THE SOCIETY FOK THE PROMOTION OF HELLENIC STUDIES SUPPLEMENTAKY PAPEE No. 4 EXCAVATIONS A'C PHYLAKOPI IN MELOS CONDUCTED BY THE BRITISH SCHOOL AT ATHENS PUBLISHED BY THE COUNCIL, AND SOLD ON THEIR BEHALF BY ]\^ MCMILLAN AND CO., Limited, ST. MARTIN'S STREET, LONDON, W.C. MDCCCCIV AU rights resei'ved PWCE THIRTY SHILUNGS NET EXCAVATIONS AT PHYLA KO Pi IN :\IELOS THE SOCIETY FOPv THE PHOMOTTON OF HELLENIC STUDIES SUPPLEMENTAHY 1>APER No. 4 EXCAVATIONS AT PHYLAKOPI IX MELOS CONDLXTED BY THE BRITISH SCHOOL AT ATHENS DESCRIBED BY T. D. ATKINSON, R. C. BOSANQUET, C. C. EDOAR, A. J. EVANS, D. a. HOGARTH, D. MACKENZIE, C. SMITH, AND F. B. WELCH II'/VA 41 /'/'(/',.■<, and lO.S III ml rat ion ■■< in the Text l'l-|U.ISHEI) BY THF. COUNCII.. AND SOLD ON THEIR BEHALF BY MACAU LLAX AXi) CO.. Limjtfd. ST. .MARTIN'S STREET. LONDON, W.O. MDCCCCIV ,1// ri'/lil-i /•fA-erct'/ RicHARiJ Clay and fSoxs, Limited. HREAU STREET HILL, E.C., AM) BUNCAY, St'FKOLK. FliEI ACE In accordance with a 23roposal of the Committee of the British School at Athens, the Council of the Society for the Promotion of Hellenic Studies, at a meeting held on Dec. 20, 1900, authorised tlie publication of a Supplemental- v Paper embodying the results of the Excavations at Phylakopi. For this purpose a special Editorial Committee was appointed, consisting of Messrs. R. C. Bosanquet, E. A. Gardner, and G. F. Hill. The process of preparing for publication the masses of pottery and other objects brought from Melos to Athens has of iiecessit\' Ijeen a slow one. It was begun by Mr. C. C. Edgar in 1897, and has been carried on since 1899 ))y Mr. Bosanquet, with occasional help from j\lr. F. B. Welch and other nieml)ers of the British School. The majority of the drawings (excluding those in Chapter II. wliich arc all Ijy Mr. Atkinson) are the work of Mr. Halvor Bacge ; others, including the coloured copy of the Flying-fish picture, have been" made by Mr. E. Gillieron, and a few by Mr. Theodore Fyfe. The difficulty of communication between contril)utors li\ing in three and sometimes four different countries, ma}' be held to excuse certain inconsistencies in nomenclature. No contributor is to be regarded as responsible for any statement not standing over his own signature. vi PREFACE The Editorial Committee desire to express their acknowledg- ments to Mr. P. Cavvadias, Ephor-General of Antiquities, and to Messrs. V. Staes and Ch. Tsountas, Ephors of the National Museum at Athens, who have done everything in their power to facilitate the preparation of this volume ; to Messrs. D. G. Hogarth and Cecil Smith, who, in addition to contributing the chapters signed by them, have in other ways rendered valual)le assistance ; and also to Mr. Arthur Evans, who, though he took no part in the excav- ations, has contributed a note on the Pottery Marks in the light of the Cretan evidence. THE EDITORIAL COMMITTEE. Feh. 1904. TA13LE OF CONTENTS Introduction. By Cecil Smith PAGE 1 Chapter I. — The Excavation. By D. G. Hogarth. § 1. — Preliminary Soundings and the Wall of Fortification §2. — The South- Western Corner § 3. — The North- Western Terrace § 4. — The North Central Terrace § 5. — The North-Eastern Terrace and the Palace §6. — The other Regions y 10 11 U 19 23 of the Three Periods Chapter II. — The Architecture. By T. D. Atkinson. § 1. — Introductory ... § 2. — Relationship of the Building:- § 3.— The Site § 4.— The Town Wall § 5.— The First Period § 6. — The Second Pei"iod § 7.