c the Healing of the Paralytic, wherein the cured man is depicted in profile holding his bed upside down on his back as he faces the Saviour, conforms exactly to the type used in the catacombs and on the sarcophagi of Rome ; 7 the Raising of Lazarus in which Christ touches with His wand the figure of Lazarus who, swathed like a mummy, stands with his head covered in the doorway of a little aedicula set upon a podium while one of the sisters of Lazarus kneels at his feet, is peculiar to the representations on Western sarcophagi and does not occur on Eastern monuments : 8 Christ enthroned on a globe, both in the scene of the Presentation of the Crowns and in the Widow’s Mite, is, as far as existing examples show, an Italian conception appearing as early as the fourth century in Rome 9 while the Presentation of the 4 See p. 19. 7 See p. 104. ’ See p. 22. * See p. 1x2. G See p. 38. r ‘ See p. 142. 212 A SCHOOL OF IVORY CARVERS IN PROVENCE Crowns was common to the monuments of Rome, Ravenna, and Gaul; in the Entry into Jerusalem Christ rides astride after the manner of representation in the West and in distinct contrast to the Eastern ren- dering where He is always depicted riding sidewise after the Oriental fashion ; 10 and finally the Last Supper sustains the Western origin by the retention in the composition of the loaves and fishes, which, as symbols of the Eucharist, replace the bread and wine of the actual event which were represented in the East . 11 At the same time that the Western origin of the covers is clear, certain iconographic types and features of composition appear that betray a dis- tinctly Eastern origin and show that the ivories were executed at a time when Hellenistic symbolism in the West was breaking down under Eastern influence and that they were done in a community where Oriental and Occidental types might naturally mingle and tend to produce an eclectic art. In the first place the Healing of the Blind, where two men instead of one present themselves to Jesus for aid, is an Eastern type . 12 The Miracle of Cana shows a transitional type wherein the simple and highly symbolic Hellenistic type begins to break down under the Oriental tendency to give a more narrative rendering to Biblical events and to make of this scene an actual representation of the wedding feast; while the figure of Christ is represented touching with His wand the jars of water, after the symbolic method in the West, a narrative interpretation has been given to the scene by the introduction of a servant pouring the water into the jars and by increasing the number of spectators from one to eight . 13 In much the same manner the Last Supper, while manifestly of Western origin, shows again a transitional type in that the scene is represented as an actual event after the Eastern manner while Western characteristics appear in the retention of the symbolic loaves and fishes and the abbreviation of the number of disciples to three . 14 The Eastern influence in the covers is also shown by the reflection in certain scenes of the Eastern liturgy. While no examples of the Virgin in the Temple appear on Eastern monuments before the eighth century, the relation of the scene with the worship of the Virgin in Syria where a particular liturgy had been conceived in memory of the life that Mary spent in the temple, ]1 See p. 136. 14 See p. 136. 10 See p. 124. 13 See p. 91. 12 See p. 100. A SCHOOL OF IVORY CARVERS IN PROVENCE 213 the fact that the event is only mentioned in the apocryphal gospels, and the absence of any feast of the Virgin in Roman liturgies before the seventh century, present evidence that such a scene could have been carved only in a community where Eastern religious forms were observed. 15 The Annunciation to the Virgin at the Spring is also an inherently Orien- tal scene, for it again is only related in the apocryphal gospels, and in subsequent art appears only on the monuments of the Orient. 16 The Appearance of the Star is an Oriental motif, for this scene had no liturgical significance in the Roman church as it had in Syria, where on the night of January 6, besides the Birth and the Adora- tion of the Shepherds, the Appearance of the Star was especially cele- brated. 17 The dependence, moreover, of the Virgin in the Temple, 18 the Annunciation at the Spring, 19 the Adoration of the Ox and the Ass, 20 the Appearance of the Star, 21 and Christ among the Doctors 22 on the Latin apocryphal Gospel of Pseudo-Matthew indicates an origin for the ivories in a mixed community that was basically Western but where East- ern forms were popular. 23 It must be granted, then, that the Milan covers present an odd mixture of Oriental and Occidental iconography and could have been produced only in a community, such as I have shown Provence to have been, where Eastern and Western forms freely intermingled. Reviewing now those features of the iconography which connect the covers most directly with Provence, the Massacre of the Innocents is above all others the most con- vincingly Provencal of the scenes on the covers. Besides connecting the covers with southern Gaul and probably with the church of St. Victor at Marseilles, it is the one scene that most closely binds together the group of ivories. The Massacre of the Innocents does not appear either in the catacombs or on the sarcophagi of Rome. 24 Of the twelve examples of the scene which occur before the ninth century, four are Eastern, one is Roman and strikingly different from all the rest, and the remainder are to be connected with Gaul. The representations from Syria and Egypt in the Rabula Gospels, the frescoes of Antinoe, a pyxis in the Louvre, and a censer at Cairo, as well as the scene on the arch mosaics of Santa Maria 15 See p. 10. 16 See p. 11. 17 See p. 33. 15 See p. 9. 19 See p. 12. 20 See p. 20. 21 See p. 35. 22 See p. 70. 23 See p. 191. 24 See p. 60. 214 A SCHOOL OF IVORY CARVERS IN PROVENCE Maggiore at Rome, all show the sword type wherein the soldiers kill the children with their swords. On the Milan ivory, however, as on the other ivories of the group, is the smashing type where the soldiers hurl the innocents to the ground. The first example of the smashing type is on a sarcophagus cover of the early fifth century in the crypt of St. Maxi- min in Provence. In the sixth century the type occurs in the miniatures inserted into a Carolingian manuscript at Munich and in the ninth century appears on two Carolingian book covers. The type was thus traditional in Gaul and the earliest example we know was found in the famous crypt of St. Maximin whither pilgrims from all parts of Gaul were wont to travel, to do reverence before its relics. The original provenience of the smashing type may be located by the close connection of the church of St. Maximin with the abbey church of St. Victor at Marseilles. The Holy Innocents were worshipped from ancient times at Marseilles and their relics were brought from the Orient to that city by Cassianus. It is likely, then, that in the church of St. Victor, whence the cult in Provence emanated, the smashing type originated ; and it is even possible that the Milan covers were executed under the auspices of this same abbey in Marseilles. The distinctive composition of the Baptism and its presence on the Werden casket, the Bodleian book cover, and the Berlin fragment, as well as on the Milan covers, places the origin of the whole group in that region where any one of the ivories may be shown to have originated. Since no doubt can be expressed as to the Western and Gallic origin of the Bodleian ivory 25 the fact that the same singular composition of the Baptism, as well as of the Massacre of the Innocents, occurs on that ivory as on the Milan covers, offers one proof of the Gallic origin of the Milan ivories. The features of the Baptism which are uniformly pre- served on all the ivories and distinguish the scene from other representa- tions are: the small Christ with long hair, the bearded John the Baptist wearing an exomis and carrying in his hand a shepherd’s crook, and the waters of the Jordan piled up behind Christ like a waterfall. 26 The Western character of the scene on the Milan ivory is evinced by several details : Jesus is immersed in the water only to His ankles after the manner of representation on the Western sarcophagi while on the Eastern 25 See p. 248 sq. 26 See p. 76. A SCHOOL OF IVORY CARVERS IN PROVENCE 2I 5 monuments He is waist deep in the Jordan; He is small of stature, which is a more common feature of Western than of Eastern scenes; the exomis which John wears instead of the pallium is more frequently used in the West than in the East . 27 The characteristic of the composition, however, which distinguishes the scene from all others and marks the group is the pedum, or shepherd's crook, which John carries. The only other example of the pedum in a scene of the Baptism is in a mosaic of Santa Maria in Cosmedin at Ravenna where the composition in other details is quite different. Another detail that clearly connects the scene with Provence is the manner of representing the waters of the Jordan as a waterfall. The waterfall occurs on no Eastern representations and appears on only one Roman sarcophagus. On the other hand it is a feature of the Baptism on two sarcophagi of Arles, on one of Soissons, one of Mas d’Aire, and on another of Madrid. With the exception of the pedum, the features of the composition on the Milan cover are those of the Baptisms on the sarcophagi of Provence. Besides the last two scenes, the Virgin in the Temple, the Adoration of the Ox and the Ass, the Appearance of the Star, and the Adoration of the Magi, which all occur on two or more of the ivories of the group, may be attributed to Provence. In the first place the Virgin in the Temple, that appears on both the Milan and Werden ivories and nowhere else in the early Christian art of either East or West, occurs in an abbreviated rendering on a fifth century slab of the crypt of St. Maximin which at once associates the scene with St. Victor at Marseilles. The representation of what is generally recognized as the Nativity but is really a scene of the Adoration of the Ox and the Ass, besides occurring in identical manner on the Milan, Werden, and Nevers ivories of the group and later on the Carolingian book cover in the Bodleian which was copied from the diptych of which the Nevers fragment was possibly a part, may be compared to a unique representation of the same scene on a sarcophagus of Arles . 28 The scene on the sarcophagus of Arles, in that the Magi are also figured adoring the Child, is a frank com- bination of the two adorations in a symbolical sense without any con- nection with the Nativity. That the same curious combination of the two adorations was in the mind of the ivory carver is shown by the way in 27 See Table IV. 28 See Fig. 14. 2 1 6 A SCHOOL OF IVORY CARVERS IN PROVENCE which he associated the two scenes; the Adoration of the Ox and the Ass, coming out of chronological order, occupies on one cover the large top panel while the same place is assigned to the Adoration of the Magi on the other cover. The links that bind the Milan scenes to Provence are further strengthened by the Appearance of the Star to the Three Magi. 29 While this is combined with the Adoration on Roman monuments, it appears as a separate scene on the Milan covers, and in like manner on two sarcophagi of Arles. 30 Its rendering on the ivory, where the Magi wear odd notched chitons and look with marked gestures of surprise at the star above their heads, is identical, even to the cut of the chitons and the specific gestures of the Magi, with the sarcophagi representations. It is nearly identical, again, with the scene of three young Hebrews before the bust of the Assyrian king on a sarcophagus of St. Gilles. 31 Besides the Appearance of the Star, which more firmly estab- lishes the attribution of the Milan covers to Gaul, it is also to be noted that the very rare representation of the Widow’s Mite, which appears on the covers, was seen in the fifth century by Perpetuus on the walls of the church of Saint Martin at Tours. As to the other scenes, which I have not at- tempted to attach directly to Provence, reference to the iconographical discussions in the first part of the book will show that none of them hinders the attribution and many of them show special details which are more similar to the scenes on Provengal sarcophagi than to those of Rome. The eclectic character of these ivory covers, whose iconography is so largely that of Roman monuments and whose scenes yet present so many marks of Eastern influence, reflects quite naturally the eclectic nature not only of the art of Provence but also of the racial, social, and religious characteristics of the region. On the other hand, the ornament on the Milan book covers manifests no conclusive arguments for the Provencal origin of the covers although it fully sustains the attribution. At first sight the curious use of orfevrerie cloisonne e would seem to complicate the question of locating the ivories. In fact it has given rise to the suggestion that the lamb and cross may have been added some centuries after the covers were carved. This ob- vious solution of the difficulties is unsatisfactory, for it does not account 29 See p. 34- 30 See Fig. 26. 31 See p. 35. A SCHOOL OF IVORY CARVERS IN PROVENCE 217 for the framed spaces which the absence of the lamb and cross would have left vacant, and it should be resorted to only when it is proved that such gold work at the beginning of the sixth century was impossible. To Strzygowski, who believes that orfevrerie cloisonnee originated in Ar- menia, 32 its presence on the covers would tend to prove that they came from Asia Minor. 33 The Eastern origin of cloisonne is readily admitted. Both Molinier 34 and Kondakoff 35 agree in attributing its inception to Persia; and it was probably known and used in the Christian art of Constantinople as early as the time of Constantine. 36 There was, however, a marked difference between the technique of the cloisonne in gold that was brought into Gaul, as Molinier 37 and Lasteyrie 38 assert, by the barbarians from the north and the Byzantine process, which was later introduced from the south. The Northern method consisted in a cloisonne setting of either semi-precious stones or bits of colored paste in gold, while the Byzantine process was the usual pure enamel cloisonne in which the enamel was spread into small spaces of the design outlined by fine gold wires and was then polished down to a perfectly smooth surface. Inasmuch, then, as the cross on the Milan covers is made of square and circular stones set in gold, and the lamb, according to Dalton, 39 is made up of stones which have not been polished down to a flat surface, the use of this Northern and un-Byzantine process is consistent with the Provengal origin of the covers. Moreover, the “barbaric art” of gold work found in the Merovingian graves of France is in many cases almost identical with the ornament on the ivories. The tomb of Childeric (d. 481) at Tournay contained the sword of a chieftain which has a hilt and handle decorated with a cloisonne not unlike that of the lamb on the Milan cover. 40 At Gourdon there was discovered a flat dish ornamented in the middle with a cloisonne 32 Van Berchem and Strzygowski, Amida, p. 352; Dalton, Byzantine Art and Archaeol- ogy, p. 59, note 1. 33 Strzygowski, Kleinasien, p. 198. 34 Molinier, L’orfdvrerie religieuse et civile, p. 17. 35 Kondakoff, Histoire et monuments des emaux byzantins, 1892, p. 63. 33 Dalton, Byz. Art and Arch., pp. 495-496. 37 Molinier, op. cit., p. 18; Marignan, Louis Courajod, p. 112. 33 Lasteyrie, Tresor de Guarrazar, Paris, i860. Note, however, the strong case made out by Riegl, in his Spatromische Kunstindustrie, for the Roman origin of Granatanlagen. 39 Dalton, op. cit., p. 57. 4 Molinier, in Michel’s Histoire de I’art chret., I, p. 407. 2l8 A SCHOOL OF IVORY CARVERS IN PROVENCE cross composed of alternating pieces of circular and rectangular stones or glass which is similar to the cross on the book covers. Other excavations scattered over France 41 have disclosed numerous examples of this work whose fabrication may be dated from the fifth to the seventh century . 42 There is also the testament of Perpetuus, which tells of the gold work which was done in Gaul as early as the fifth century , 43 and the evidence of Gregory of Tours that beautiful pieces of gold work set with stones existed in Narbonne and Avignon . 44 While Labarte would show that the earliest examples of this work in Gaul came from Constantinople, he points out that Gaul in the early Christian times had already acquired a great reputation in the art of goldwork and the cities of Arles, Rheims, and Trier had ateliers where work was done for the emperors . 45 It therefore seems justifiable, as far as the cloisonne work on the covers bears on the attribution, to assume that the Milan covers were executed in Provence for, if anything, this particular form of ornament appears to sustain the attribution. Even the point made by Strzygowski, that the lamb and the cross were Palestinian motives, does not bear on the ques- tion for Schonewolf has already remarked Syro-Palestinian influence on the Gallic sarcophagi in the frequent use of the “crux gemmata .” 46 In fact, even if these motives be Eastern, the corresponding representation of the lamb on a sarcophagus from St. Maximin 47 would help confirm the attribution to Provence. Aside from the cloisonne, the ornament on the Milan book covers is not particularly instructive in locating the ivories. The Corinthian capitals and the acanthus mouldings which frame the panels are simply a debased rendering of classical motives such as might be found in any artistic centre where Hellenistic traditions still survived. In the central panel the two curtains draped back to the columns, while a common enough motive in the East, was not very common on the monuments of Rome and seldom appears on the Roman sarcophagi. At the same time this motive occurs 41 Molinier, op. cit., I, pp. 408-410. 