— The Third Period . ExplaHutioH of Plan of Fir>0 11. The Excavated Part of the Site looking towaixls the S. End of the Palace 21 12. Early House in J 2 from above 22 13. Ruined Walls with the Cultivated Plain and the Hills in the distance 26 14. Plan and Section of Buildings in H 2, showing Work of Three Periods and Modern Terrace 28 15. The Town Wall looking West, showing an Offset in the Work of Period II and Bastion of Period III built up against it 30 1 (). Plan of Part of the Town Wall, and Enlarged Plan showing Staircase, etc 31 17. View of Passage through the Town AVall looking South 32 18. View of Staircase 32 19. Suggested Restoration of Doorway in the Town Wall 33 20. Angle of the Town Wall 34 2 1 . Remains of Period I 35 22. ,, ,, „ showing Pavement and Doorway 30 23. House of Period I 37 •2i. „ „ „ 38 25. Block-Plan of Part of the Town of Period II 39 26. Plan of House of Period TI, with Stone Column 40 27. Plan of House of Period II 41 CONTENTS. no. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67, 69, 71. T2, 74. View of Rooms in Fig. 27 looking towards the Sea ,, ,, ,, looking South ... Diagram of Doorway in Fig. 27 Plan of House of Period II View of Room in Fig. 32 Plan of House of Period II View of Staircase in Fig. 34 Section of Street and Drain (Plate I., H 3 : 13) and of D Plan of House of Period II View of Wall in H 1 (Plate I) Diagram of Doorway in H 2 (Plate I) View of Doorway in H 2 Details of Plaster Ceiling Block-Plan of part of the Town of Period III A Street Corner (G 3) Plan and Section of Small Chamber in G 3 : 4 ( = Fig. 7) View of the Chamber in G 3 : 4 Outlet from Chamber in G 3 : 4 Plan of House of Period III Plan of Detached House of Period I CI Plan of the Palace View of Step of Portico and Plinth of Anta Earthenware Cylinder from the AVell Plan of a Doorway with Rebated Jambs . . . Step and Threshold of Doorway in E 3 Conjectural Restoration of Doorway in E 3 Basalt Base of Column (E 3 : A) Stone Base of Column (G 3) Earthenware Cliannel (G 2 : r) General Sections : Lines LM, NO Lines PQRS, TU, WX, YZ Fragments of Painted Plaster : Marine Design ,, ,, : Body and Arms of a Seated Man ,, ,, ,, : Arms and Shoulder of a Hum Figure Fragment of Painted Plaster Design of Lilies (^reconstructed) Fragment of Painted Plaster : Head and Back of a Bird Spiral Design (reconstructed) 68. Prehistoric Vases from Pelos {B.S.A. iii. pp. 41, 44) 70. Prehistoric Vases from Pelos {B.S.A. iii. p. 45) Vase of Type developed from Type 1 73. Pyxis-lids Vase developed from Duck-vase Type 1-A(.E 42 42 43 43 44 45 45 46 46 47 48 48 49 50 51 52 52 53 53 54 55 56 57 58 58 59 60 61 61 61 62 63 72 73 74 75 76 77 78 83 84 85 88 90 CONTENTS. 75. Front view of Animal-vase 76. Amphora with Geometric design and Hollow Foot ,,. 77-7i>. Flat, Two-handled Vases of Early Mycenaean Style <^0, .SI, 82-84. Sliallow Cups 85-87. „ „ 88. Panelled Cup with Floral Design 89. Fragment with Frieze of Animals 90. Iled-faced Jug with Design in White 91. Vase with Design in Black and Red 92. Fragment with Flying Swallow ... 93. Fragment with Bird on Dolphin 94. Amphora with Floral Design in Black and Red 95. Figure from Fishermen Vase 96. Jug with Floral Design in Black and Red 97 98. Top of Small Jug with Floral Design 99, 100. Fragments with Designs in Black and Red 101. Jug of Later Melian Fabric with Floral Design 102. Fragment with Ivy-leaf Pattern 103. Cup with Rosette Pattern in Red ami l>lack 104. 105. Cups with Patterns inside 1 06. Restoration of Cup 107. 'Shell-vase" , 108. Small Askos of Native Melian Fabric 109. ,, ,, Imported Mycenaean Fabric 1 10. Pedestal Vase ill. Pierced Terracotta Vase of shape of Stone Vases 112. Bath, showing Interior Decollation 1 13. Large Basin or Foot-bath 1 It. Fragment from Bath or other large Vessel, with Du 1 I."). Similar fragment, with Duck and Reeds ... 