42 Molinier, op. cit., p. 410. 43 Labarte, Histoire dcs arts industrials, I, p. 416. 44 Labarte, op. cit., pp. 420-423 ; Gregory of Tours, Hist. Franc., Ill, 10; VII, 36 sq. 45 Labarte, op. cit., p. 415. 46 Schonewolf, Die Darstcllung dcr Auferstehung Christi, p. 32. 47 Le Blant, op. cit., pi. LVI/i. A SCHOOL OF IVORY CARVERS IN PROVENCE 219 frequently on the sarcophagi 48 of southern Gaul which show marked East- ern influence, including a sarcophagus of St. Maximin, and became later one of the most distinctive features of Carolingian art. While the symbols of the Evangelists no doubt spread from Egypt and originated 49 from the liturgical cycle of Egypt, their use as personal em- blems with the figures of the Evangelists, as they occur on the Milan covers, was Western and their first appearance with the Evangelists oc- curred in a mosaic of the Baptistery of St. John Lateran.'" A very cogent argument against the suggested origin of the Milan covers in Asia Minor is the fact that in the Rossano Gospels from Anatolia the writing figure of the Evangelist Mark appears without the emblem. The costumes of some of the figures on the ivories are either similar to the traditional dress of Gaul or like the dress seen on certain Gallic sarcophagi. During the early centuries of Christian art the form of dress was Hellenistic. In the sixth century, however, the rich Byzantine costumes of the Orient already appear in art. The first repre- sentation in Rome of the Madonna in the rich robes of a Byzantine em- press is found in the sixth century at Santa Maria Antiqua and in Ravenna we see these new forms of dress in the mosaics of Sant’ Apollinare Nuovo, and Sant’ Apollinare in Classe. Marignan, in discussing the various aspects of Merovingian society , 51 writes that the Gallo-Roman women wore the tunic with the long, straight sleeves, the dalmatica, the paenula, and the pallium; these different vestments were generally ornamented with embroidery. According to the same authority , 52 the women also wore their hair either hanging loose upon their necks or bound in a knot on the top of their heads. W 7 ere it not for this description of the costume of the Gallic women, which so accurately accords with the dress of the young Virgin, in the two scenes of the Virgin in the Temple and the 4 - Le Blant, op. cit., pis. XIX/i (Clermont), XXXIV (St. Guilhem-du-Desert) , XXXVII, XL/i (Toulouse), XLVIII/3 (Le Mas-Saint- Antonin), and LV/3 (St. Maximin). 4S Bye. Zcit., XV, p. 702. Garr., op. cit., IV, pi. 239. While the mosaic is now destroyed, this method of de- picting the Evangelists became customary in the West. In the East, however, the artists sometimes depicted the four symbols alone, often combining them in a single figure known as the Tetramorph, “but never, during the great period, associating them severally with the Evangelists’ - (C. R. Morey, East Christian Paintings in the Freer Collection, 1914. p. 39). 51 Marignan, La Socicte Merovingienne, 1899, p. 127. r ' 2 Marignan, op. cit., p. 129. 220 A SCHOOL OF IVORY CARVERS IN PROVENCE Annunciation by the Spring, on the Milan covers, one would almost think her costume Byzantine. It accounts for the modish manner in which her hair is dressed and for the heavy embroideries which border her garments. Regarding the male garments as they appear on the covers, it is interesting to note that the costume of the soldiers of Herod in the Massacre of the Innocents is identical with the costume of the soldiers in the same scene on the sarcophagus of St. Maximin and is preserved in Merovingian times in the miniatures of the Codex Purpureus of Munich and also in the Judg- ment of Solomon on a Carolingian ivory in the Louvre. 53 Besides the dress of the Magi, whose similarity in costume to the Magi on Provencal sarcophagi has already been mentioned, the costumes of the other figures are either the regular Hellenistic dress used in the art of the period or the traditional dress for particular figures. The date of the Milan book covers may be estimated fairly closely to be about the beginning of the sixth century. The earliest date which has been assigned is the fifth century and the principal authorities to sus- tain it are De Rossi, 54 Dobbert, 55 and Stuhlfauth. 56 Bugati 57 held that they were executed about the end of the fifth century or the beginning of the sixth century; Kondakoff 38 and Schmid 39 assign them to the beginning of the sixth century; Labarte, 60 Westwood, 61 and Dalton 62 attribute them to the sixth century while Grimouard de Saint-Laurent 63 puts them as late as the close of the same century. The average then, of the dates which have been assigned to the covers would lie between the close of the fifth century and the beginning of the sixth century. Such a date, c. 500 A. D., may be sustained on other grounds than the appear- ance of the ivories and the vague criteria of style. 53 Hermann Hieber, Die Miniaturen des frilhen Mittelalters, fig. 53, illustrates the manuscript and Molinier, Hist, des Arts Appliques, I, Ivoires, pi. XIII, the ivory. 64 De Rossi, Bull., 1865, p. 26. 53 Dobbert, Repertorium, VIII, 1885, p. 173. 56 Stuhlfauth, Die altcliristliche Elfenbeinplastik, pp. 70-71. 5T Bugati, Memorie storiche, etc., 1782, p. 280. 58 Kondakoff, Hist, ct nvon. des emaux byz., 1892, p. 78. 59 Schmid, Die Darstellung der Geburt Christi, p. 111. co Labarte, Histoirc des arts industriels, I, p. 43. 61 Westwood, Fictile Ivories, no. 95, p. 38. 02 Dalton, Byz. Art and Arch., p. 202. 63 Grimouard de Saint-Laurent, Revue de l’ art chretien, XV, 1872, p. 408; to these should be added: Strzygowski, Kleinasien, p. 198; Venturi, Storia, I, p. 509: Molinier, Hist, des arts appliques, I, Ivoires, p. 61 ; Haseloff, lb. Preuss. Kunsts., XXIV, p. 49. A SCHOOL OF IVORY CARVERS IN PROVENCE 221 In the first place, most of the scenes on the covers dealing with the early life of Christ and the Virgin are based upon the account in the Latin Pseudo-Matthew. By the end of the fifth century the popularity of the Eastern apocryphal gospels in the West had become so great that Pope Gelasius (492-496) issued a decree against their use, which was recalled under the following pontiff. The Gospel of Pseudo-Matthew, clearly the source of many scenes on the Milan covers, was compiled during the course of the fifth century, according to the general view, and probably, as I have pointed out, at the beginning thereof. 64 Furthermore, the first feast of the Virgin was not introduced from the East into Gaul before the sixth century and it is doubtful if a work like the Milan cover, where two special scenes of the Virgin are repre- sented, would have been executed much before her cult was established. That the covers could not have been carved very late in the sixth century is shown not only by the fact that the style is of the early sarcophagi of the West but also because they present so many types of scenes peculiar to the sarcophagi of the fourth and fifth centuries and show only traces of Eastern iconography. Also, by the fifth century in Provence we have seen this same mingling of Eastern and Western iconography and orna- ment on the sarcophagi from which some of the peculiar scenes of the covers, as the Virgin in the Temple, the Massacre of the Innocents, the Appearance of the Star, and the Entry into Jerusalem, were derived. From the dependence of the covers on the Latin Pseudo-Matthew, from the Virgin scenes, and from their relation with the best sarcophagi types of the fifth century, I would, therefore, place the execution of the Milan book covers about 500 A. D. IV WERDEN CASKET IN THE VICTORIA AND ALBERT MUSEUM Casket South Kensington Date. . .500-525. Victoria and Albert Museum. In the \ ictoria and Albert Museum in South Kensington there are pre- served three sides of a small ivory casket (Figs. 1 5 7- 159) which came from 4 See p. 191. 222 A SCHOOL OF IVORY CARVERS IN PROVENCE the abbey of Werden in Westphalia. 1 Commencing with the long panel (Fig. 157), which was probably the front of the casket, and reading to the left, the scenes run as follows: the Virgin in the Temple, the Visita- tion, the Departure of the Virgin from the Home of Elizabeth, Joseph’s Dream, and the Annunciation by the Spring. The short panel of what was the left hand side of the box is lost but, from the sequence of the scenes and from their similarity to the scenes on the Milan covers, it must have borne the representation of the Massacre of the Innocents. On the other long panel (Fig. 158) occur the Appearance of the Star to the Magi, the Adoration of the Ox and the Ass, and the Adoration of the Magi. Fig. 157. S. Kensington: Victoria and Albert Museum, IVORY CASKET FROM WERDEN. FRONT. Finally, on the short panel (Fig. 159) of the right side of the casket are depicted the coming of the Pharisees and Sadducees to John the Baptist at the Jordan, the Laying of the Ax to the Unfruitful Tree, and the Baptism of Christ. The principal reason for connecting the Werden casket with Provence is its marked similarity to the Milan covers. While some of the scenes are different and there are certain slight variations in the iconography of the scenes which are common to both ivories, the likeness of the two works is so close that several authorities have considered them to have been carved by the same hand and all agree that they came from the same art centre. Inasmuch as the greater part of the scenes on the casket are the same in subject and rendering as those on the covers, the casket necessarily shows an eclectic iconography, a tendency to follow the account of the scenes in the Gospel of Pseudo-Matthew, and is to be attributed along with the Milan covers to a Provengal atelier. Such new scenes as appear on the casket are so uncommon in early Christian art of the sixth century 1 Garr., Storia, VI, pi. 447/1, 2, 3. A SCHOOL OF IVORY CARVERS IN PROVENCE 223 that they may well be neglected as far as the solution of the origin of the box is concerned. The scene of the Virgin in the Temple is, with the exception of the priest who reads by the steps of the temple, similar to the same scene on the Milan cover. While it has been interpreted by Westwood 2 to represent Zacharias, book in hand, standing by the temple, such an interpretation, as well as that of Stuhlfauth , 3 who calls the scene ‘‘the annunciation of the birth of Mary to Zacharias,” is impossible, for no recognition is made of the figure of the young maiden toward whom the angel looks as he points heavenward. That this maiden is meant to represent the Virgin is partly shown by the similarity of her costume and hair with the figure of the Virgin in the scene of the Annunciation. The interpretation of De Waal 4 and Leclercq , 5 that it depicts the Presentation of the Virgin in the Temple, is, so far as the original meaning of the scene goes, the same as my own, which has already been set forth . 6 The assertion of Stuhlfauth that the explanation offered by De Waal is impossible because the Presentation of the Virgin was unknown in early Christian art does not take into account the fifth century slab found at St. Maximin in Provence, where the Virgin in the Temple is represented with an explana- tory inscription. While the priest, as I have said, is omitted in the scene on the Milan covers (see Eig. 1), the two renderings are otherwise iden- tical and. in spite of this one difference, are clearly related. With this in mind, it is interesting to note that the source of the scene is the Gospel of Pseudo-Matthew as in the case of so many of the scenes on the Milan covers, a fact which is also pointed out by De Waal. Following the Virgin in the Temple should come the Annunciation by the Spring; but the demands of composition were such that apparently the artist chose to use this scene as a suitable terminus for his panel. Other- wise. the scenes as they are depicted on the casket adhere to the order of events as related in the Gospels. Of the Annunciation nothing more can be said than has already been advanced regarding the same scene on the 2 \\ estwood, Fictile Ivories, p. 42. 3 Stuhlfauth, Die altcliristliche Elfenbeinplastik, p. 72. 4 De Waal, Rom. Quar., I, p. 185. Leclercq in Cabrol’s Dictionnaire d’archeologie chritienne et de liturgie, s. v. “Apocryphes,” col. 2557. 6 See p. 7. 224 A SCHOOL OF IVORY CARVERS IN PROVENCE Milan cover (see Fig. 4), for both are identical even to the embroidered dress of the Virgin. 7 The Visitation, where Mary and Elizabeth meet before a building with two towers, which represents the “city of Juda” wherein Elizabeth dwelt, 8 does not occur on the Milan covers. Elizabeth, standing with her back to the towers as she extends her arms to the approaching Virgin, wears a heavy veil after the fashion of a matron, and a mantle whose long sleeves hang down at her side. The Virgin, on the other hand, as befits her age and position at the time, is clothed in more simple and youthful garments and her hair is dressed in a more stylish manner. As a means of locating or dating an early Christian monument, the Visitation is wellnigh useless. Outside the fact that the scene does not occur before the sixth century and only appears on monuments which were either of Eastern origin or may be connected with the East, it offers no distinctive iconography. While there have always been two methods of representing the scene, one, as on the ivory, where the two women meet without em- bracing, and the other where they clasp each other in their arms, neither type seems to have been peculiar to any specific region or century. Since the greater part of the examples occur on monuments from Syria or Egypt, we may only assume, if we accept the Provenqal origin of the Werden casket, that the scene, like so many others we have seen in Provence, is to be accounted for by the large Oriental population in Pro- vence and the intimate relation of the region with the East. The Visitation is followed by a s,cene which is even less instructive and helpful and warrants only the briefest description, as it is unique in the Christian art of the first eight centuries, save for the curious fresco of the Departure of the Virgin from Elizabeth’s House, at Bawit. 9 The two standing figures, it is to be noted, are the same in every detail of their costumes as the two women in the Visitation, the figure at the right corresponding in respect to her heavy veil and matronly garb to the mother of John, while the other woman at the left is similar, in her more maidenly attire, to the Virgin. The composition of the group is, moreover, similar to that of the Visitation, save for the fact that the two women instead of ad- 7 See Figs. 4 and 5. 8 This is the interpretation given to the scene by Westwood (op. cit., p. 42) and by Stuhlfauth (op. cit., p. 72). 9 Cledat, C. R. Acad. Inscr., 1904, p. 5 2 5 - A SCHOOL OF IVORY CARVERS IN PROVENCE vancing to greet one another appear to have turned their backs as if the event commemorated were a farewell. This representation, then, can only be explained as the departure of Mary, after her three months visit, from the home of Elizabeth. If this is the correct interpretation, the source of the scene, as of the Visitation, would probably be the account in the Gospel of Luke (i, 56). After the departure of the Virgin there is depicted on this panel the Appearance of an Angel to Joseph. While this does not occur on the Milan covers, there are reasons for connecting it with Provence. In the first place it follows the account in both the Latin Pseudo-Matthew and the Gospel of Matthew, where it is related that Joseph, being of a mind to put Mary aside when he found her upon her return from the home of Elizabeth heavy with child, was visited in his sleep by an angel who assured him that he need have no fear. Furthermore, the only example of the scene which is similar to that on the ivory occurs on a sarcophagus of Le Puy 10 (Fig. 154). Besides being very rare in early Christian art, its narrative character and lack of symbolism would suggest its origin in a region where the Oriental fondness for historical representations of Biblical events was more active than in Rome. Moreover the other ex- amples of the scene which appear on Egyptian monuments present a ren- dering quite different from that on either the ivory or the sarcophagus. On the Maximianus chair (Fig. 150) and in a fresco of Antinoe (Fig. 1 5 1 ) , Joseph sleeps in Eastern fashion on a mattress. On the ivory, however, Joseph lies on the ground, while on the sarcophagus he is curled up on a little couch. His costume on both the ivory and the sarcophagus is the same, while it is different from his dress on the two Egyptian examples. The type would then seem to be a local and Gallic one. The next scene occupied the short panel of what was the left side of the casket. This panel, however, is gone and, although I think from the sequence of the scenes on the casket and their general similarity to those on the Milan covers that it contained the Massacre of the Innocents, it is useless to speculate. Passing, then, to the long panel (Fig. 158) which was at the back of the box, the next scene is the Appearance of the Star. The Appearance of the Star to the Three Magi is, in the first place, 10 Le Blant, Les sarcophagcs chretiens de la Gaulc, pi. XVII/4. 226 A SCHOOL OF IVORY CARVERS IN PROVENCE identical with the scene on the Milan covers (see Fig. 27), even to the attitude and odd chitons of the Magi; second, like the scene on the covers, it is similar to the rendering of the Appearance of the Star on the Provencal sarcophagi and is, therefore, to be connected with that section of southern Gaul where the sarcophagi and the covers were executed. 11 I he Adoration of the Ox and the Ass, which is next in sequence, as in the case of the rendering of the same scene on the Milan covers (see Fig 12), which it seems to follow, may be connected with the account of the event in the Latin version of Pseudo-Matthew, and related to the Fig. 158. S. Kensington: Victoria and Albert Museum, ivory casket from VVerden. Back. representations on the sarcophagi of Provence. 1 - Its close similarity to the Milan scenes appears in practically every detail of the composition, but especially in the presence of Joseph carrying a curiously shaped saw and in the marked form of the manger crib covered with straw or a soft material whose folds radiate from under the Child. Whatever is true for the Milan scene holds good, then, for the representation on the casket. The relationship between the Adoration of the Magi on the casket and the same subject on the Milan covers (see Fig. 35) is apparent. As on the Milan ivory, the Virgin, heavily veiled and seated on a cathedra with a round back like those on the Roman sarcophagi, rests her feet on a suppedaneum and holds in her lap the young Jesus who is clothed in a long tunic as a child of two years; from the left the Magi, dressed in the distinctive notched chitons and flowing chlamydes, advance in a line toward the Child, bearing their gifts. The fact that the gifts of the Magi differ in the two representations is of no importance, for no artistic tradi- tion as to the kinds of gifts which the Wise Men brought from the East was constantly maintained. 13 The type, where the Magi are all beard- less 14 and advance in a line, is the Hellenistic type peculiar chiefly to the 11 See p. 216. 12 See p. 215. 13 See p. 40. 14 It has been said that the first Magus is bearded. Since the ivory has been badly A SCHOOL OF IVORY CARVERS IN PROVENCE sarcophagi representations in the West. 10 The rendering on the casket appears to have been based rather on the Adoration of the Milan hook covers than on any text, and the characteristic which connects this repre- sentation on the Werden ivory with Provence is its obvious and symbolic combination of the Adoration of the Magi 16 with the Adoration of the Ox and the Ass. This symbolic combination, while occurring on the ivories, appears most convincingly on the sarcophagi of Provence. 1 ' Finally, on the last panel (Fig. 159) there is represented a combination scene wherein one part illustrates the words of John the Baptist when he was preaching to the people of Judea on the banks of the Jordan prior to the coming of the Saviour, and the complementary portion depicts the subsequent Baptism of Jesus. The scene at the left of the panel shows three men fleeing from the gates of a city and approaching one who buries his ax in the coils of a viper wrapped about the roots of a tree, while a second man. on the further side of the tree, raises his ax in preparation for cutting down the unfruitful tree. The representation is a direct pic- torial rendering of the words of John, recounted in the third chapter of Matthew and Luke : “Oh generation of vipers, who has warned you to flee from the wrath to come? . . . and now also the ax is laid unto the root of the trees : therefore every tree which bringeth not forth good fruit is hewn down and cast into the fire.” While unique, to my knowledge, among the scenes of Christian art before the ninth century and, therefore, valueless for comparative iconography, this composition shows that the casket was executed in an art centre which was not subservient to estab- lished traditions and only makes more manifest the eclectic and varied nature of this school of ivory carvers in Provence. The Baptism, which completes the scenes on the Werden casket, offers a composition which, regardless of two added features, resembles so closely the scene on the Milan covers as well as on the other ivories of the Pro- vencal group, that I have called the type Provencal. 18 Although the Werden scene is obviously to be connected with the tvpe worn the original does not give, as a photograph does, the impression that the Magus wears a beard. Because the three Magi in the scene of the Appearance of the Star on the same panel are beardless and because the peculiar length of chin of all the faces on the casket gives the impression of beards, I believe that what looks to he a heard in the photograph is no more than a few irregularities caused by wear. la See p. 38. See p. 22. 17 See Fig. 14. See Baptism, p. 76, and Milan Book Covers, p. 214 228 A SCHOOL OF IVORY CARVERS IN PROVENCE on the other ivories of the group, it has two added features in its composi- tion which must be explained. The cruciform nimbus on the head Fig. 159. S. Kensington : Victoria and Albert Museum, ivory casket from Werden. Side. of Christ must have been inscribed on the ivory at a later date, for it is not given to Christ in this scene on the Western monuments before the eighth century and does not appear in Oriental Baptisms until after the sixth century. 19 The other odd feature is the personification of the Jordan, a jovial reminiscence of a pagan deity who sits upon a mass of rocks with his arms full of water plants and has at his feet an over- turned urn from which the waters of the river flow down the side of the hill. The use of the personification of the river in this scene was, in all probability, an original characteristic of the Alexandrian-Coptic representa- tions, 20 although as early as the fifth century the Jordan was personified in both the mosaics of S. Giovanni in Fonte and S. Maria in Cosmedin at at Ravenna. 21 The custom of personifying rivers and mountains was in itself a tradition introduced from the Hellenistic centres of the Orient into the West at an early date. While its use in the Christian scenes of the West was rare, there occurs on a sarcophagus of Syracuse, in the scene of Moses Striking Water from the Rock, a personification of the mountain. 22 As a Hellenistic feature of uncertain origin its appear- ance in Provence is no more surprising than its use in any other art centre. From Gallo-Roman times Provence was in intimate relationship with Alex- andria and the other Hellenistic cities of the East and the Provenqal fea- ture in the scene of the Baptism, of representing the waters of the Jordan piled back upon their source until they appear like a cataract, was derived from an Eastern account of the Baptism. 23 So far a review of the iconography on the Werden casket has not 19 See Table IV. 20 See p. 79. 21 See p. 82. ' 22 See p. 12, note 7. 23 See p. 81. A SCHOOL OF IVORY CARVERS IN PROVENCE 229 revealed any new links which attach the ivory or the group of ivories to Provence. Nevertheless, the comparison of such scenes as the Virgin in the Temple, the Annunciation, the Appearance of the Star, the Adoration of the Animals, the Adoration of the Magi, and the Baptism, with the same scenes on the Milan covers has shown beyond question that the origin of the Werden casket is the same as that of the Milan book covers. While all the authorities agree as to the close relationship between the two ivories and Westwood 24 and Strzygowski 25 see in the casket the work of the same hand that carved the Milan covers, Schmid 26 and Stuhlfauth 27 notice cer- tain variations in the execution which make it necessary to suppose that a second artist did the casket. They assume, however, that both artists either worked from the same model or that the Werden casket was a weaker copy of the covers. Therefore, if the attribution of the Milan covers to Provence has been sustained in the previous chapter, the same origin must be ascribed to the Werden casket. None of the new scenes occurring on the casket and not on the covers has offered any iconography inconsistent with the traditions of Provence or incompatible with the eclectic art of that country. Moreover, the scene of Joseph’s Dream finds its closest parallel on a sarcophagus of Le Puy and gives some evidence, at least, of the Gallic origin of the casket. A small but important detail on the ivory is the structure with a facade flanked by two round towers in the scene of the Visitation (Fig. 157. Strzygowski, in discussing the origin of Romanesque art in the West, ascribes the Werden casket to Milan and claims the two-towered fa9ade as an Oriental characteristic. 28 In evidence of this he cites the prevalence of this architectural form in the churches of Syria, its use on the Brescia casket which he ascribes to Anatolia, and its occurrence on the doors of Santa Sabina which were also of Eastern fabrication. At once we may ask. on the same grounds, why this feature could not have been introduced into Marseilles by Cassianus or one of his Eastern monks? Marseilles, as well as the church and monastery of St. Victor, was as closely connected with the Orient and with Syria as was Milan and any of its churches. 24 Westwood, op. cit., p. 42. 2o Strzygowski, Iconographie der Taufe Christi, p. 13. 24 Schmid, Die Darstellung der Geburt Christi, p. 109. 27 Stuhlfauth, op. cit., p. 73. 25 Strzygowski, Kleiitasien, pp. 214-215. 2 3 o A SCHOOL OF IVORY CARVERS IN PROVENCE The question answers itself ; for the use of the two-towered facade to represent a city was known and used in Gaul. On a Merovingian gem (Fig. 160) is represented a facade flanked by two towers not unlike those on the ivory. 29 On either side of the building stand the martyrs Gervasius (GERBA) and Protasius (PROTA), as the inscription tells us, and be- low is the inscription CAENOM(ANI), the tribe whose capital became the city of Le Mans. Thus the building on the gem does not neces- Fig. 160. Engraved gem : Sts. Gervasius and Protasius before the city of Le Mans. sarily represent a church, as one might imagine, but symbolizes the city of Le Mans where the remains of Gervasius and Protasius were preserved. This same use of a towered facade to symbolize a city occurs on the cover of the Lorsch Gospels, of the eighth or ninth century, in the Vatican; Herod is represented seated before a facade with flanking round towers as he re- ceives the three Wise Men who are on their way in search of the new born Babe. 30 It is to be noted that the towers on all these Western monuments are round, while the towers on the Syrian churches are usually square, as also on the Brescia casket and the doors of Santa Sabina. It has been pointed out by Enlart that the circular tower was commonly used in Caro- lingian times in preference to the square tower and that it seems to have been more ancient in Gaul than the square form. 31 Since the facade on the Werden casket must be interpreted to represent the city of Juda where Elizabeth lived, there is more reason for associating the circular towers with Gaul than with either Milan or the East. While the relation of the Werden casket to the Milan covers is obvious, not only in iconography but also in the ornament and costumes of the fig- 20 Garr., VI, pi. 478/40. 30 Garr., VI, pi. 457/2. 31 Enlart, Manuel d’archeologie frangaise, I, pp. 124-125. A SCHOOL OF IVORY CARVERS IN PROVENCE 23 1 ures, I would say neither that they were by the same hand nor that the Werden casket was a weaker copy. It is true that there do appear certain differences in the actual carving of the two ivories; the figures on the Milan covers are far more fiat and sketchy in their rendering and they entirely lack the snap and form of those figures on the casket where there seems to have been an effort to obtain plastic effect and action. Nor do I feel that the casket can be called a “weaker copy.” While a less pretentious work than the covers, it seems to me to be more virile and to show in in- cipient form those characteristics of style, such as the exaggerated anima- tion and movement, which were later apparent in Carolingian art. Without doubt the casket was executed at a date posterior to the carving of the covers. The diminished Hellenism of the style would show it, as also would the fact that on the casket the story of the Virgin is more fully developed than on the Milan ivories. Although somewhat later, the casket, nevertheless, could not have been done much after the beginning of the sixth century. The style is not Carolingian ,and it is impossible to ascribe a work of this character to the seventh or eighth century. There- fore, accepting the date of the execution of the Milan book covers as about the year 500, I consider the Werden casket to have been carved at a time well within the first half of the sixth century. 32 V DIPTYCH IN THE LIBRARY OF ROUEN Diptych. Rouen Date, c. 475 A. D. Library. 1 he diptych of Rouen, 1 which is today in the town library, has been in the cathedral of that city at least since the twelfth century when it was used to adorn the binding of a manuscript containing a miscellany of docu- ments relating to the episcopal see of Rouen. 2 Each leaf (Fig. 161) is or- namented with a pediment sustained by channelled pilasters with Corinthian capitals: in each tympanum is figured a shell and on the raking cornices ” 2 This date is the same as that given by Stuhlfauth, Schmid, and De Rossi. 1 C. de Linas, Gazette archcologigite, 1886, pi. IV; Laurent, Lcs ii'oires prcgotliiqucs, figs- 7, 8. 2 De Linas, op. cit., p. 25. 232 A SCHOOL OF IVORY CARVERS IN PROVENCE Fig. 161. Rouen: Public Library, ivory diptych. Saints Peter and Paul. A SCHOOL OF IVORY CARVERS IN PROVENCE 233 stands a bird. Within the niches are two apostles : St. Peter, who is nimbed and holds in his hand a key, stands in a three-quarters position; St. Paul, who is also nimbed and advances toward the right holding in his left hand a roll, while with his right hand he makes a gesture of either speech or blessing, is designated in the customary fashion by his bald head and long beard. The striking similarity, which has been frequently remarked by art critics, of the two apostles on the ivories to the rendering of the same Saints on a sarcophagus of St. Victor at Marseilles (Fig. 162) together with the resemblance of the figure of Paul to the same figure on another sarcophagus from St. Victor, 3 indicate that the Rouen diptych must have been executed in Saint Cassianus’ monastery of St. Victor at Marseilles. The connection with Marseilles is sustained by the style and architectural details of the ivories which are similar to the pseudo-Eastern sarcophagus type of Provence. The face of the first named sarcophagus from Mar- seilles is broken into seven rectangular panels, of which the last at the left is now gone. The central panel is framed by Corinthian capped pilasters. The figures represent Christ and six of His apostles arranged in groups of three on either side. In the space to the left of the Saviour is the figure of St. Paul executed in a manner almost identical with the same personage on the diptych, and in the panel to the right is St. Peter. That some relation must have existed between the sarcophagus and the diptych has already been recognized by Molinier, 4 Laurent, 5 and Dalton. 6 Molinier, who praises highly the fine workmanship of the ivories which shows that the artist had under his eye a good model, suggests that who- ever, at the end of the Roman period, executed the ivories was inspired by the same model as the artists who carved the sarcophagus. Instead of agreeing with De Linas, who first published the ivories and held them to be the work of Greek image makers, he prefers to consider them the product of Occidental artists, either Italian or Gallo-Roman. Stuhlfauth, 7 on the other hand, who ignores the parallel with the sarcophagi, claims, on the basis of the architectural niche and a vague resemblance of the 3 Le Blant, Les sarcophages chretiens de la Gaule, pi. XVI/i. 4 Molinier, Hist. gen. des arts appliques, I, Ivoires, p. 53. 5 Laurent, op. cit., pp. 23-25. 6 Dalton, Byzantine Art and Archaeology, p. 194. 234 A SCHOOL OF IVORY CARVERS IN PROVENCE ivories to the Archangel panel 7 8 in the British Museum, that the diptych was a product of Byzantine art and done by a Greek artist. While accepting a direct relation between the figures of the sarcophagus and the ivories, Laurent says that the type of the figures is not at all the same as the ordinary type of the Western sarcophagi and that both monu- ments show the influence of Oriental art upon the West. In explaining the Eastern type of figures, as he calls them, on the sarcophagus, he pro- pounds three theories. First, the sarcophagus may have been done in the East. Then, on the contrary, it may have been executed in Gaul, the sculptor copying the ivories. Or better, as Molinier thinks, both the sar- cophagus and the ivories were inspired by the same model. Fig. 162. Marseilles : Museum, sarcophagus. Christ and Apostles. This last conclusion of Molinier's seems to me to be correct, save that he has not sufficiently emphasized the connection of both ivories and sar- cophagi with Marseilles. In the first place, the sarcophagus of St. Victor, on which Peter and Paul are represented, is not in the least similar, as Laurent notes, either to the characteristic Roman sarcophagi or to those of the atelier of Arles which followed Roman models. There was, however, a second school of sarcophagus carvers in southern Gaul which did not par- ticipate in the styles of Arles and Rome, but attached itself, by contact more or less direct, to the Christian art of the Orient. This Eastern art stream, coming largely from Syria and transmitting to the Provencal sarcophagus carvers many Syrian motives, spread into Gaul from Marseilles . 9 While the vine motive and the conch shell borne by pilasters are not features 7 Stuhlfauth, Elfenbcinplastik, p. 176. 8 Garr., op. cif., VI, pi. 457/1. 9 See Orientalizing of Gaul. A SCHOOL OF IVORY CARVERS IN PROVENCE 2 35 found upon Roman sarcophagi, they are characteristic of the sarcophagi of southern Gaul . 10 The setting of the figures in panels or shallow niches and the general low relief of the carving on the St. Victor sarcophagus are again features peculiar to this Provencal type . 