1 16-122. Series of Local Vases (B.S.A. iv. p. 36) 123. Mycenaean Vase of comparatively Eai'ly Styl 124, 125. Late Mycenaean Pottery 126—130. Kamares Pottery 131-133. Designs on Kamares Pottery 134. Jug with Side-spout 135, 1.36. Quadruple Vases 137. Goblet 138. Fragment with Spiral Ribl)ing 139. 140. Vases of dark Clay 141. Pattern on inside of large shallow \'as( 142. Shallow V'essel with Interior Handle Pap 91 98 110 111 112 1 1 3 115 116 117 119 120 121 122 124 126 127 128 128 131 132 133 134 134 135 136 136 138 139 140 140 141 142 114 146 147 149 150 152 153 154 154 155 155 155 CONTENTS. 143. Rim and Shoulder of large Vase 144. Interior of shallow Yase 1 45. Wash-hand Vase ? 146. Vase with turned-over Rim 147. Censer? 148. Fragments of Cypriote Bowl 149. Cretan Wild Goat 150. Impressed Signs — Table of Signs 151. Scratched Signs 152. Geometrical Marks 153. Fish-sign from Dictaean Libation-table 154. Early Babylonian Signs for Fish 155. Pair of Linear Characters from Base of Vessel 156. Sign-groups from Cnossian Tablets 157. Sign-group from Seal-Impression 158. Identical Forms at Phylakopi and Cnossos... 159. Non-Cnossian Group of Signs 160. Terracotta Statuette from Cameiros 161. Stone Mould and Axe-head cast in it 162. Ivory Signet-Ring 163. Ivory Finial 164. Marble Jar 165. Stone Cup 166. 167. Steatite 'Blossom-Bowls' 168. Steatite Bowl with Oblique Fluting 169. ,, „ with Horizontal Grooves 170. „ ,, (hemispherical) 171. Dish of Micaceous Schist 172. Grooved Implement of Slate 173. Pestles and Mortars 174. Plummet of Porous Trachyte 175. Forked Object of Tufa 176. Hindquarters of Terracotta Cow in Native Melian V 177. Foi'epart of Terracotta Cow in Imported M3'cenaean 178,179. Heads of Bulls, completed 1 80. Terracotta Boat 181. Miniature Stool 182. Box-like vessel, with three Openings at Top 1 83. Box with Rounded Feet 184. Steatite Lamp 1 85. Section of Claj- Lamp 186. Standard Lamp 187. Lamp of Tiifa are War PAGE 156 156 157 157 158 158 176 177 179 180 182 182 182 183 18^ 183 184 184 188 191 193 193 195 196 197 198 198 198 198 199 200 201 201 204 204 205 206 206 207 208 209 209 210 211 CONTENTS. 188. Section of Clay Fireholder . . . 189. Clay ' Spit-rest ' 190. Earthenware Table 191. Sketch Map of Melon (B.S'.A. iii 192. Specimens of Worked Obsidian 193. Plans and Sections of Tombs -') 211 212 212 217 219 235 INTKODUCTL(,L\. Thk narrative of the School undertaking in Melos (in 1896) which h-il to the excavation of Phvkkopi has ah-eady appeared in the Annual of the British School ; all that it is necessary to recall here may be restated in a few words. Tile discovery of the marble Poseidon, now in the National 3Iuseum at Atlicns, had led lis to believe that a systematic excavation of Klima (a V-shaped patch of garden land hard by the sea, below the classical town) might be fruitful in results of similar importance. It was, more ver, a condition impo.sed by the Greek Government that that site should be selected for the first excava- tion. In this, the 23oi)U d'appui of the ^Melian campaign, we were disappointed ; the excavations we made seemed to prove that the site was unoccupied, and perhaps even was covered by the sea, until the late Roman i)eriod. We next proceeded up the hill to the site near the village of Trypete, which was already well known as covering the remains of the classical town. The whole of this ground is under cultivation, and lying as it does mostly on the side of a terraced slope, offers in most parts only a meagre depth of soil to the excavator's spade. Moreover, ever