11 Accepting, then, a relationship between the sarcophagus and the dip- tych, it is impossible, when we consider the similarity of the architectural details as well as of the figures, to suppose that the Rouen diptych was carved by an Italian artist. On the other hand, it is equally unlikely that the diptych was executed in the East, or was even copied, at least directly, from an Oriental model. It is true that niches with birds on the pediment, as they occur on the ivories, are Eastern features, but they were also com- mon in Provence. Furthermore, the niche on the diptych with only a small shell ornamenting the centre of the tympanum is quite different from the customary Eastern niche wherein a large conch shell, supported by columns with spiral channellings, takes the place of a pediment. The types of the two apostles, Laurent and Stuhlfauth to the contrary notwithstanding, are not primarily Eastern. Comparing the two Saints with the same per- sonages on the Maximianus chair , 12 and the Tongres-Brussels diptych , 13 there appear certain marked differences. On the Eastern ivories St. Paul wears a pointed instead of a round beard and his head is not so bald as it is on the Rouen diptych. The apostles on all the Oriental ivories are figured in a rather rigid frontality while the profile and three-quarters pose of the two Saints on our diptych recall a much finer Hellenistic tradition. The costume, also, of these apostles does not agree in the least with that of the figures on the Eastern ivory carvings. Thus, although there are certain Eastern reminiscences on the Rouen diptych, there is nothing, to my mind, which justifies either attributing it to the East or assuming that it was copied from an Eastern model. More- over. if the figure of Paul on the diptych be compared with the same per- sonage on a sarcophagus of St. Victor 14 and a sarcophagus of Aix 15 it will be seen that the heads, postures, and costumes are practically iden- 10 See pp. 198-199. 11 Le Blant, op. cit., Marseilles (XII/4, XYI/2), Rodez (XXII/i), Toulouse (XXXVII, XXXIX. XL, XLI, XLII), Le Mas-Saint-Antonin (XLVTII/i), and Apt (XLIX). 12 Garr., op. cit., VI, pi. 416. 13 One leaf is in the Treasury ot Xotre-Dame de Tongres, the other in the Musee du Cinquantenaire at Brussels; Molinier, op. cit., pp. 54, 55; Laurent, op. cit., figs. 1, 2. 14 Le Blant, op. cit., pi. XVI/i. 13 Le Blant, op. cit., pi. LI/2. 236 A SCHOOL OF IVORY CARVERS IN PROVENCE tical on all three examples. The analogies, then, are all with Provence and, what is more, with monuments which are to be connected with the church and monastery of St. Victor at Marseilles. By thus attributing the Rouen diptych to Marseilles the objections to its Provencal origin are to a large extent removed. Provence, we have seen, was overrun with Greeks and Syrians, and the church of St. Victor, which Cassianus, after prolonged sojourn in the East, founded about 414, was more intimately connected with the Orient than with the West. 10 Therefore the conviction of De Linas and Stuhlfauth that the diptych came from the hands of a Greek artist and the contention of Laurent that it was copied from an Eastern model, may be all admitted without weaken- ing the attribution to Marseilles or discarding Molinier’s opinion that it was carved by a Gallo-Roman. The date of the Rouen diptych should be somewhere about the begin- ning of the last quarter of the fifth century. As compared with the Milan book covers, the Rouen diptych presents a style quite similar in general lowness of relief, in the channelled pilasters with Corinthian capitals, and in the egg and dart mouldings, though it seems superior in design and workmanship. The freedom of draperies, the careful modelling, the de- tails of the costumes such as the tassels which ornament the angles of the pallium, and the exactitude of the movements, point to a date earlier than the Milan covers before art had begun to abandon the finer technique and classical freedom of the Hellenistic examples for the rather slack sketchiness of the Oriental style. Besides the style we have the approximate date of the sarcophagus in St. Victor for pur- poses of comparison. Le Blant in publishing the sarcophagus intimates that it could not have been earlier than the end of the fourth century 17 and De Rossi states that the particular form of the cruciform monogram carved on it originated in Gaul in the first half of the fifth century, or at the earliest at the end of the fourth century. 18 Therefore since the sarcophagus was found in the crypt of St. Victor and as there is every reason to suppose that it had been executed in Marseilles for that church, it would be quite consistent to place its execution after 414, when the church and monastery were founded by Cassianus. Inas- 10 See p. 196. 17 Le Blant, op. cit., p. 47. 18 De Rossi, Bull, di arch, christ., 1880, p. 155. A SCHOOL OF IVORY CARVERS IN PROVENCE 237 much as it is not likely that a school of art would have grown up around a new order until some time after it had been thoroughly estab- lished, and as the style of the monogram was more common at the end of the fifth century than at its beginning, I am inclined to believe that the sarcophagus was carved between 450 and 475 A. D. and that the Rouen diptych could not have been done much later. Moreover, the style and execution of the diptych, as in the case of the Milan diptych, are less plastic, bold, and hence less Hellenistic than the work on the Berlin and Nevers fragments, which in the next chapter I shall show to have been executed about the second quarter of the fifth century. TWO IVORY FRAGMENTS IN THE BERLIN AND NEVERS Among the ivories to be considered in the Provencal group are two fragments, of which one 1 is the side panel of a five part diptych (Fig. 163) and is preserved in the Berlin Museum ; the other, possibly belonging to the same diptych, is in the Museum of Nevers (Fig. 164). The ivory fragment which is now in the Berlin Museum was previously in the pos- session of M. Mallet at Amiens. The three scenes on this panel are sep- aratelv framed by an egg and dart moulding, and on the right hand edge of the ivory is a narrow bead and reel moulding which originally sur- rounded the whole diptych. Reading down, the scenes are the smashing type of the Massacre of the Innocents, the Baptism, and the Miracle of Cana. That the Nevers fragment 2 was possibly part of the same diptych as the Berlin panel is suggested not only by their similarity in style and in 1 Haseloff, Jb. Preuss. Kunsts., XXIV, 1903, p. 47 sq; Abbe Drival, Revue de I’art chretien 2e Serie (1875), pi. XIX; Stuhlfauth, Elfenbeinplastik, p. 75, pi. IV/i ; Strzy- gowski, Kleinasien, p. 199; Fleury, La Messe, VI, p. 1 19 ; Dalton, Byz. Art and Arch., p. 203. 2 Barbier de Montault, “L’ivoire latin du musee de Nevers,” Bulletin Monumental, 1885, no. 8; A. Darcel, Tresor des eglises et objets d’art franqais appartenant aux viusees exposes en 1889 au palais du Trocadero, vol. I, no. 14; Haseloff, op. cit., p. 52, gives the dimensions of both fragments and points out the relation; Molinier, Ivoires, p. 60, note 6; Stuhlfauth, op. cit., p. 77, wrongly attributes the ivory to Toulouse; Dalton, op. cit., p. 203. VI MUSEUMS Two Diptych Fragments. Date, c. 450 A. D. Berlin and Nevers Museums. 238 A SCHOOL OF IVORY CARVERS IN PROVENCE the mouldings which the}' both employ, but also by the measurements of the two ivories. The dimensions of the Nevers fragment, which are 7.4 by 12.9 cm., indicate that it might fit on to the top of the Berlin fragment, whose dimensions are 20 by 8.1 cm. Thus the Adoration of the Magi on the Nevers fragment would be set above the Massacre of the Innocents on the Berlin panel, while the Nativity, with the missing Joseph and Mary (p. 18), and a third scene, would complete the crowning frieze of one leaf of a five part diptych. The intimate relation of the Nevers and Berlin fragments with the Milan book covers and the Werden casket in the South Kensington Museum has been readily accepted by all the writers on these ivories. 3 In fact, the iconographical connection between them is so apparent that they have all been considered products of the same school. While attributing both fragments to the same atelier as the Milan covers, Stuhlfauth failed to recognize the relation between them and dated the Nevers panel as the earliest example of what he called a Milan school, while he set the execu- tion of the Berlin fragment between the Milan covers and the Werden casket. 4 His principal reason for dating the Berlin ivory after the Milan covers and before the Werden casket was the fact that the Saviour wears a plain nimbus on the Berlin panel and a cruciform nimbus on the Werden casket. I11 marked contrast with Stuhlfauth's views is the theory of Haseloff, who attributes the whole group of ivories to Rome and dates the fragments in the closing years of the fourth century. Inasmuch as a modification of Haseloff’s theory seems to me to be the truth, I will first give his reasons for placing the execution of the Berlin fragment at Rome and for dating it so long before the time when the Milan covers are supposed to have been done. It is at once apparent, when compared with the Milan covers, that the Berlin and Nevers ivories are superior in workmanship and must have been done at a time when better traditions still existed among the ivory craftsmen. To find a working basis from which to approach the date and place of execution of the Berlin fragment, Haseloff traces the development of ivory carving in Rome. Beginning with the Symmachi diptych of either 3 See especially the previous references to Haseloff, Stuhlfauth, and Strzygowski. 4 Stuhlfauth, op. cit., pp. 76-8. 5 Haseloff, op. cit., p. 55; O. Seeck, De Symmachi vita. Monumenta Germaniae Hist.: auctores antiquissimi, VI (1883), p. lix; Molinier, op. cit., p. 43, no. 58. Fig. 163. Berlin Museum, ivory. Massacre of Innocents, Baptism, Miracle at Cana. 24 O A SCHOOL OF IVORY CARVERS IN PROVENCE 392-394 or 401, 5 he follows the continuation of certain ornamental motifs and points of style, through the Probianus diptych of the latter part of the fourth or the very first of the fifth century, 6 the “Lampadiorum” diptych at Brescia of the middle of the fifth century, 7 to the Trivulzio group which according to him consists of the leaf with the scene of the Holy Women at the Tomb, in the collection of Prince Trivulzio at Milan, 8 the four plaques, with scenes of the Passion, in the British Museum, 9 and the three panels, with scenes from the lives of Moses, Paul, and Peter, in the same museum. 10 When Haseloff contrasts the two fragments with the Trivulzio group, he finds that they are cruder and smoother in execution and differ in the construction of the figures and in the handling of the draperies. But when the same fragments are compared with the pagan group of consular diptychs, they appear to be remarkably similar to the Probianus diptych. They are similar to this diptych in the broad, flat background behind the figures, in the shape of the heads, and in the treatment of the hair, hands and costumes. Therefore, from this marked likeness to the Probianus diptych and from the fact that after the beginning of the fifth century all the Roman diptychs become very crude, flat, and inferior, 11 he attri- butes the Berlin and Nevers fragments to a period in Rome just before the date of the Probianus diptych, which would have been the end of the fourth century. Thus, according to his theory, these ivories were the earliest examples of a Roman school of Christian ivory carvers who later produced the Milan covers and the Werden casket. This connection of the group with Rome would be further borne out by the fact that the Milan covers have the same outer moulding of palmette ornament which occurs on the Symmachi diptych, the Probianus diptych, and the Trivulzio ivory. While keeping in mind the analogies which Haseloff has shown to exist between the two fragments and the Roman ivories of the first part 6 Haseloff, op. cit.; Molinier, op. cit., pi. IV. 7 Meyer, Zwei antike Elfenbeintafeln der Koniglichen Staatsbibliothek in Munchen. Abh. d. K. Bayr. Akad. d. IViss., I. Kl. xv. Bd. I Abt. Munich, 1879, Taf. II ; Molinier, op. cit., p. 32, no. 33. 8 Molinier, op. cit., pi. VI. 9 Dalton, Catalogue of the Early Christian Antiquities in the British Museum, pi. VI. See p. 187, note 1. 10 Dalton, op. cit., pi. VII, no. 292. 11 Haseloff, op. cit., pp. 59-60, cites as examples of this decline in Roman ivory carving the Basilius diptych of the year 480, the diptych of Boethius of 487, and the Orestes diptych of 513. A SCHOOL OF IVORY CARVERS IN PROVENCE 241 of the fifth century, I wish to consider their iconography before continuing the discussion of style. The Adoration of the Ox and the Ass on the Nevers panel , 12 of which only half the scene is preserved, is the same Provenqal type, following the account in Pseudo-Matthew, which occurs on the Milan book covers and the Werden casket . 13 Even in the half- scene there are figured the stable, the adoring animals, and the distinctive stone crib which we have seen were peculiar to the group. The Adoration of the Magi on the same panel is again similar in iconography to the Adoration on the Milan covers . 14 In both examples the Virgin is seated in a high backed, round headed cathedra, the Magi are dressed in the curious tri-notched chiton which I have shown to be distinctive of the monuments of Provence , 15 the Magi offer practically the same gifts consisting of two platters of uncertain objects and a cornucopia, and on both monuments the scene is set in a house . 16 Furthermore, the combination of the Adora- tion of the Ox and the Ass with the Adoration of the Magi I have already pointed out as peculiar to the Provencal monuments . 17 Turning to the iconography of the Berlin fragment, the scene of the Massacre of the Innocents besides presenting the best proof of the Pro- vengal origin of the ivory also offers a terminus post quern for its date of execution. The “smashing'’ Massacre of the Innocents, in which the soldiers of Herod smash the children to the ground instead of cutting them down with swords, I have already shown was peculiar to Provence in early Christian times and to Gaul during the Carolingian period . 18 Since the remains of the first martyrs were brought by Cassianus from Palestine to Provence during the first quarter of the fifth century, and since it was not until after the installation of their remains in Cassianus’ church of St. \ ictor at Marseilles that their cult became established in the region, it is unlikely that the first example of the “smashing” Massacre on a sar- cophagus cover in the church of St. Maximin, which was a dependency of St. A ictor, dates before the end of the first quarter of the fifth century . 19 At Rome, moreover, no special recognition was made of the Innocents until the sixth century, and the only representation there of the Massacre, on the arch mosaics of Santa Maria Maggiore, showed a symbolic rendering of the sword type. Therefore, inasmuch as I have elsewhere pointed out 20 12 See The A ativily, p. 17. 13 See The Nativity, p. 20. 14 See Adoration of the Magi, p. 44. 33 See pp. 34, 43. 14 See Adoration of the Magi, p. 44. 17 See p. 21. 18 See Massacre of the Innocents, p. 65. in See p. 65 . 20 See p. 66. 242 A SCHOOL OF IVORY CARVERS IN PROVENCE wherein the iconography of the Massacre on the sarcophagus of St. Maxi- min is more primitive and earlier than either the Berlin or Milan scene, we have a terminus post quern of the first quarter of the fifth century for the execution of the Berlin and Nevers fragments. The Baptism on the Berlin panel, although it offers little evidence re- garding either the date or provenience of the original diptych, indisputably connects the fragment with the Milan and Werden ivories . 21 The scene, with the exception of the nimbus on the head of Jesus, is practically identical with the same scene on the Milan covers. The presence of a nimbus on the head of Christ, which led Stuhlfauth to date the ivory after the Milan covers, has little or no bearing on the date, for it seems certain to me that it must have been added sometime after the ivory had been carved. In confirmation of this assumption it will appear, on careful observation, that the waves of the Jordan behind Jesus run straight across the nimbus, as if the nimbus were a circle inscribed on the ivory regardless of the background. The Miracle of Cana represents the transitional or Alexandrian-Coptic type introduced from the East into the West sometime during the fifth century . 22 At the end of the fourth century, when Haseloff supposes the Berlin ivory to have been carved, the Hellenistic type, wherein Jesus touches with His wand the jars of water and in which no servant is in- cluded. was faithfully followed on the sarcophagi of Rome. Moreover, this scene on the Berlin fragment, while of the same type as the scene on the Milan covers, depicts an earlier and simpler rendering of it . 23 Here again the nimbus is of little importance, for if the nimbus in the Baptism was added later it is presumable that this one on the head of Christ was also inscribed at a later date. Viewed as a whole, then, the iconography shows that the Berlin and Nevers fragments, close enough in style to be parts of the same diptych, are so intimately related to the Milan book covers that they must have been exe- cuted in the same art centre, if not in the same atelier, as were the covers. Therefore, if the connection of the Milan covers with Provence has been established, the Provencal origin of the Berlin and Nevers panels is at once proved. Independently of this relation with the ivories already connected 21 See The Baptism, p. 76. 23 See p. 91. 22 See The Miracle of Cana, p. 90. A SCHOOL OF IVORY CARVERS IN PROVENCE 243 with Provence, the combination scene of the two Adorations on the Nevers fragment and the “smashing” Massacre on the Berlin ivory show Pro- vencal types which themselves fix the place of origin. As to the question of date, the iconography at once makes it clear that the two fragments are earlier than the Milan covers and at the same time are posterior to the first quarter of the fifth century. This date is suggested by the improbability that the Massacre of the Innocents would have been carved on an ivory of Provence before the cult of the Innocents was intro- Fig. 164. Nevers : Museum, ivory. The Nativity and Adoration of the Magi. duced into the country during the first part of the fifth century . 24 Thus the iconography establishes the Provencal origin of the fragments and sets their execution somewhere between the first quarter of the fifth century and the beginning of the sixth century when the Milan covers were prob- ably carved. While it seems to me that the iconography offers the most conclusive evidence as to the date and provenience of the ivories, let us return to Haseloff’s stylistic theory as to the fourth century date of the Berlin and Nevers fragment. It has already been admitted that an obvious gap sep- arates the style of the two fragments from the style of the Milan covers. The difference is manifest in the more sketchy carving on the Milan covers, in the more intense and angular bodies and the more energetic movement of the figures. In other words, the work on the Berlin and Nevers frag- 24 See p. 67 . 2 44 A SCHOOL OF IVORY CARVERS IN PROVENCE ments is superior to, and therefore earlier than, the work on the book covers ; it shows a superior execution, greater plastic sense, and a more classic handling of draperies. While earlier than the Milan covers, is the work also slightly anterior to the Probianus diptych (Fig. 165), as Haseloff asserts? Waiving the evidence as to date adduced above, I would say, “No.” Although a marked likeness, and one that must be considered, exists between the two works in the flat backgrounds, in the build of the figures, in the shape of the heads, in the cut of the hair, and in the cos- tumes, the style of the two fragments seems inferior to that of the Probi- anus diptych. They appear to represent a direct, though later, continuation or reminiscence of the better style that occurs on the Roman diptych; the figures are fatter, more stubby, and in certain cases are almost repulsively bloated. Since the Berlin and Nevers fragments are no longer to be considered as Roman works, but as Provengal ivories, how can any stylistic similarity between them and a Roman diptych bear on their date? It is here that I feel that the attribution of the ivories to Provence removes the weak point in Haseloff’s theory. His principal reason for setting the Berlin fragment at the end of the fourth century was the fact that after the be- ginning of the fifth century, when the Probianus diptych was carved, the work on Roman consular diptychs appears to have ceased and that, when they commenced to be carved again, their execution was coarse, crude, and hasty. It is this sudden cessation of Roman ivory carving that solves the problem and bridges the gap between Provence and Rome. For about a generation after the beginning of the fifth century artistic production stopped at Rome. Even before Alaric actually entered Rome in 410, fear drove away a large part of the population. All classes, though principally the patricians and the artisans, fled to other parts of the Empire. So complete was the flight of craftsmen and so slight was the inducement to return to Rome after the state had been freed from the invaders, that legislation was necessary to force the artists to return. 25 At this time sarcophagus carving practically stopped at Rome; and the craftsmen ap- pear to have emigrated in great numbers to Provence, for about the be- ginning of the fifth century there sprang up a fully developed school of 25 Frothingham, Monuments of Christian Rome, p. 52. A SCHOOL OF IVORY CARVERS IN PROVENCE 245 Fig. 165. Berlin: Museum, diptych of Probianus. 246 A SCHOOL OF IVORY CARVERS IN PROVENCE sarcophagus carvers at Arles who carried on the traditions of the Roman school. 26 The bearing of such an exodus of artists from Rome at the beginning of the fifth century upon the Provencal school is obvious. While the Berlin and Nevers fragments are Provenqal because of their affinities with the Milan and Werden ivories from that region, we have also noted that these two fragments bear a marked similarity to early Roman ivory carving such as appears on the Probianus diptych of c. 400. Furthermore, the rather curious palmette ornament on the Milan covers and also on the Rouen diptych is similar to the moulding on the Symmachi and Probianus Fig. 166. Marseilles : Museum, sarcophagus. diptychs. The conclusion to be deduced seems to satisfy all these problems. The ivory carvers of Rome, like the sarcophagus carvers, must have also fled to Gaul and have continued in that region many traditions of the imperial school. If the monastery of St. Victor, or even Marseilles, is accepted as the probable centre of this ivory school, it seems likely that the monastery, which was founded shortly after 414, or certainly the city, received some of these Roman artisans, who founded a local school. Sometime between the end of the first quarter of the fifth century and the end thereof, as the iconography would prove, these artists or their native pupils carved the diptych, or diptychs, of which there remain only the Nevers and Berlin fragments, the earliest examples of the Provencal group. While they preserved, instinctively, in this diptych the stylistic tradition of the Roman school before the year 410, they, at the same time, had to adapt themselves to the semi-Oriental taste of the 26 See The Orientalizing of Gaul. A SCHOOL OF IVORY CARVERS IN PROVENCE 2 47 community. Therefore, in certain scenes they followed the newly com- posed and popular Pseudo-Matthew, executed a particular type of the Massacre of the Innocents created after their relics had been brought to Provence by Cassianus, and combined the Eastern iconography they found with the Hellenistic forms they had brought with them from Rome. Hence the confusing eclectic character of the ivories. About all has been said in regard to the date of the Berlin and Nevers fragments that can be said; being earlier than the Milan covers and Fig. 167. St. Maximix : Sarcophagus. posterior to the end of the first quarter of the century, they would date 425-500. Again, the style of the two fragments appears to be a little earlier than that of the Rouen diptych, which I dated between 450 and 475 A. D. The style on the Rouen diptych, while superior to the work on the Milan covers, seems closer to the more nervous work on the covers than to the heavy and stolid style of the Berlin and Nevers fragments. I have therefore assigned the fragments to a date c. 450; the bearing of this date on that of Pseudo-Matthew has already been pointed out (p. 191). The thickset character of the work on the fragments recalls in many ways a small class of sarcophagi in Provence. Certain sarcophagi from Mar- seilles 2 ' (Fig. 166), St. Maximin 28 (Fig. 167), Arles 29 (Fig. 168), and \ ienne 30 show figures whose large hands, bloated dimensions, features and costumes are noticeably similar in general accent to the work on the Berlin and Nevers fragments. In the original diptychs (or diptych) of which these fragments were parts we find the first fruits of what I have called 27 Le Blant, Lcs sarcophagcs chrctiens de la Gaule, pi. XVI/2. 2 * Le Blant, op. cif., pi. LIV/l. 29 Le Blant, op. cit., pi. X/i. 30 Le Blant, op. cit., pi. V/4. 248 A SCHOOL OF IVORY CARVERS IN PROVENCE Fig. 168 Arles : Sarcophagus cover. the Provencal School of Ivory Carvers, and it is on the assumption that Roman ivory workers came to Provence that we may best explain the curious eclectic character of the work on the group. 31 VII IVORY BOOK COVER IN THE BODLEIAN LIBRARY, OXFORD Book cover. Oxford Date, end of eighth or beginning Bodleian Library, of ninth century. In the Bodleian Library at Oxford is one leaf of a small diptych (Fig. 169) which, although it is a Carolingian work of the eighth or ninth cen- tury, is related to the Provengal school of ivories and has often been in- cluded in the group under discussion. This relation consists in its being a partial copy of the Provencal diptych of which the Berlin fragment was a part. A similar relation exists between the fragment of a Carolingian plaque at Munich and the Berlin ivory (see p. 64, note 15). The interest of this cover increases as its connection with Provence and its use of cer- tain Alexandrian-Coptic types of scenes help to explain in what manner Oriental and so-called Syro-Egyptian motifs and iconographical types en- tered Carolingian and Ottoman art. 1 The cover is divided into twelve compartments, three on each side, which form a border about a large rectangular panel. Within this central panel the Redeemer, young and beardless, is depicted trampling under foot the lion and the dragon, the asp and the basilisk. In the small panels across the top of the cover, reading from left to right, are figured the Prophet Isaiah standing by a tree and holding a scroll inscribed with the 31 It is interesting to note that Dalton, Byz. Art and Arch., p. 203, remarks that the Berlin and Nevers fragments are the earliest of the group. 1 Sauerland and Haseloff, Der Psalter Erzbisclwf Egberts von Trier, 1901, pp. 104, 136. A SCHOOL OF IVORY CARVERS IN PROVENCE 249 words ECCE VIRG(O) CONCI(PIET ), 2 and the scenes of the Annun- ciation and the Nativity. Continuing down the right hand side of the cover are carved the Adoration of the Magi, the Massacre of the Innocents, and the Baptism. Across the bottom, from right to left, are the Miracle of Cana, Christ Asleep in a Ship with Three Apostles on the Sea of Galilee, and the Raising of Jairus’ Daughter. Reading up the left side of the cover are Christ Driving the Demons into the Herd of Swine, the Para- lytic, and the Woman with an Issue of Blood. It is curious that an ivory whose iconography so clearly establishes its provenience and in a general way its date should have been so poorly attributed as the Bodleian cover. Many writers have attributed it to Italy : Westwood 3 and later Masked 4 have called it an Italian work of the ninth or tenth century; Durand 5 has said that it was in the Roman style and could not be dated before the end of the eighth century; and Strzygowski 6 has called it sixth century work from Ravenna. Stuhlfauth , 7 realizing that there was something odd and apparently inexplicable about the ivory, declared that it was an “offenbare moderne Falschung.” Although he refrained from either dating or attributing the cover, Beissel 8 ' said that it was a work done in the manner of the five part diptychs from Syria and Egypt, but was copied from a Roman or Gallic model of the fifth or sixth century. This analysis is remarkably near the truth, as is ETaselofFs as- sertion that the Bodleian cover is a Carolingian ivory, to be connected with certain Carolingian ivories from Metz, and was copied from the “Roman” diptych of which the Berlin fragment was a part . 9 Inasmuch as the Bodleian cover is not a member of the Provence group. I intend only to point out its connections with Provence and to discuss those scenes which prove its Carolingian origin. Much of the iconography on the cover, if not copied directly from the Provenqal diptych, of which the Berlin fragment was a part, was derived either from the scenes on the 2 Is. vii, 14. 3 Westwood, Fictile Ivories, p. 55, plate VI. 4 Masked, Ivories, 1905, p. 94. ’ Durand, Annales archeologiques, i860, XX, p. 122. c Strzygowski, Byz. Denk., I, p. 46. 7 Stuhlfauth, Die altchristliche Elfenbeinplastik, p. 191. 8 Beissel, Geschichte der Evangelienbiicher in der ersten Hdlfte des Mittelalters, Frei- burg im Breisgau, 1906, p. 304. p Haseloff, Jb. Preuss. Kunsts., XXIV, 1903, p. 60. 250 A SCHOOL OF IVORY CARVERS IN PROVENCE Fig. 169- Oxford : Bodleian Library. Ivory book cover. A SCHOOL OF IVORY CARVERS IN PROVENCE other Provencal ivories or from the scenes on the early Christian sar- cophagi of Gaul. In the first place the three scenes on the Berlin fragment, the smashing type of the Massacre of the Innocents , 10 the Provenqal type of the Baptism , 11 and the Alexandrian-Coptic type of the Miracle of Cana , 12 occur in the same place and order on the Bodleian cover. Aside from a difference in style the Massacre and the Baptism on the two ivories are nearly identical, while the Miracle of Cana on the Carolingian ivory shows the jars increased from four to six after the customary number on Eastern monuments from the sixth century on. Furthermore, the Nativ- ity 13 and the Adoration 14 on the cover recall by their rendering the same scenes on the Milan book covers. Other scenes on the ivory, like the Woman with an Issue of Blood and the Raising of Jairus' Daughter, were taken from the types on the sarcophagi of southern Gaul . 15 It is not impossible, then, so far as the iconography of these scenes leads us, to consider the Bodleian book cover a Provenqal ivory of the fifth century. Turning to the distinctively Carolingian types on the cover, Christ Triumphant over the Four Beasts is the most valuable scene in establishing the date and origin of the ivory, and for this reason I have treated it separately . 16 The results of this study of Christ Triumphant are clear and conclusive. The type itself originated in Alexandria and then, like other Alexandrian types, passed into Coptic art. Thence it must have been introduced directly into Gaul and subsequently transmitted into Caro- lingian art. for the developed type, where Christ tramples on the four beasts, outside the early Christian monuments of Egypt only occurs on the Frankish monuments of the Carolingian and Ottoman periods. The scene on the Bodleian cover, in its composition, its beardless type of Christ, and in its nervous draperies, finds its only convincing parallels in the minia- tures of the Godescalc school and two Carolingian ivory book covers which, in their turn, are connected with this same Carolingian school of miniature painting. Therefore, on the basis of this scene alone the Bodleian cover appears to date from the end of the eighth century or the very beginning of the ninth and to have come from that centre in northern France where the famous Godescalc. Ada, and Soissons Gospels were executed . 17 10 See p. 64. 11 See p. 78. 12 See p. 9 2. 13 See p. 18. 14 See p. 44. 1 ’ Le Blant, Lcs sarcophages chrctiens dc la Gaule, pi. V/i. 16 See Christ Triumphant, p. 146. . 17 See p. 157. 252 A SCHOOL OF IVORY CARVERS IN PROVENCE Other scenes which attest the Carol ingian workmanship are the An- nunciation and the scene of Christ Asleep with Three Apostles on the Sea of Galilee. The two distinctive features in the scene of the Annunciation, wherein the Virgin is seated as the angel accosts her, are the female at- tendant who stands behind the Virgin, and the curious tower which rises at the right of the composition. While the Annunciation was frequently- represented on the monuments of early Christian art, the attendant in the scene does not occur until the eighth or ninth century and then on Caro- lingian monuments. A Carolingian reliquary of the eighth century in the Sancta Sanctorum 18 and another ivory of the tenth century in the Louvre 19 figure an attendant behind the seated Virgin. Again in the type of the Annunciation with the standing Virgin, the first use of the attendant is on a Carolingian ivory in the Bibliotheque Nationale in Paris. 20 A minia- ture of the Soissons Gospels, 21 a work of the Godescalc school, represents an Annunciation, which, with the one exception that the attendant does not appear in the composition, is noticeably close in other respects to the scene on the ivory; the pose and draperies of the Virgin and the form of the chair in which she is seated are about the same in the two examples. While no examples of Christ Asleep in a Ship on the Sea of Galilee exist, to my knowledge, on any early Christian monuments, the scene oc- curs frequently in the manuscripts of the Ottonian period in northwestern Europe. These Ottonian examples, whose iconography was naturally de- rived in large measure from Carolingian monuments, are very similar to the rendering of the scene on the Bodleian book cover. 22 While the scene seems to be absent from early Christian monuments, we have nevertheless the testimony of Perpetuus that it was represented in the fifth century on the walls of the church of St. Martin at Tours. 23 A summary of the iconographical types on the cover supports the con- clusion indicated above. While the Carolingian character of the work is evinced by the scenes of Christ Triumphant, the Annunciation, and Christ 18 Lauer, Monuments Piot, XV, 1903, pi. VII. 19 Fleury, La Salute Vierge, I, pi. XIII. 20 Labarte, Histoire des arts industriels, Album I, pi. V. 21 Laurent, Les ivoires pregothiques, pi. IV. 22 Codex Egberti (Haseloff, op. cit., pi. XXIV), Pericopes at Munich (Swarzenski, Regensburger Buchmalerei, pi. XXVI), Ms. of Emperor Otto at Aachen (Beissel, Die Bilder der Handschrift des Kaisers Otto im Munster zu Aachen, 1886, pi. VI). 23 J. von Schlosser, Quellenbuch zur Kunstgeschichte, p. 32; Kraus , Geschichte der christlichen Kunst, I, p. 602 sq. A SCHOOL OF IVORY CARVERS IN PROVENCE 2 53 Asleep in a Ship, its connection with the Provengal ivories appears in the Nativity, the Adoration, the Massacre of the Innocents, the Baptism, and the Miracle of Cana. Further connection with the early Christian types of Provence lies in the Raising of Jairus’ Daughter and in the scene of the Woman with an Issue of Blood, which both appear on Gallic sar- cophagi. Besides the Carolingian and Provenqal types, the cover presents in the scenes of Christ Triumphant, the Healing of the Paralytic, and the Miracle of Cana certain Alexandrian-Coptic types. These Egyptian types, arriving by the trade route which existed between Egypt and the ports of southern Gaul, 24 must have entered Provence through Marseilles, as Strzygowski has suggested in the case of Egyptian motifs appearing on the ivory pulpit of Aachen, 25 and were then transmitted through Mero- vingian times to Carolingian art. It is interesting, in the light of these Egyptian types appearing on the ivories, to see how my iconographical deductions were already grasped in a general way by Beissel. While Haseloff claimed that the Bodleian cover was a Carolingian copy of a Roman ivory, Beissel recognized in the cover a Syro-Egyptian manner, but said that it had been copied from either a Roman or Gallic model of the fifth or sixth century. In discussing the Berlin fragment, which was the fifth century prototype for the Bodleian cover, I have shown how the Massacre of the Innocents and the Baptism on it were Provengal types, how the Miracle of Cana was an Alexandrian- Coptic type, and how the work resembles the Roman ivory carving of the beginning of the fifth century. The explanation of this curious mixture of Gallic and Egyptian types and a Roman style may well lie in the emi- gration of Roman ivory carvers, after the sack of Rome in 410, to Pro- vence, where they commenced to work for a semi-Oriental population and consequently had to adopt many Eastern ideas and motifs. Thus my arguments for the Provencal origin of the Berlin fragment, from which the Carolingian ivory was partly copied, account for the conflicting and 24 Besides the commercial and monastic affinities which united Marseilles and Egypt (see Orientalizing of Gaul, p. 195) Egyptians were numerous enough at Arles to have an oratory in honor of St. Menas (De Rossi, Bull, di arch, crist., 1869, p. 32) and in Marseilles we have examples of the name Menas (Le Blant, Inscript, chret. de la Gaule, 1865, II, p. 309, no. 55 la ; “Bonam requiem habeat in die futuro maritus meus Menas”) ; Cabrol, Dictionnaire, etc., s. v. Ampoules, col. 1 727. 