since the discovery of the famous Aphrodite, this part of Melos has been the happy hunting ground for casual excavators: a fact which we were soon able to verify. Our efforts thus were necessarily limited ; the season was spent partly in searching for tombs, partly in the oi)euing up of a late house with a fine mosaic, and in the identi- fication of certain topographical features, including the probable agora and stadion. Such time as could be spared from excavation was usually occupied, especially by Mr. Bosanquet and Mr. Mackenzie, in exploring the rest of the island and examining likely sites for excavation. Chief among these was a site close to the village of Phylakopi on the north-east coast, which Ross (Insdreisai iii. 13) had heard of in 1848 but had not visited, and of which Weil and Diimmler in A(h. Mitfk. 1876, p. 246, and 1886, p. 26 give an account. Both Ross and Diimmler call the site 'arov Kc'nrpov : but this is apparently a misconception, the real name being 'arov KuTrrov, because of the white spray that blows over it when the wind is strong from the north.' We have called it Phylakopi after the village which stands not far away.- Some antiquities which we had seen offered for sale in Athens were said to come from this site, and seemed to promise interesting results from an ' Edgixi- in B.S.A. v. p. l.">. to 'arh Karpo as 'eiiie ^'i«.'l■telstuIule wfittT - \^\\\ also fivlk-d it •fi'AaKt.'Trj), Init ichi ifd siilwjuts.' 2 CECIL SMITH examination of the ground. Diinimler had carefully described the site, which consisted of a necropolis of the Mycenaean period, pre-Mycenaean graves, and on the edge of a clifif partly eaten away by the sea a prehistoric settlement within strong walls of cyclopean masonry. The necropolis, as he states, was completely cleared out at the time of the war of indepen- dence ; the prehistoric tombs in the tufa hill south-west of the town had equally been excavated, but some at any rate more recently, as the traces of excavation were still fresh when we arrived. The town itself had naturally not escaped, but, being a bigger job, had not been seriously undertaken by the excavators, who had contented themselves with a hole here and there, which had already been sunk when Diimmler visited the site. The main portion of the town covered the long gradual slope down- wards to the east ; apart from the fragments of pottery which littered the surface, the traces of ancient occupation were not very noticeable in this portion, consisting of stones which might be observed projecting here and there from the soil, but showing occasional traces of regular arrangement. The more jsrominent part lay to the west at the highest point of the slope, where unluckily, owing to the proximity of a soft tufa stratum, the sea had in course of time eroded a considerable slice of cliff and city. Here on the west and south the strong walls of polygonal masonry were visible on their outer face, standing uncovered on this side to a height of some metres above the surface of the soil. Within the walls, a group of buildings remained still undestroyed by the sea, occupying a width varying from 10 metres at the western extremity to about 30 at the eastern end, at which point the edge of the cliff takes a sharp trend in a northerly direction. It was in this block that we first broke ground on May 7, 1896. During the first season, owing to the fact that we were simultaneously engaged in other parts of the island and that I was obliged to return to England at the end of May, it was impossible to do more than partly clear this quarter, and make preparations f