25 Strzygowski, Hell, und Kopt. Kunst., p. 69. 254 A SCHOOL OF IVORY CARVERS IN PROVENCE apparently inexplicable elements in the Bodleian cover which Beissel recog- nized and which probably led Stuhlfauth to call the work a forgery. The style of the Bodleian cover confirms its attribution. The mouldings around the panels are conventionalized and crudely exe- cuted copies of the egg and dart and bead and reel mouldings on the Berlin fragment. The nervous drapery of the figures, which I have dis- cussed more fully in my treatment of Christ Triumphant, 26 is the distinctly Carolingian feature of the style. Inasmuch as it is not as animated as the drapery of the Carolingian work of the later ninth and the tenth cen- turies and as it resembles very closely the form of drapery in the miniatures of the Godescalc school, I would date the work at the end of the eighth or beginning of the ninth century. This date is sustained by a comparison of the Bodleian cover with later Carolingian ivories. Although cer- tain points of iconography associate the Bodleian ivory with two Caro- lingian book covers in the Bibliotheque Nationale at Paris, 27 which were probably carved at Metz, 28 the styles are entirely different. The two Paris ivories, which are very accomplished and pretentions pieces of ivory carv- ing, can not be much, if any, earlier than the tenth century, while the poor modelling, the lack of animation in the figures, the comparatively mild drapery, and the rather hesitating execution of the Bodleian cover date it about a century earlier than the Paris examples. 29 Therefore the Bodleian ivory may be said to date from about 800 and to have been carved in that centre of northern France where the Godescalc school of miniaturists flourished. 26 See Christ Triumphant, p. 158. 27 Covers of Mss. lat. 9393 and 9388. 28 Haseloff, op. cit., p. 61. 29 Another ivory in the Victoria and Albert Museum (Graeven, no. 67) is to be classi- fied with these two covers in the Bibliotheque Nationale. INDEX INDEX Monuments are printed in italics and listed under place, which is printed in CAPITALS. Collections, Churches, Libraries, Museums, etc., are printed in roman small capitals. Authorities and sources, ancient and modern, are printed in CAPITALS. Iconography is indicated by printing the subjects of scenes in ITALIC CAPITALS, but details of iconography are printed like ordinary topics in Roman type. Full face Numerals indicate special discussions of topics. AACHEN ivory reliefs, 147, 253 Abbeville, see MIANNAY Abraham, St. 196 Ada Gospels, see TRIER ADORATION OF MAGI, see MAGI; Type of NATIVITY 13, 18, 22, 23, 161, 166; OF OX AND ASS, see OX AND ASS; OF SHEPHERDS, see SHEPHERDS ADULTERESS, see WOMAN TAKEN IN ADULTERY aedicula 100; in RAISING OF LAZARUS 108-m, 113, 119, 163, 167, 211 AFRICA lamps from 151-153 AINALOFF 4, 1 14, 167 AIX sarcophagus 235 AKMlM lamp from 151 ; medallion from 83, 166; textile from 152 Alaric 244 Alexandria 3, 4, 6, 24, 25, 28, 39, 40, 57, 81, 82, 88, 89, 104-106, 114, 126, 147, 160, 167, 168, 169, 170, 174, 196, 197, 200, 202, 204 note, 228, 251 ALEXANDRIA catacomb frescoes 6, 131-133, 148, 149, 151-154; copper cover from 87; Hellenistic portraits 48 Alexandrian-Coptic School 6, 159, 167, 186, 248 Type of ADORATION OF THE MAGI 37, 47, 48, 175, 184 of ANNUNCIATION, 171, 175 of BAPTISM 72, 79, 82, 84, 176, 180, 228 of ENTRY INTO JERUSALEM 122, 126, 177 of HEALING OF THE BLIND 94, 98, 176 of HEALING OF THE PARALYTIC 10 2, 105, 180, 253 of JOSEPH ASSURED BY AN ANGEL 173 of JOURNEY TO BETHLEHEM 174 of MASSACRE OF THE INNOCENTS 176 of MIRACLE OF CANA 85, 88, 90, 106, 166, 176, 242, 251 of NATIVITY 13, 17, 22, 23, 174 of RAISING OF LAZARUS 108, 109, 117, 177, 181 of SAMARITAN WOMAN AT THE WELL 204 and note of TESTING OF THE VIRGIN 173 ALGIERS sarcophagus 86 altar in COMMUNION OF THE APOSTLES 140; in NATIVITY 14, 162 AMBROSE, St. 20 AMIENS mallet Collection : ivory fragment 237 ampullae 178; see MONZA 255 256 INDEX amulets, see ROME Vatican ANANIAS 204 ANASTASIS, in Martyrion at Jerusalem 178 Anatolia see Asia Minor anaxyrides, see Costume Andrew, St., apostle 131 angel, in ANNUNCIATION n-13, 172, 173, 252; in BAPTISM 79-81, 83; in JOSEPH ASSURED BY AN ANGEL 202, 22 5; with MAGI 34, 47-50, 52, 56, 57, 175, 180, 184; in VIRGIN IN THE TEMPLE 8, 9, 223; bearded 83; nimbed, see Nimbus angels, in ADORATION OF THE MAGI 37, 52, 53! in BAPTISM 72, 74/78-81, 83, 84. 176; in CHRIST TRIUMPHANT 148, 150, 151, 154, 157 ; in COMMUNION OF THE APOSTLES 140; in NATIVITY 31, 32, 59, 184; nimbed, see Nimbus Anglo-Saxon Gospels see BRUSSELS Animal Kingdom, Adoration of 13, 19 ANNUNCIATION 171, 175, 249, 252; AT THE SPRING ix, 12 note, 208, 210, 211, 213, 220, 222, 223; TO SHEPHERDS 32; to Zacharias 223 ANTINOE comb from 118, 119; frescoes 5, 62, 92, 171, 174, 176, 180, 202, 213, 225; textile from 127 Antioch 3. 4, 160, 197, 200 APOCRYPHAL GOSPELS 19, 22, 30, 175, 178, 189-191, 201-203, 206, 210, 211, 213; see also APOCRYPHUM, HISTORY, PROTEVANGELIUM, PSEUDO-MAT- THEW, TWELVE APOSTLES APOCRYPHUM JOSEPHI 27 APOCRYPHUM ZACHARIAE 27 Apollo, Apa, Monastery of 27 Apostles, see Andrew, COMMUNION OF APOSTLES, Disciples, HOLY APOSTLES, John, Paul, Peter, TWELVE APOSTLES APPEARANCE OF THE STAR 33, 43, 57, 203, 208, 210, 21 1, 213, 215, 216, 221, 222, 225, 226, 229 apse 1 17, 185 Aquitaine 194 Arabs 168, 183 arcade, ornamental motif 199 arch 1 15, 1 16, 167 Arenberg, Gospels of, see BRUSSELS Arianism 168 Arles 193-195, 201, 218 ARLES sarcophagi 34, 66, 78, 86, in, 113, 197, 203, 215, 216, 234, 247; S. TROPHIME: sarcophagus 21 Armenia 217 Armenian Miniatures, see ETSCHMIADZIN ars barbaricaria 200 ASCENSION , in church at Jerusalem 178 Asia Minor 1, 2, 5, 39, 99-101, 107, 109, 114-116, 120, 122, 128, 135, 140, 181-183, 188, 201, 204, 217, 219, 229; Type of Christ, see Christ asp, in CHRIST TRIUMPHANT 146, 148, 156, 158, 248 ass, in ENTRY INTO JERUSALEM 123, 126, 164, 177, 181 ; in NATIVITY 23; see also OX AND ASS: head of, in MASSACRE OF THE INNOCENTS 62 Athele 194 Athens 150; see also DAPHNI ATHOS, Mt. manuscript 121 ; steatite panel 121; monasteries: chilandari 140: vato- pedi 140 attendant, see Servant AUCH sarcophagus 113 Augustine, St. 19 AVIGNON gold work 218 ax 22 / INDEX -57 B BAGAI lamp from 151 BAGAWAT frescoes 6, 169 BALAEUS 55 , 58 banquet, see Feast baptism 103, 163 BAPTISM, of Christ 71 , 162 , 176 , 180 , 208, 21 x, 214, 215, 222, 227, 229, 237, 242, 249, 251 Baptist, see John the Baptist Baptistery, see NAPLES, RAVENNA barberini Collection, see ROME Bartimeus 95, 96 Basil II, Mniologium of, see ROME: Vatican basilewsky Collection, see PETROGRAD basilisk, in CHRIST TRIUMPHANT 146 , 248 BASILIUS, bishop of Caesarea 34 baskets, see Loaves Bawit 169 BAWlT frescoes 5, 25-27, 62, 79, 81, 170-172, 175, 176, 224 BAYET 36 beard, see Angel, Christ, MAGI, Paul beasts, Christ trampling upon, see CHRIST TRIUMPHANT bed, in HEALING OF THE PARALYTIC 102-106, 163, 165, 185, 211 BEISSEL 249, 253, 254 BENEVENTO cathedral: bronze doors 70, 94 BERLIN museum : ivory fragment 18, 62, 64, 67, 76, 77, 90-92, 187, 191, 203, 214, 237 , 249, 251, 254; date 241-243, 247: technique 238, 244, 247; Alexandrian pyxis 105, 198; ampulla 83 and note; bone carving 105; diptych of Probianus 240, 244, 246; German ivories 28; Lombard relief 44; Minden pyxis 25, 174 Gans Collection : medallion 89, 172, 176 Bernward, Gospels of, see HILDESHEIM Berthramnus, 195 Bes, mask of 148 Bethesda, Pool of 102, 106 Bethlehem 19, 22, 190 BETHLEHEM, JOURNEY TO, see JOURNEY TO BETHLEHEM BETHLEHEM church of the nativity: 5, 178; mosaic 29, 53, 55, 179, 180 bianchixi Collection, see ROME BI6NKOWSKI 38 BLIND, HEALING OF 94 , 163 , 166 , 176 , 184 , 208, 210, 212 BLIND AND HALT, HEALING OF 98 blind man, men, in HEALING OF THE BLIND 94-98, 100, 101, 163, 166, 184, 212; eyes of, touched by Christ, see Eyes ; head of, touched by Christ, see Head BOLOGNA Coptic ivory 25, 36, 114; pyxis 97, 104, 118, 166, 167 bolster 53, 138 BONN Coptic pyxis 118 book 152, 156, 157 book covers, see MANCHESTER, MILAN, OXFORD, PARTS, RAVENNA, ROME, SOUTH KENSINGTON Bordeaux 194 borghese, Villa, see ROME INDEX 258 BOTTARI, sarcophagus described by 95 braid, ornamental motif 198 bread, in COMMUNION OF THE APOSTLES 140; in LAST SUPPER 129, 130, 134- 136, 139 , 212 BRESCIA ivory casket 9 7, 114, 115, 166, 167, 204 and note, 229, 230; Lampadiorum diptych 240 British museum, see London Brittany 194 BRIVIO, casket from, see PARIS bruls Collection, see ROME BRUSSELS Gospels of Arenberg 154; diptych of Genoels-Elderen 156-158, 251; musee du cinquantenaire : ivory book cover 154; leaf of diptych 235 BUGATI 220 Bulgaria 116 BUST OF KING, THREE HEBREWS BEFORE 216 Byzantine art 1, 57, 140, 154, 155, 165, 181, 182, 185, 188, 234; cloisonne 217; costume, see Costume; ivories 93; manuscripts 11 Type of ADORATION OF THE MAGI, 37, 47, 5L 58, 184 of BAPTISM 84 of CHRIST AMONG THE DOCTORS 68, 69 of ENTRY INTO JERUSALEM 122, 124, 125, 127, 185 of HEALING OF THE BLIND 95, 100, 101, 184 of HEALING OF THE PARALYTIC 102, 106, 185 of LAST SUPPER 135, 140, 185 of MASSACRE OF THE INNOCENTS 62 of MIRACLE OF CANA 85, 88, 93, 184 of NATIVITY 13, 22, 26, 30, 184 of RAISING OF LAZARUS 108, 109, 120, 185 of SAMARITAN WOMAN AT THE WELL 204 and note Byzantium, see Constantinople C Caenomani 230 Cairo 126, 127, 1 71 , 177 CAIRO museum: 6, 170; censer 62, 213: cover of earthen vessel 1 5 1 CALENDAR of 354, 42 CALLIXTUS, see ROME catacombs CAMBRIDGE Gospels 124, 137 CANA, MIRACLE OF 85, 105, 132, 133, 163, 166, 176, 180, 184, 208, 210, 212, 237, 242, 249, 251, 253 Canatha (Kanawat) 194 Capernaum 102, 106 CAPPADOCIA churches: frescoes 140 Carolingian Art 28, 44, 67, 90, 92, 146, 1 53-155, 157, 192, 219, 230, 231, 241, 248, 251-254; ivories, see BRUSSELS, METZ, OXFORD, PARIS bibliotheque nationale, ROME VATICAN, SOUTH KENSINGTON carpet, in ENTRY INTO JERUSALEM 126, 177 CARTHAGE lamps from 151, 152; relief from 48 Cassianus, abbot of S. Victor, Marseilles 67, 68, 196, 202, 203, 214, 233, 236, 241, 247; bishop of Autun 196 catacombs, see ALEXANDRIA, NAPLES, ROME cathedra (or chair) 44, 49, 64, 70, 172, 173, 175, 226, 241 ; of Maximianus, see RAVENNA cave, NATIVITY in 13, 19, 20, 22, 32, 45, 184, 190; in RAISING OF LAZARUS no, 120, 185 INDEX 259 censer, see CAIRO, KERTCH, LONDON, ODESSA centurion, in CRUCIFIXION 188 note chair, see Cathedra chalice, in COMMUNION OF THE APOSTLES 139. 140 CHAPEL OF JOHN VII, See ROME CHURCHES chilandri, see ATHOS Child, Christ, see Christ Childeric, tomb of, see TOURNAY children, in MASSACRE OF THE INNOCENTS 61-64, 66 chlamys, chlamydes, see Costume CHORIKIOS 141 Chosroes 54 Christ, bearded 72, 78, 79, 83, 84, 176, 178, 180, 205; beardless 73, 76, 77, 79, 81, 162, 176, 248, 251 ; Child, see ADORATION OF OX AND ASS, ADORATION OF SHEP- HERDS, CHRIST AMONG THE DOCTORS, MAGI, NATIVITY; costume of, see Costume; enthroned 70, 131, 154; on globe, see Globe; nimbed, see Nimbus; Asia Minor type of 205; Palestinian type of 205; in BAPTISM 72-84, 162, 176, 180, 214, 228, 242; in CHRIST TRIUMPHANT 146, 148-158, 248, 251; in COM- MUNION OF THE APOSTLES 139, 140, 185; in CRUCIFIXION 187 note, 205; in ENTRY INTO JERUSALEM 122, 123-128, 164, 177, 181, 185, 212; in HEALING OF THE BLIND 94-98, 100, 101, 163, 166, 176, 184, 212; in HEALING OF THE PARALYTIC 102-106, 163, 180, 21 1; in LAST SUPPER 129, 134-137, 185; in MIRACLE OF CANA 85-87, 89-93, 132, 163, 176, 184, 212; in MIRACLE OF LOAVES AND FISHES 130-133 ; in PRESENTATION OF THE CROWNS 141-144; in RAISING OF LAZARUS 108-113, 117-120, 163, 177, 181, 185, 211; in SAMARITAN WOMAN AT THE WELL 204 and note; in WIDOW'S MITE 144-146. CHRIST AMONG THE DOCTORS 68, 208, 211, 213; DRIVING DEMONS INTO SWINE 249; ENTERING JERUSALEM, see ENTRY INTO JERUSALEM; FEEDING THE MULTITUDE , see LOAVES AND FISHES; HEALING THE BLIND, see BLIND; HEALING THE PARALYTIC, see PARALYTIC; RAIS- ING LAZARUS, see LAZARUS; AND SAMARITAN WOMAN, see SAMARI- TAN WOMAN; IN SHIP ON SEA OF GALILEE 249, 252; TRIUMPHANT 106, 146, 187, 251-254 CHRONICON Paschale 81 ciborium 140; columns, see VENICE s. marco CINQUANTENAIRE, MUSEE DU, See BRUSSELS city, representation of 227, 230; in ENTRY INTO JERUSALEM 123, 124, 128, 164; in VISITATION 224, 230 CIVITA CASTELLANA sarcophagus 89 Claudius Julianus 193 CLEMENT, of Alexandria 40, 168 Clermont-Ferrand 194, 196 CLERMONT-FERRAND sarcophagus 1 1 3 cloisonne, see Enamel cluny, Musee de, see PARIS Codex Cambricensis, see CAMBRIDGE; Egbert i, see TRIER; Purpureus, see MUNICH; Rossanensis, see ROSSANO coenobium, 195 coiffure, Egyptian 153; Gallo-Roman 219 colobium, worn by Christ in CRUCIFIXION , 205 Cologne 194 Columbanus, St. 194 column, drum of, see CONSTANTINOPLE ottoman museum 26 o INDEX COMING OF THE ORIENT TO CHRIST 37 COMMUNION OF THE APOSTLES 129, 139. 185 conch shell, ornamental motif, 199, 234 Constantine, Emperor, 182, 217; of Germanicia 200 Constantinople 3, 4, 39, 59. 84, 88, 93, 124, 181-184, 188, 194, 196, 217, 218 CONSTANTINOPLE hagia Sofia: 182, 204 note; holy apostles, Church of: mosaics 5, 31 note, 121, 128, 183; kahrie djami: mosaics 11; relief 49 ottoman museum: drum of column 84; encolpium 22, 47, 97, 105, 117, 165-167, 204 note; Phrygian relief 43, 115, 116, 167; relief 117, 167 consular diptychs 240, 244 controversies, iconoclastic, see Iconoclastic Coptic Art 51 , 56, 62, 11 7, 147, 166, 169 , 170, 174, 179, 251; cross, see Cross; influence on NATIVITY 26 ; literature 149, 175; medallions, see Medallions ; seal, see LONDON British museum; textiles, see Textiles Type of HEALING OF THE PARALYTIC 102, 106 , 180 ; RAISING OF LAZARUS 108, 109, 1 14, 1 18 , 180 Cosmas Indicopleustes, manuscript 6 costume of Baptist 73, 77, 79, 162, 214, 215; Byzantine 219, 220; on Chair of Maximianus 170; of Christ 37, 120, 153, 158, 185, 205; of Elizabeth 224; of Joseph 225; of Magi 34, 35, 38, 43, 44, 49, 57, 175, 216, 220, 226, 241; Merovingian 219; on Milan book covers 219; of Orientals 50; on Rouen diptych 236; of soldiers, in MAS- SACRE OF THE INNOCENTS 220; of tribune, in MASSACRE OF THE INNOCENTS 64, 66, of Virgin 7, 219, 220, 223, 224, 226, 252 couch in ADORATION OF THE MAGI 39: in JOSEPH ASSURED BY AN ANGEL 225; in LAST SUPPER 129, 130, 134, 136, 138, 140 COURAJOD 196, 198 crib, introduction of, in NATIVITY 15 , 162; ‘‘Roman,” 15; see also MAGI, OX AND ASS, NATIVITY crocodiles 148 crook, shepherd’s, see Pedum cross 82, 94, 98, 103, 106, 109, 118, 119, 126, 127, 1 33, 150-153, 156, 157, 173, 176, 177, 180, 181, 218; on dish from Gourdon 218; on Milan book covers 208, 216-218 crowd, see Spectators crown, crowns 141-143; worn by MAGI 57, 58; crenellated 126 CROWNS, PRESENTATION OF 141 , 208, 210, 211 CRUCIFIXION 187 note; “colobium” Type of, Narbonne 205 curtains, ornamental motif 199; on Milan book covers 208, 218 Cyprus 1 16 CYPRUS, encolpium from, see strzygowski Collection D dais, in MASSACRE OF THE INNOCENTS 62 dalmatica, see Costume DALTON 4, 11, 23, 54, 189, 217, 220, 233 DAPHNI monastery: mosaics 11, 51 DARMSTADT pyxis 105, 113 Deir, in Lebanon 193 Deir-Abou-Hennys, see ANTINOE DELATTRE 49 demons, see CHRIST DRIVING DEMONS INTO SWINE DE ROSSI 69, 150, 188, 220, 236 DE WAAL 223 DIEHL, 4, 22, 24 INDEX 261 Diocletian, emperor 188 Diptychs, see BERLIN, BRESCIA, BRUSSELS, Consular, MILAN, PARIS, ROUEN, SOUTH KENSINGTON, SYMMACHI disciples (Apostles) in CHRIST TRIUMPHANT 148, 152, 154; in COMMUNION OF THE APOSTLES 139, 140; in ENTRY INTO JERUSALEM 123, 124, 127, 128, 164, 185; in HEALING OF THE BLIND 100, 101, 184; in HEALING OF THE PARALYTIC 102, 104; in LAST SUPPER 129, 134, 136, 137, 185, 212; in MIRA- CLE OF CANA 85-87, 89-92, 163, 180; in MIRACLE OF LOAVES AND FISHES 131; in RAISING OF LAZARUS 109, 112, 121, 185; in WIDOW’S MITE 145 DOBBERT 137, 140, 141, 188, 220 doctors 69, 70 DOCTORS, CHRIST AMONG, see CHRIST DOMITILLA, CATACOMB of, See ROME doors, bronze, see BENEVENTO, MONREALE, PISA; wooden, of s. sabina, see ROME doria panfili, Villa, see ROME dove, in BAPTISM 73, 77, 80 dragon, in CHRIST TRIUMPH ANT 146, 148, 150, 152-154, 156, 158, 248 drapery, Carolingian 158, 251, 254 DREAM. JOSEPH’S, see JOSEPH ASSURED BY AN ANGEL drinking cup, Anglo-Saxon, see LONDON British museum DUTSCHKE 1 15 DURAND 249 E EBERSOLT 115 Ecclesia ex Circumcisione 37 Egypt 1, 2, 25, 29, 43, 50, 56, 57, 89, 92, 1 16, 119, 120, 122, 146-154, 157, 167-171, 174-183, 186, 193, 201, 213, 219, 224, 225, 249, 251, 253 Elizabeth 222, 224, 225, 230 EMMAUS, SUPPER AT 137 enamel, cloisonne 206, 208, 216-218 encolpium, see CONSTANTINOPLE, REGGIO, strzygowski Collection ENLART 230 ENTRY ISTO JERUSALEM 121, 164, 177, 181, 185, 208. 212, 221 Ephesus 3, 101, 160, 195 EPHRAEM SYRUS 20, 34, 55 Epiphany, Feast of, see Feast EPIPHANY, see MAGI; Type of ADORATION OF TH EMAGI 38, 41, 45 epitaphs, see Inscriptions ESTELLA, Spain, s. pedro de la rua : pyxis 133 ETSCHMIADZIN monastery: Gospels 5, 29, 49, 54, 56, 79, 105, 126, 157, 170, 172-175, 177, 178, 181 Eucharist 130, 133, 139, 141, 185, 212 Eucharistic Repast 129, 133; symbolism, see Symbolism; Transubstantiation 85 iu 163. 164 Euphron 194 Evangelists, see John, Luke, Mark, Matthew; symbols of 208, 219 exomis, see Costume eyes, of Blind Man, touched by Christ 94-97, 101, 163, 166, 184 F FAILLON 65, 66 262 INDEX FAYUM, portraits from 114, 120 Feast, Feasts 1 77, 178, 201; of Baptism 58; at Constantinople, at Christmas 59, 184; Eastern, in Gaul 196; of Epiphany 39, 55, 59, 191, 203; governor of, in MIRACLE OF CANA, 93; of Holy Innocents 67, 203; of Nativity 58, 191, 213; of Virgin, in Provence 10, 201, 221 ; in Rome 202, 213 fish 130, 131, 133-136 FLEURY, ROHAULT de 138 FLIGHT INTO EGYPT 52 FLORENCE pyxis 47; bargello: Rhenish ivory 155; laurentiana: Rabula Gospels from ZAGBA 23, 61, 62, 64, 79, 81, 83, 92, 98, 104, 127, 139, 157, 165, 167, 173. 180, 181, 184, 185, 204 note, 213 foliate design, in RAISING OF LAZARUS 109, 113, 164 forrer Collection, see STRASSBURG frankincense, see Gifts Freer Gospels, see WASHINGTON frescoes, see ALEXANDRIA, ANTINOE, el BAGAWAT, BAWIT, CAPPADOCIA, GAZA, MONTE RAPARO, NAPLES, S. ANGELO IN FORMIS, S. URBANO, SAKKARAH G Galilee, Sea of, see CHRIST IN SHIP ON SEA OF GALILEE GARRUCCI 142 Gaul 39, 68, 142, 146, 152-154, 157, 193-201, 203, 208, 212-219, 226, 230, 234, 236, 241, 246, 251, 253; Eastern feasts in 196; Eastern monastic influence in 195, 196, 205; Oriental iconography in 201; Orientalizing of 189, 192; relations of, with Syria 193-195, 199, 205; sarcophagi of 77, 92, 112, 113, 197, 198, 204 note, 218, 219, 251, 253; see also AIX, ARLES, AUCH, CLERMONT-FERRAND, MARSEIL- LES, MAS d’AIRE, MAS-ST. ANTONIN, NARBONNE, NIMES, PROVENCE, PUY, ST. GILLES, SAINT-RUF, SOISSONS, TOULOUSE, VIENNE; Sy- rian artists in 200; Syrian ornament in 198, 199, 206, 234 GAZA church of st. sergius : frescoes 141 Gelasius, Pope 221 gem, Merovingian 230 GENOELS-ELDEREN, diptych from, see BRUSSELS Gentiles, Adoration of 13, 19 Germanicia 200 GERONA sarcophagus 86 Gervasius, St. 142 gifts, offered by MAGI 38-41, 46, 53, 162, 191, 226, 241 glass, "Roman” 87, 103, 109-111, 113, 114, 164 globe, Christ enthroned upon 141-146, 208, 21 1 Gnostic sources 27, 189 God, representation of 148; Hand of 72, 78-80; Lamb of, on Milan book covers 206, 216- 218, on sarcophagus of ST. MAXIMIN 218 Godescalc, school of 157, 158, 251, 252, 254; Gospels of, see PARIS gold, see Gifts GOSPELS, canonical 85, 94, 102, 113, 127, 145, 190, 200, 201, 206, 208, 210, 223; see also John, Luke, Mark, Matthew; apocryphal, see APOCRYPHAL GOSPELS, APOC- RYPHUM, HISTORY, JAMES, PROTEVAN GELIUM, PSEUDO-MATTHEW, TWELVE APOSTLES gospels (manuscripts), see BRUSSELS, ETSCHMIADZIN, FLORENCE, HILDES- HEIM, LONDON, MUNICH, PARIS, ROME, ROSSANO, TRIER, WASH- INGTON INDEX 263 GOURDON cloisonne dish from 217 governor of feast, see Feast gradual, see PRUM GRAEVEN 4 grapevine, ornamental motif 198, 234 Greece 116, 182, 196 Greek chapel, see ROME GREGORY of Nazianzus 20, 58; Homilies of, see PARIS GREGORY of Nyssa 20 GREGORY of Tours 193, 194, 199, 205, 218 GRIMOtiARD de Saint-Laurent 220 GRISAR 151 H Hadelin, St., reliquary of, see VISE hagia Sofia, see CONSTANTINOPLE Hand of God, see God HASELOFF 188, 238, 240, 242-244, 249, 253 head, of Blind Man, touched by Christ 94, 95, 97, 163 HEALING, see BLIND, BLIND AND HALT, PARALYTIC, WOMAN WITH ISSUE OF BLOOD Hebrews, see Jews HEISENBERG 31 note, 32, 121, 183 helix, ornamental motif 198 Hellenistic School 2, 6, 72, 159, 200 Type of ADORATION OF THE MAGI 38, 162, 226 of BAPTISM 72, 73, 81, 162 of ENTRY INTO JERUSALEM 122, 125, 127, 164, 181 of HEALING OF THE BLIND 94, 95, 163, 176 of HEALING OF THE PARALYTIC 102, 103, 107, 163 of MASSACRE OF THE INNOCENTS 60, 162 of MIRACLE OF CANA 85, 163, 166, 176, 242 of NATIVITY, see ADORATION Type of NATIVITY of RAISING OF LAZARUS 108, 109, 114, 118, 163, 177 of SAMARITAN WOMAN AT THE WELL 204 and note Henry II, Gospels of, see MUNICH heresies 178; see also Arianism, Monophysite, Nestorianism Herod 46, 60-63, 208, 220, 230, 241 HILDESHEIM Gospels of Bernward 124; Missal 155 Historical Type of LAST SUPPER 129, 140, 185 HISTORY OF THE BLESSED VIRGIN MARY 22 HOLY apostles, Church of, see CONSTANTINOPLE Holy Innocents 67, 68, 203, 204, 214, 241, 243, 247; sarcophagus of, see ST. MAXIMIN HOLY WOMEN AT THE TOMB 240 Honoratus 196 horseman, lancing animal 147 Horus 147-149 house, in ADORATION OF THE MAGI 18, 43, 44, 46, 241; in ENTRY INTO JERU- SALEM 128; in HEALING OF THE PARALYTIC 106, 185 Hypatia 169 I Iconoclastic controversies 183 iconography, passim; value of, in determining schools 4 264 INDEX incense, see Gifts Innocents, see Holy Innocents; MASSACRE OF, see MASSACRE inscription, inscriptions of ARLES 193; on Bodleian book cover 156 157; in frescoes at ALEXANDRIA 132, 149; on lamp 151 ; of LYON 194, 200; on Merovingian gem 230; on scroll held by Isaiah 249; of Syrians, in Gaul 194, 198 Insulani 196 interlacings, ornamental motif 198 IRENAEUS, St. 39, 40, 96 Isaac of Armenia, Exarch, sarcophagus of, see RAVENNA Isaiah 248 Isis and Anubis, cult of, at Nimes 197 Italo-byzantine art 31, 36, 59, 69, 93, 94, 108, 121, 140, 184, 204 note Italy 1, 140, 154, 155, 188, 194, 195, 249 Type of MIRACLE OF CANA in 90 ivories, passim; Alexandrian 81 ; Byzantine 93; Provencal 7, 11, 21, 35, 43, 78, 91, see also BERLIN, MILAN, NEVERS, ROUEN, S. KENSINGTON J Jairus, house of 95; RAISING OF DAUGHTER OF, 112 note, 249, 251, 253 James, Gospel of 189 jars, in MIRACLE OF CANA 85-87, 89-92, 133, 163, 166, 176, 180, 212, 242, 251 Jericho 95 JEROME, St. 189, 193, 194 Jerusalem 46, 54, 124, 190; ENTRY INTO, see ENTRY JERUSALEM church of the ascension 5, 178; martyrion 178 Jesus, see Christ Jews no, 216; Adoration of 13, 19, 161, 162 John the Baptist, in BAPTISM 72-80, 82, 162, 214, 215; COMING OF PHARISEES AND SADDUCEES TO, see PHARISEES ; costume of, see Costume; nimbed, see Nimbus JOHN CHRYSOSTOM 58 John, St. 129, 141, 186; on Milan book covers 208; GOSPEL of 7, 85, 95, 96, 101, 102, 106, 109, no, 112, 120, 125, 130, 204 note JOHN VII, chapel of see ROME churches JONAS 194 Jordan 72-77, 162, 214, 215, 222, 228, 242; personification of 72, 74, 77-82, 84, 176, 180, 228 Joseph ASSURED BY AN ANGEL 173, 202, 222, 225, 229; in ADORATION OF THE MAGI 38, 39, 45, 47, 49! in CHRIST AMONG THE DOCTORS 69, 70; in JOURNEY TO BETHLEHEM 174; in NATIVITY or ADORATION OF THE OX AND ASS 17-19, 21-25, 2 7, 2 9, 3 2 > 166, 174, 184, 190, 191, 226; RECONCILIA- TION OF, WITH MARY 203 Joshua Rotulus, see ROME Vatican JOURNEY TO BETHLEHEM 23, 174 Juda, city of 224 Judas, in LAST SUPPER 129, 140, 185, 186 Judea 196 JUDGMENT OF SOLOMON, see SOLOMON JUSTIN MARTYR 58 Justinian, emperor 141, 182, 188 K KEHRER 42 KERTCH censer from, see ODESSA INDEX 265 key, held by St. Peter, Rouen diptych 233 KIEFF cathedral: mosaics 140 kings, MAGI as 58 knees, Christ immersed to, in BAPTISM 72, 73, 76, 162 KOXDAKOFF 217, 220 L LABARTE 4, 188, 218, 220 Lamb of God, see God Lampadiorum diptych, see BRESCIA lamps, Early Christian, from Egypt, 150, 171 lance 152, 188 note LAST SUPPER 129 , 164. 185. 208, 210, 212 LASTEYRIE 217 later an, see ROME LAURENT 157, 233, 234 Lazarus, CHRIST DINING WITH 137; with head bare 109, 113-115, 118, 119, 164, 167, 181 ; with head covered 108-m, 117, 120, 163, 185, 211; nimbed, see Nimbus; sisters of, see Sisters LAZARUS, RAISING OF 43, 108, 163, 167, 177, 180, 185, 208, 210, 21 1 leaf, heart-shaped, ornamental motif 198, 199 LE BLANT 65, 236 LECLERCQ 9, 24, 81, 223 Lerins 196 Levi, master of Jesus 69 Libri canonum 195 Liguge 195 Limoges, Missal of 92 LINAS de 233, 236 lion, in CHRIST TRIUMPHANT 146, 148, 150, 153. 154, 156, 158, 248 LIPSIUS 189 Liturgical Type of LAST SUPPER 129, 139, 185 liturgies, Byzantine 59; Gallican 195; Egyptian 219; influence of, on Byzantine Epiphany 58, 183, 184; Nestorian 58; Roman 196; Syrian 10, 34, 53, 55, 177, 178, 212, 213, lizard 148 loaves, baskets of 130-132 loaves and fishes 130, 131, 135-138, 212 LOAVES AND FISHES, MIRACLE OF 85, 130, 133-135, 139, 141, 163, 164 loggia 156 Logos 40 Lombard relief 51 LONDON British museum: ivories: Alexandrian-C optic panel 79, 80, 176; Archangel panel 234: diptych 56; St. Menas pyxis 50, 170; panel 71 note, 83; plaques (4) with Passion scenes 187 note; 205, 240; panels (3) with scenes from lives of Moses, Paul and Peter 240 Anglo-Saxon drinking cup 92; Coptic seal 127; textile 50, 175; Ms. Harl. 2788 ( Caro - lingian gospels ) 158 LORSCH, gospels from, see ROME Vatican louvre, see PARIS Lucifer 34 Luke. St., on Milan hook covers 208; GOSPEL of 18, 19, 68, 70, 73, 74. 102, 106, 125, 225, LYCIA churches: frescoes 140 Ly°n 39, 194, 195, 200; inscriptions of, see Inscriptions 266 INDEX M MADRID Christian lamp 150; sarcophagus 87, 215 MAGDALEN COLLEGE, See OXFORD MAGI, ADORATION OF 13, 16, 18, 19, 21-23, 33, 34, 36 , 162 , 175 , 180 , 184 , 189, 190, 208, 210, 21 1, 215, 216, 222, 226, 227, 229, 230, 238, 241, 249, 251 ; arrangement 40, 51-54, 162, 166, 175, 184; costume, see Costume; names of 41; prostrate 49; two 38, 40, 41 ; four 40, 41 MAGI, APPEARANCE OF THE STAR TO, see APPEARANCE OF THE STAR magician, Christ as 85, 87, 113, 176 MANCHESTER fragment of Murano book cover 25, 56, 175 manger, see NATIVITY , ADORATION ; introduction of, in NATIVITY 15 , 162 Mans, Le 230 mantle, in ENTRY INTO JERUSALEM 123, 126-128, 164, 177, 185; see also Costume Marcella 65 MARDIN, Syrian gospels from, see PARIS Mariolatry 201, 202, 212; see also Feasts, Virgin Mark, St., representation of, on Milan book covers 208; in Rossano Gospels 219; GOS- PEL of 73, 95, 102, 106, 125 Marseilles 67, 138, 146, 187, 192-194, 196, 197, 203, 206, 213-215, 229, 233, 234, 236, 241, 246, 253 MARSEILLES relief 198; sarcophagi 86, 197, 234, 247; abbey of st. victor 66-68, 187, 196, 202, 203, 213-215, 229, 233-236, 241, 246, sarcophagi of 233-237 Martha, sister of Lazarus, see Sisters Martin, St. 195 martyrion, see JERUSALEM martyrs 142 Mary, sister of Lazarus, see Sisters; Virgin, see Virgin Mary; Magdalene 65 and note, 67, 68; life of 68 MAS d’AIRE sarcophagus 78, 215 MASKELL 249 MASSACRE OF THE INNOCENTS 59 , 162 , 176 , 203 , 208, 21 1, 213, 214, 220-222, 225, 2 37, 238, 241-243, 247, 249, 251, 253 MAS-ST. ANTONIN sarcophagus 199 master, of Christ 71 ; see also Levi, Zachyas Matthew, representation of, on Milan book covers 208; GOSPEL of 18, 19, 40, 46, 73, 95, 125, 225; HEBREW GOSPEL OF 189 mattress in JOSEPH ASSURED BY AN ANGEL 225; in NATIVITY 24, 26-29, 32, 59, 165, 174, 179 , 184 Maximianus , Chair of, see RAVENNA Maximin, St. 65, 68 medallions, Coptic 50, 171, 175; terra cotta, see MONZA; see also Encolpium Menas, St. pyxis , see LONDON British museum menologium, see ROME Vatican Merovingian art 146, 152, 217, 218, 220, 253; costume, see Costume; gem 230 Metz 194; Carolingian ivories from 249, 254 MIANNAY, wooden pail from 154 MICHEL 189 micheli Collection, see PARIS midwives, in NATIVITY 30-32, 59, 183, 184, 190; see also Salome, Zelemi Milan 3, 188, 229, 230, 238 MILAN cathedral: Ivory book covers 7, 11, 16, 18, 21, 22, 35, 43, 44, 63, 64, 67, 69-71, 76, 77, 91, 95, 100, 101, 104, 1 12, 124, 130, 136-146, 187-189, 201, 203, 206 , 222-230, 236, 238, 240-247, 251 ; date 220, 221 ; technique 231, 243 INDEX 267 s. ambrogio: Paliotto 93. s. celso: sarcophagus 34. s. nazzaros 57, 87 and note. sarcophagi 14, 21, 188 trivulzio Collection : ivory plaque 240 miracle, see CANA, BLIND, JAIRUS, LAZARUS, LOAVES AND FISHES, PAR- ALYTIC, TABITHA, WIDOW'S SON missal, see HILDESHEIM, LIMOGES MOLINIER 137, 144, 189, 217, 233, 234, 236 monks, eastern, in Gaul 153, 157, 195, 196 monogram 236, 237 monophysite heresy 28, 168 MONREALE cathedral: bronze doors 36; mosaics 94 MONTE RAPARO s. angelo: fresco 140 Monza 3 MONZA basilica : ampullae 5, 29, 30, 34, 53, 83, 173, 178, 180; terra cotta medallion 12 mosaics, see BETHLEHEM, CONSTANTINOPLE, JERUSALEM, KIEFF, MON- REALE, NAPLES, PARENZO, RAVENNA, ROME, SERRES, VENICE MOSCOW uwaroff Collection: ivory plaque 172, 173, 204 note MOSES, STRIKING WATER FROM MOUNTAIN 12 note, 228; scenes from life of 240 mothers, in MASSACRE OF THE INNOCENTS 60-66, 162 mountain 77; see also MOSES MU’ALLAKA el, church : wooden lintel 126, 127, 170, 177 MULTIPLICATION, see LOAVES AND FISHES multitude, in MIRACLE OF LOAVES AND FISHES 131-133; see also Spectators MUNICH staatsbibliothek : Codex purpureus 64, 67, 68, 214, 220; Gospels of Henry II 37; ivory plaque (fragment) 64 note, 76, 248 myrrh, see Gifts N NAPLES baptistery: mosaics 90, 143, 204 note, catacomb of s. gennaro: fresco 83 Narbonne 144, 194, 198, 199, 206, 218 NARBONNE sarcophagi 199, 204. genesius church: CRUCIFIXION 205 NATIVITY 13, 33, 45, 54, 55, 58, 59, 161, 166, 174, 178, 179, 183, 189, 190, 203, 208, 210, 211, 215, 249, 251, 253; see also OX AND ASS NAXOS, marble relief 23, 166 Neoplatonism 136, 160, 168 NEROUTSOS-BEY 148, 149 Nestorianism 168 NEVERS museum: ivory fragment 17, 18, 21, 22, 44, 187, 191, 215, 237; date 241-243, 247 ; technique 238, 240, 244, 247 nimbus, worn by Christ 14, in CHRIST TRIUMPHANT 148, 150, 156, in ENTRY INTO JERUSALEM 128, in LAST SUPPER 135, in RAISING OF LAZ- ARUS 112 note, 117, 120, 185; worn by Christ, John, or angels in BAPTISM 72, 77, 78, 80; God the Father in CHRIST TRIUMPHANT 148; Lamb of God, on Milan book covers 206; Lazarus 119, 121 ; Paul and Peter on Rouen diptych 233; Virgin or Child in ADORATION OF THE MAGI 43, 44, 52, 53, 57: Virgin and angel in ANNUNCIATION 1 73 Nimes 197 NTMES, sarcophagus 205 NUMIDIA, sarcophagus from, see ALGIERS O ODESSA, censer from KERTCH 183 Odo 68 268 INDEX Olives, Mount of, church of the ascension, see JERUSALEM open, NATIVITY in 13, 22, 23, 29, 166, 179, 184, 211 orant 131 ; Virgin as 10 Oriental-Hellenistic School 6, 159, 164, 181 Type of ADORATION OF THE MAGI, 37, 47, 48, 166 of BAPTISM 165 of COMMUNION OF THE APOSTLES 165 of CRUCIFIXION 187 note, 205 of HEALING OF THE BLIND 94, 97, 165, 166 of HEALING OF THE PARALYTIC 102, 104, 165, 167 of MASSACRE OF THE INNOCENTS 165 of MIRACLE OF CANA 165, 166 of NATIVITY 13, 22, 165, 166 of RAISING OF LAZARUS 108, 109,114, 167 of SAMARITAN WOMAN AT THE WELL 165, 204 note ORIGEN 168 ORLEANS, fragment of vase from 152 ornament, Syrian, in Gaul 148, 199, 206; on Berlin ivory 2 37; Bodleian book cover 254; Milan book covers 206, 208, 216, 218, 230, 240, 246; Nevers ivory 238; sarcophagi of Provence 208, 221, 233-235; Roman ivories 238, 240, 246; Rouen diptych 231, 233, 235, 236, 246 Orpheus 160 ottoman museum, see CONSTANTINOPLE Ottonian art 28, 69, 101, 108, 121, 124, 154, 248, 251; ivories 48; manuscripts 36, 57, 108, 128, 252 OX AND ASS, ADORATION OF 13, 14, 22, 46, 59, 162, 166, 174, 179, 184, 190, 211, 213, 215, 216, 222, 226, 229, 238, 241, 243, 251, 253 OXFORD bodleian library: ivory book cover 17, 18, 21, 22, 44, 64, 67, 76, 78, 92, 105, 106, 146, 154, 155, 157, 158, 191, 214, 215, 248; date 249, 251; style 254 magdalen college : Ms. Life of the Magdalene 68 P palace, portal of 152 PALERMO ivory 93 palatine, lamp from, see ROME Palestine 5, 22, 29, 43, 48, 53 - 55 . 81, 88, 151, 157, 170, 175 . 177-iSo, 183, 186, 241 Palestinian-Coptic School 6, 159, 170, 177, 181, 186 Type of ADORATION OF THE MAGI 3 7 , 47 , 5L U 5 , 180 of ANNUNCIATION 173 of BAPTISM 72, 80, 82, 176, 180 of Christ, see Christ of ENTRY INTO JERUSALEM 122, 127, 181 of HEALING OF THE PARALYTIC 103, 106, 180 of MIRACLE OF CANA 85, 90, 92, 180, 184 of NATIVITY 13, 22, 29, 179 of RAISING OF LAZARUS 180 of SAMARITAN WOMAN AT THE WELL 204 note pallium, see Costume palmette, ornamental motif 198 PANAGIA, Virgin as 54 panel, see PETROGRAD golenisheff Collection ; ROME sancta sanctorum ; SALERNO paralytic 102-106, 163, 165, 180 PARALYTIC, HEALING OF 102, 163, 167, 180, 185, 208, 210, 211, 249, 253 INDEX 269 PARENZO cathedral: mosaic 143 Paris 194 PARIS bibliotheque nationale : manuscripts: syr. 33, gospels from MARDIN 5, 173, gr. 74, gospels 11, 69, 84, 94, 108; gr. 510, Homilies of Gregory of Nazianzus 51, 69, 99, 100, 101, 108, 128, 184, 185; gr. 1208, Homilies of Jacobus 11; frag- ment of Matthew from SINOPE 99, 135, 183, 184; lat. 8850, Gospels of Soissons 93. 15 7, 251, 252; lat. 9428, Sacramentary of Drogo 28, 44; nouv. acq. lat. 1203, Gospels of Godcscalc 157, 158, 251 ivories: Alexandrian-C optic book covers 105, 118, 126, 170, 172-174, 177, 204 note; Carolingian book covers 65, 67, 68, 214, 252, 254 cluny, musee de : Coptic pyxis 98, 106, 119, 204 note louvre: silver casket from BRIVIO 112 note; Carolingian ivories 44, 202, 252; Coptic pyxis 62, 204 note, 213 micheli Collection: ivories 98, 105, 118 Passion, scenes of 240 patera 82 Paul, St. 141, 142, 144, 233-235, 240 pedum 72, 77, 78, 82, 214, 215 PERPETUUS 145, 216, 218, 252 Persia, Persians 183, 217 personifications 80, 228; see also Jordan PERUGIA sarcophagus 69 PESARO Coptic pyxis 98 Peter, St. 131, 141, 142, 144. 233-235, 240 PETER OF RAVENNA 20 PETROGRAD basilewsky Collection: pyxis 106, 119, 204 note, golenisheff Collec- tion: panel 30, 83, 180; World Chronicle 6, 49 note PHARISEES AND SADDUCEES, COMING OF, to John the Baptist 222, 227 Phoenicians 192 PHRYGIA, relief from, see CONSTANTINOPLE PISA cathedral: bronze doors 36 pitcher, in ANNUNCIATION AT THE SPRING n, 13 PODGORITZA, glass dish from hi podium, in ADORATION OF THE MAGI 43; in RAISING OF LAZARUS, 108, in, 211 Poitiers 194 POUSSAY, Gospels of 155 PRESENTATION IN THE TEMPLE 9, 223; OF THE CROWNS, see CROWNS PRETEXTATUS, CATACOMB of, See ROME priest, in VIRGIN IN THE TEMPLE 8, 223 Priscilla, catacomb of, see ROME Probianus, diptych of, see BERLIN prophet, prophets, in ADORATION OF THE MAGI 49; Isaiah, see Isaiah Protasius, St. 142, 230 PROTEVAN GELIUM OF JAMES 9, 13, 22, 26, 27, 31, 47 Proto-byzantine School 5, 6, 99, 107, 140, 181 Type of BAPTISM 84 of ENTRY INTO JERUSALEM 128 of HEALING OF THE BLIND 184 of LAST SUPPER 135, 140, 185 of NATIVITY 31, 184 of RAISING OF LAZARUS 109, 120, 185 Provencal School 1, 6, 159, 167, 186, 187, 201, 246 270 INDEX Type of ADORATION OF THE MAGI 38, 43, 211, 215, 216, 222, 226, 229, 238, 241, 243, 251 of ADORATION OF THE OX AND ASS 16, 21, 174, 211, 213, 215, 216, 222, 226, 229, 238, 241, 243, 251, 253 of ANNUNCIATION AT THE SPRING 11, 208, 210, 211, 213, 220, 222, 223 of APPEARANCE OF THE STAR 33, 43, 57, 203, 211, 213, 215, 216, 221, 222, 225, 226, 229 of BAPTISM 72, 76, 211, 214, 222, 227, 229, 237, 242, 251, 253 of CHRIST AMONG THE DOCTORS 70, 211, 213 of ENTRY INTO JERUSALEM 124, 212, 221 of JOSEPH ASSURED BY AN ANGEL 202, 222 of LAST SUPPER 136, 208, 210, 212 of MASSACRE OF TIL EINNOCENTS 60, 62, 203, 208, 21 1, 213, 220, 221, 237, 238, 241-243, 247, 249, 251, 253 of MIRACLE OF CANA 90, 212, 237, 242, 251, 253 Types of SAMARITAN WOMAN AT THE WELL 204 Types of VIRGIN IN THE TEMPLE 7, 201, 211, 213, 215, 219, 221-223, 229 Provence 6, 33-35, 65, 6 7, 68, 91, 100, 104, 138, 146, 174, 187, 189, 191-193, 196-198, 203, 205, 208, 210, 21 1, 213-216, 218, 221-229, 235, 236, 241-244, 247, 248, 253; Oriental iconography in 201; see also Gaul PROVENCE, sarcophagi of 35, 36, 66, 78, 112, 116, 189, 199, 204, 208, 215, 220, 221, 226-227, 235, 247; see also AIX, ARLES, AUCH, CLERMONT, MARSEIL- LES, MAS d’AIRE, MAS St. ANTONIN, NARBONNE, NIMES, PUY, St. GILLES, St. RUF, SOISSONS, TOULOUSE, VIENNE PRUDENTIUS 20, 33, 40 PRUM, Gradual of 93 Psalms 148, 156 PSEUDO-MATTHEW, GOSPEL of 9, 10, 13, 18, 26, 27, 35, 45, 69, 70, 189, 210, 211, 213, 221, 222, 225, 226, 241, 247 pupils, in CHRIST AMONG THE DOCTORS 70 putti 160 PUY, Le sarcophagus 1 73, 202, 225, 229 pyxis, see BERLIN, BOLOGNA, BONN, DARMSTADT, ESTELLA, FLORENCE, PARIS louvre, PESARO, PETROGRAD, ROME Vatican, ROUEN, VIENNA, WERDEN, YOULGRAVE R RABANUS MAURUS 68 Rabula, Gospels of, see FLORENCE RAISING, see JAIRUS, LAZARUS, TABITHA, WIDOW’S SON Ravenna 3, 4, 6, 48, 100, 136, 138, 142, 144, 152-154, 196, 212, 249 Type of BAPTISM 82 RAVENNA ivories: Chair of Maximianus 3, 24, 26, 29, 48-50, 62, 79, 81, 89, 98, 105, 118, 126, 127, 133, 170, 171-177, 186, 202, 204 note, 225, 235; Murano book cover 98, 106, 119, 171, 176, 186; see also ROME stroganoff Collection, and MANCHESTER s. apollinare in classe: mosaic 219; s. apollinare nuovo: mosaics 56, 90, 99, 100, 105, 107, 134, 136, 137, 145, 152, 204 note, 219; s. giovanni in fonte: mosaic 82, 228; stucco relief 153; s. maria in cosmf.din (Arian Baptistery) : mosaic 77, 82, 215, 228; s. vitale: mosaic 143, 144 sarcophaagi 6, 47, 115, 116, 167, 197; of Isaac the Exarch 43; Pignatta 153 rays of light, from Hand of God 72, 78; six, on star, see Star RECONCILIATION OF JOSEPH AND MARY, see JOSEPH INDEX 271 REGGIO museum : encolpium 22, 23, 166 reliquaries, Carolingian, see ROME sancta sanctorum ; Oriental, in Gaul 199, 206 Repast, see Eucharistic Resurrection 96 REVEILLOUT 26, 27, 149 Rheims 218 Rhenish ivory, see FLORENCE bargello RICHTER 37, 61 Rivers, Four, of Life 208 “Roman” glass, see Glass Roman School, of ivory carvers 91, 238, 240, 244, 246-248, 253 Rome 6, 39, 45, 91, 104, 123, 142-144, 150, 161, 167, 182, 196, 200-203, 206, 21 1, 212, 216, 218, 225, 238, 240, 241, 244, 246, 247 ROME catacombs: frescoes 6, 40, 60, 73, 103, no, hi, 115, 129-131, 159, 161-163, 165, 167, 200, 211, 213; in callixtus 41, no, 130, 131 ; domitilla 41, 96, 97; Greek chapel 40; pretextatus no; priscilla 103, no, 130, 143; ss. pietro e marcel- lino 41, 87, 133; ST. PONTIANUS 83; S. SEBASTIANO 14, 19; S. VALENTINO 28 churches: chapel of john vii (Old St. Peter’s), mosaic 28, 51, 127, 137; s. agata in subura, mosaic 143; s. costanza, mosaic 143; sts. felix and adauctis, fresco 143; st. john lateran, mosaic 219; s. lorenzo, mosaic 143-144; s. maria antiqua, fresco 202, 219; s. maria in cosmedin, mosaic 51; s. maria maggiore, mosaics 4, 37, 60, 63, 162, 213, 241 ; s. saba, frescoes 107, 185 ; s. sabina, wooden doors 42, 229, 230; S. TEODORO, mosaic 143, 144; S. URBANO ALLA CAFFARRELLA, frescoes 36, 51 collections: barberini, diptych 50, 147, 170; bianchini, silver vase 89; bruls, lamp 130; stroganoff, fragment of Murano hook cover 172-174 lateran museum: fresco in palatine, lamp from 150, 151 sancta sanctorum: Carolingian coffer 93; Carolingian reliquary 252; Coptic ivory 98, textile 30, 180; panel 83; silver casket 44 sarcophagi 33, 42, 60, 86, 87, 91, 100, 111-113, 123, 124, 131, 159, 161, 167, 197, 211, 213, 215, 218, 226, 234, 235, 242, 244 Vatican: manuscripts: gr. 1613, gospels 51; Joshua Rotulus 126; Menologium of Basil II 58, 84 ivories: Coptic pyxis 98, 106, 119; covers of Lorsch Gospels 154, 158, 230, 251 villa borghese : relief 38 ; doria panfili, sarcophagus 39 rosettes, ornamental motif 198 ROSSANO, Gospels of 5, 99-101, 107, 120, 121, 128, 135, 140, 141, 165, 183-185, 219 Rotulus, of Joshua, see ROME Vatican ROUEN library: ivory diptych 187, 231; date 236; techniue 236, 237, 247; ivory pyxis 1 7, 47 S sagum, Gallic 153 saints 141 s. agata in subura, see ROME churches s. ambrogio, see MILAN S. ANGELO IN FORMIS frescoes 69, 101, 128, 140, 145, 146 s. apollinare in classe, nuovo, see RAVENNA s. costanza, see ROME churches STS. FELIX AND ADAUCTUS, See ROME CHURCHES s. gennaro, catacomb of, see NAPLES ST. GILLES sarcophagus 35, 43, 216 st. john lateran, see ROME churches s. lorenzo, see ROME churches INDEX s. marco, see VENICE s. maria antiqua, m aggiore, see ROME churches; in cosmedin, see RAVENNA and ROME churches ST. MAXIMIN abbey church: 202, 204; carved slab 10, 201, 215, 223; sarcophagi 65-68, 203, 214, 218-220, 241, 247 s. nazzaro, see MILAN s. pedro de la rua, see ESTELLA ss. pietro e marcellino, catacomb of, see ROME st. pontianus, catacomb of, see ROME ■ST. RUF abbey : sarcophagus 204 s. saba, see ROME churches s. sabina, see ROME churches S. SEBASTIANO, CATACOMB of, See ROME s. teodoro, see ROME churches s. urbano alla caffarella, see ROME churches S. VALENTINO, CATACOMB of, See ROME s. victor, abbey of, see MARSEILLES Sakkarah 169 SALERNO ivory panels 94 Salome, midwife 23, 25, 26, 30 note, 174, 175 SALONICA arnbo 36, 55 SALVIANUS 194 SAMARITAN WOMAN AT THE WELL 90, 144, 204 and note SARAGOSSA sarcophagus 86 sarcophagus, sarcophagi, of Holy Innocents 66, 67; with representation of ADORATION OF THE MAGI 40, 41, 227; BAPTISM 73, 74, 214; ENTRY INTO JERUSA- LEM 123 ; HEALING OF THE BLIND 97 ; HEALING OF THE PARALYTIC 102, 103, 105; MIRACLE OF CANA 91, 92; MIRACLE OF LOAVES AND FISHES 131; NATIVITY 17, 21, 26; RAISING OF LAZARUS, hi, 115, 211; see also ALGIERS, BOTTARI, CIVITA CASTELLANA, GERONA, MADRID, MILAN, NUMIDIA, PERUGIA, PROVENCE, RAVENNA, ROME, SARA- GOSSA, Sidamara, SYRACUSE saw, held by Joseph, in NATIVITY 17, 226 SBETTLA fragment of vase from 116, 167 SCHMID 15, 188, 220, 229 Schnoudi 169 SCHONEWOLF 205, 218 Schools, of Early Christian Art 159, see also Alexandrian-Coptic, Asia Minor, Byzan- tine, Coptic, Hellenistic, Palestinian-Coptic, Proto-byzantine, Provengal, Ravenna, Roman, Syrian scorpions 148 seat, see Cathedra SEDULIUS 40 sella curulis 66 Serapenm 169 sergius, st., see GAZA SERMONS, Theophany 81 serpent, serpents 148, 150-153, 227 SERRES. Macedonia, cathedral : mosaic 140 servant, servants, in ANNUNCIATION 252; EUCHARISTIC REPAST 133: MIRA- CLE OF CANA 85, 88-93, 166, 180, 212, 242; RAISING OF LAZARUS 109, 120, 185 shed, in Western NATIVITIES 13, 14, 20, 21, 162, 166, 190, 211 Shepherd, Good 160 INDEX 273 SHEPHERDS, ADORATION OF 13, 14, 16, 18, 19, 21, 23, 32, 34, 52-55, 58, 59, 162, 166, 180, 184 Sibyl, in ADORATION OF THE MAGI 37 Sidamara Type of sarcophagi 5 Sidion 65 SIDONIUS APOLLINARIS 196 Siloam, Pool of 95, 101, 184 Simeon 27 SIXOPE, fragment of Matthew from, see PARIS sisters, of Lazarus, in RAISING OF LAZARUS 109, no, 112, 113, 120, 121, 137, 163, 185, 21 1 Smashing Type, of MASSACRE OF THE INNOCENTS, see Provencal SOISSOXS, Gospels of, see PARIS; sarcophagus 78, 215 soldiers, in MASSACRE OF THE INNOCENTS 60-64, 66, 214, 220, 241 solium 60, 61 SOLOMON , JUDGMENT OF, on ivory in louvre 220 Soter, bishop, baptistery of, see NAPLES SOUTH KENSINGTON victoria and albert museum: Werden casket 8, n, 17, 18, 21, 22, 35, 44, 76, 77, 173, 187, 188, 201-203, 214, 215, 221, 238, 240-242, 246, date 231, technique 231; Carolingian ivory 65, 67, 68, 214; Coptic textile 172; French ivory 155 Spain 39, 194 spectators 91, 101, 121, 123, 125, 128, 144, 184, 185, 212 SPELUNCA THESAURORUM 41, 55 Spirit, Holy 73, 77 SPiTZER Collection : ivory 28 spring, in ANNUNCIATION AT THE SPRING 12, 13 stable, in NATIVITY or ADORATION OF OX AND ASS 19, 21, 190, 241 staff, in HEALING OF THE BUND 94, 97, 163, 166 star, appearing to MAGI 33-35, 216; see APPEARANCE OF THE STAR; appearing to THREE HEBREWS 35! in ADORATION OF THE MAGI 37, 53, 55; in NATIVITY 179; in VIRGIN IN THE TEMPLE 9; six-rayed, ornamental motif 198 STRASSBURG forrer Collection : Coptic textile 50, 175 stroganoff Collection, see ROME STRZYGOWSKI 1, 4, 23, 54, 56, 82, 114, 115, 126, 147, 182, 188, 204, 217, 218, 229, 249, 253 strzygowski Collection : encolpium 23, 47, 80, 166 STUHLFAUTH 4, 12 note, 137, 188, 189, 220, 223, 229, 233, 236, 238, 242, 249, 254 suppedaneum 37, 63, 226 supper, see Eucharistic Repast, LAST SUPPER swine, see CHRIST DRIVING DEMONS INTO SWINE switches, in CHRIST AMONG THE DOCTORS 71 sword, see MASSACRE OF THE INNOCENTS, TOURNAY symbolism 61, 86, 96, 103, 123, 160, 164, 168, 200, 212, 215, 227; eucharistic 130, 131, 133-138, 140, 141, 212 symbols, of Evangelists, see Evangelists Symmachi, diptych of 238, 246 Synod of 836 A.D. 54 SYRACUSE sarcophagus 12 note, 228 Syria 1, 33-36, 39, 54, 57, 67, 81, 92, 104, 115, 116, 122, 127, 177-184, 192-194, 198, 199, 201, 203, 212, 213, 224. 229, 230, 249; manuscripts or miniatures from, see ETSCHMI- ADZIX, FLORENCE, PARIS; influence of 55, 93, 99, 104, 144, 169, 179, 196- 198, 205, 206, 208, 210, 212, 234; liturgy of, see Liturgies INDEX 274 Syrian artists, in Gaul 200; calendar, in Gaul 194; commerce, with Gaul 193-195, 199; ornament, in Gaul 198, 199, 206 School 6, 159, 181 Type of ANNUNCIATION 173 of BAPTISM 72, 79 of HEALING OF THE BLIND 94, 98 , 184 of MIRACLE OF CANA 92, 184 of SAMARITAN WOMAN AT THE WELL 204 note Syro-Anatolian Type of RAISING OF LAZARUS 1 15, 116 Syro-Egyptian influence 146, 248, 253; Type of MASSACRE OF THE INNOCENTS 60, 61 T TABITHA, RAISING OF 112 note table, in LAST SUPPER 129, 135-138; in WIDOW’S MITE 144, 145 temple, in CHRIST AMONG THE DOCTORS 69; in VIRGIN IN THE TEMPLE 8, 9 TERTULLIAN 103 TESTING OF THE VIRGIN, see VIRGIN textiles, Coptic, see AKMIM, ANTINOE, LONDON, ROME sancta sanctorum, SOUTH KENSINGTON, STRASSBURG Theophany, see SERMONS Theotokos, Virgin as, See Virgin THREE HEBREWS BEFORE BUST OF NEBUCHADNEZZAR 35 throne 37, 69, 70; of Christ, see Christ enthroned TOMB, HOLY WOMEN AT, see HOLY WOMEN tomb, in RAISING OF LAZARUS 109, no, 113, 115-120, 181, 185; see also Aedicula, Arch TONGRES notre-dame: leaf of diptych 235 and note tonsure 142 TOULOUSE sarcophagi 86, 197 TOURNAY Tomb of Childeric, sword from 217 Tours 194, 195 TOURS st. martin 145, 216, 252 tower, towers 224, 229, 230, 252 TRADITIO LEGIS 142-144 Transitional Type of LAST SUPPER 129, 134 , 212; of MIRACLE OF CANA, see Alexandrian-Coptic tree, in ANNUNCIATION AT THE SPRING 12; in ENTRY INTO JERUSALEM 1 23. 124, 128, 164, 185 tribune, in MASSACRE OF THE INNOCENTS 61-65 TRIER stadtbibliothek : Ada Gospels 157, 158, 251 ; Codex Egherti 57, 69, 93, 124, 125 TRINITY 28, 79 tripod 131, 133 trivulzio Collection, see MILAN ; group of ivories 240 TROMBELLI 49, 62 tunic, see Costume TWELVE APOSTLES, GOSPEL of 26, 27 Tyche, in ENTRY INTO JERUSALEM 126 Types, of scenes in Early Christian Art, see Alexandrian-Coptic, Asia Minor, Byzantine, Coptic, Historical, Liturgical, Oriental-Hellenistic, Palestinian, Palestinian-Coptic, Proto-byzantine, Syrian, Syro-Anatolian, Syro-Egyptian, Transitional U urn, in BAPTISM 228 uwaroff Collection, see MOSCOW INDEX 275 V vase 38; ornamental motif 198, 199 Vatican, see ROME vatopedi, see ATHOS veil 172, 178 VENICE s. marco: ciborium columns 26, 57, 93, 97, 105, 121, 127, 137, 138, 204 note; mosaics 101 VENTURI 4, 137, 142, 188 victoria and albert museum, see SOUTH KENSINGTON VIENNA pyxis 25, 56 Vienne 194, 195 VIENNE altar 138; sarcophagus 247 vine, see Grapevine viper, see Serpent Virgin Mary 88, 149, 178, 191, 201, 202, 210, 221, 223, 225, 231; in ANNUNCIATION n-13, 172; CHRIST AMONG THE DOCTORS 69, 70; DEPARTURE FROM HOUSE OF ELIZABETH 224, 225; JOURNEY TO BETHLEHEM 174; MIRACLE OF CANA 85, 88, 92, 93, 132, 180; NATIVITY 15, 17-19, 21-24, 26-28, 32, 34, 162, 166. 174, 175, 179, 184, 190; VIRGIN IN THE TEMPLE 8, 219; VISI- TATION 224; costume of, see Costume; nimbed, see Nimbus; posture of, in ADORATION OF THE MAGI 37-39, 40, 42-44, 47,-49. 51-57, 59, 175, 180, 226, 241, in ANNUNCIATION 11-13, 172, 173, 252; as Theotokos 52 VIRGIN, DEPARTURE OF, FROM HOUSE OF ELIZABETH 222, 224; RECONCIL- IATION OF, WITH JOSEPH, see JOSEPH; IN THE TEMPLE 7, 201, 208, 210-213, 215, 219, 221-223, 229; TESTING OF, WITH WATER 173, 174 VISfi Reliquary of St. Hadelin 155 VISITATION 222, 224, 229 voice, in Annunciation 13 W wafer, in COMMUNION OF THE APOSTLES 139 waist, Christ immersed to, in BAPTISM 72, 74, 78, 79, 82, 215 wand 85, 86, 91, 108, 109, hi, 113, 117, 118, 133, 163, 173, 175, 176, 181, 211, 212, 242 WASHINGTON Freer Gospels 6 WATER, TEST BY, see VIRGIN, TESTING OF waterfall, Jordan represented as, in BAPTISM 72, 74, 77, 78, 162, 214, 215, 228 WEDDING AT CANA, see CANA wedding feast, in MIRACLE OF CAR 7 A 85, 88, 93, 184, 212 well, in SAMARITAN WOMAN AT THE WELL 204 and note WERDEN, ivory casket from, see SOUTH KENSINGTON ; pyxis 25 WESTWOOD 137, 188, 220, 223, 229, 249 widow, in WIDOW’S MITE 144, 145 WIDOW’S MITE 144. 208, 21 1, 216 WIDOW’S SON, RAISING OF hi note William, Count 68 WILPERT 41, 103, no, 133 wine, in LAST SUPPER 129, 130, 134, 139, 212 WOMAN, SAMARITAN, AT THE WELL, see SAMARITAN; TAKEN IN ADUL- TERY 144; WITH ISSUE OF BLOOD, HEALING OF 249, 251, 253 WOMEN, HOLY, see HOLY WOMEN AT THE TOMB World Chronicle, see PETROGRAD golenisheff Collection Worms 194 wreath 38, 40; see also Crown 276 INDEX Y YOULGRAVE pyxis 106, 204 note Z Zacharias 223 Zachyas 70 ZAGBA, see FLORENCE Rabnla Gospels ZARA Lombard relief 48 Zelemi, midwife 27, 30 note Zion, Mt., on Milan book covers 208 TABLES NOTE In the Tables, a dash indicates the occurrence of the motif named at the head of the column in which the dash occurs. The addition of a circle to the dash indicates a nimbus ; if the circle is crossed vertically, the nimbus is cruciform. TABLE I. 0A1ENTAL-HELLEHIST1C AlZXANDM AN- COPTIC PALESTINIAN- COPTIC *.T / ^ vSj> ’ nr r - or. .i\+u * *Y?|V£? AM . . ' :rr: «-’j &*fJ fg’. ~a«r~n^r.r t. • ■'• 1 ', ./ .vfTVv^t .' ttsr : pqt^ w »^ r «r r 7 v^T r ././•■'• C ,*• -’•..•*• I i ,y *. Si -k ■ V' .*«•.' < ,,*• •• ■' . ■•■ •'/* **i . ■..■■•• •'., *. .V , U • , * , «y£A 4 jV I J '■ , * 'V* • •' * • •>. V ^ ot i tA . . vJ ,xt aioor-A ' tur? tcib'A ? •; .osl \2 -’V-' .IV ,n ,* »-{«-!■< ii-.i:'! .2' ;•> • i ,: : :a . . , . i ■ .• /v * . • .,i.i J *»:.■■ . " .. • •' r. .-.1 •)'. \ v;c i • •;• : . ?i>e - u ?Wi i ■ < <■ ^ Z'ieWTa?: .- j , V. 7 ■ . SJfteV* . .1 p ' } u . . •!. n a* ' a iiiM J tX ■<< " ■ , r*«rt tf uJtf . ??•> • j •, »: ' . . ■ .* 3 - ,.a v-:. } •. ;* n ■-:•••. •-> : ixs.-rx; •» .. ; yv.*.', io, ^ i. i.'v'. t > . •.• .... ■'V. r .. 2 * .) ; • v-.-f.w, r.1 J "jrt^CD i? i w i mA i W5» I * ■ » i Q« •. io . vC* . V 1. •; .. . {.y ■ t b jt ^ TZ « :xx\ s' *« ■J. j «5t.y* ; r-; , c a . r.;. v „ B V« . * ,v i - ! •; : ; : .■ '.-id oiv u , '«:• U' rt . . O'”' 1 " ■ . ! . ■ r . ' i i? ,'r:- ■ •ce-j -vi'l ' ' •- 'i.b' . ’ ' •■*'.- v-’. .-1 ;V»? I J*: I) '■ \ '*! •’ . .TV ;• *•: . V . ■ ■ • ■ ;* j | {.* '.M 3 *- ; . - ..V ic- ‘ ’ .• • • ■ •• ■ 4r. . 0.‘< . T.t, : 1 » i L ajiw- vtwet. '-v fj L B'tMOS-W'W’fr > ! !W t ■ . •. : ' • 7 . 7(7 : f fl HELUHI6TIC TTFS KXLLENI ETI C-0R1 DJTAL palestiniah-coptic SYZARTIKJ Errata: No. S3. In citation of Laucr, for "XIV,” read “XV.” No. 57. For "Morgan Collection," read “Strzygowski Collection"; add, as reference, Slrzygowski, Oriens Christ. V, 1915, p. 100. I The lies signifies three Kagi unleie othsrsiss designated. Signifies a niabus. ^ Signifies o cruel forre niabus. • ■■■'■ i >** ■ ■ ’ • • • • ' ■ ■ ■ TABLE III. TABLE IV. Errata: No. 26, in citation of Dalton, for "298,” read "294." No. 28, for "Capital,” read "Carved drum of column." Nos. 29-30. for "Fresco,” read "Mosaic"; in location, add "Ravenna." No. 32 belongs in Alexandrian-Coptic Group after No. 26; for "Morgan Collection, read "Strzygowski Collection”; add, probably bearded. Add to Group 6: Berlin, Kaiscr-Fricdrich Museum; ampulla and bronze relief (Pergamon). Sec p. 83, note 23. reference, Strzygowski, Orient Christ. V, 1915, p. 96; Christ i T . i' vs . , , V, . :•» Ju>0‘ /-i i " ' . : i 4 ’ ' . Vi. • : . • , f '* l site** :• cAli < . Ml A > ■ ™' . ' ;'V '.if. ■’ SVis : ■' , ■ :■ . :: > '■ : !: ... -• ’ *' . ; ».l t : . »s*. i •*‘ 1 »’ : /T • J :i aG. ’ ■! >!'■ . • ,• .••>!. ..’j'x’H'A . ss. , si' i'jVT .• 1 . ! p • < . '. t’l t ML - 1 .vs ,11 ■ ■ ■■■■■■ ! b - j 1 K n 'i VITalWABlii haWM ’ . M-l'JUIMXtJi tstc.«r«-nsww TABLE V. if.. v-aji TABLE VI. TABLE VII. KKALIHQ Ot H» PARALYTIC. >SKM*M Sri ■ to ft*.',, -is -■ >.■ ■ , -v'---— --y- <,( ■■ ’ ' • • ' ■ ' . , • ■■ 2*i?j TfifiRT r .- • ' iV ..... -i»( •••'.■ «r. ■ f «S*\ * *br..'.L r.‘ «;",M 7 - > ■ i . : . --.r- y". f . . _ y . ■ .s”. .' 'l .'; y 1 . .. : :■ ■. .. | v | , fWUJ*K . •; r:- . • . .- .-is=r*s* - *• ' _ ; * vc c* _ fd ~ ;wa*.ap-oi : •«••- 5moC>^l*lKltUCi .1 wrrsucts table VIII. HELLENISTIC ORIENTAL -HELLENISTIC ALSXANDU AH-COPTIC PALESTINIAN -COPTIC Errata: No. 71, in citation of Le Biant, for “PI. XXVIII,” read "PI. XXVII." No. 99, (Garr. 171/1) is to be omitted. No. 107, for “Sarcophagus,” read "Relief": add 107a, marble vase-fragment from Sbeitla (Tunisie) (see d. 116) No. 128, in indication of date, for IX, read VII. ' . TABLE IX. HlLLBnSTIC ALEXANDRIAN -COPTIC paledtinian-coptic B7ZANHNS Errata: No. 8, add indication of "city.” No. 23, for "il,” read "el.” No. 24. for “PI. 133,” read “PI. 137.” No. 28, in date, for "VII," read "VIII.” no i i i '* } \f j ( .1 o / n- n » GETTY CENTER LIBRARY 3 3125 00774 8243 .y^ywgfi. ■ S ■ ■P iW*7> £'»• .