NAUKRATIS. PART II. ERNEST A. GARDNER, M.A., FRTjIiOW of GONVILLE and CAIUS COLLEGE, CRAVEN STUDENT AND FORMERLY WORTS STUDENT OP THE UNIVERSITY OF CAMBRIDGE DIRECTOR OF THE BRITISH SCHOOL OK ARCHAEOLOGY AT ATHENS. WITH AN APPENDIX, BY P. Ll. GRIFFITH, B.A., OF THE BRITISH MUSEUM, FORMERLY STUDENT OF THE EGYPT EXPLORATION FUND. WITH FORTY-FIVE PLATES AND PLANS. . SIXTH MEMOIR OF THE EGYPT EXPLORATION FUND. PUBLISHED BY ORDER OF THE COMMITTEE. LONDON: TRUiJNER & CO. 57 & 59, LUDGATE HILL, E.C. THE UNIVERSITY OF ILLINOIS LIBRARY NAUKEATIS. PART II.- 1885-6. Digitized by the Internet Archive in 2017 with funding from University of Illinois Urbana-Champaign Alternates i https://archive.org/details/naukratis02petr NAUKRATIS. PAET 11. BV EENEST A. GAEDNEE, M.A., FELLOW OF GONVILLE AND CAIDS COLLEGE, CRAVEN STUDENT AND FORMERLY WORTS STUDENT OP THE UNIVERSITY OF CAMBRIDGE DIRECTOR OF THE BRITISH SCHOOL OF ARCHiEOLOGY AT ATHENS. WITH AN APPENDIX, BY F. LE. GRIFFITH, B.A., OK TEE BRITISH MUSEUM, FORMERLY STUDENT OF THE EGYPT EXPLORATION FUND. SIXTH MEMOIR OF THE EGYPT EXPLORATION FUND. PUBLISHED BY ORDER OF THE COMMITTEE. LONDON : TRUBNER & CO., 57 & 59, LUDGATE HILL, E.C, 1888. LONDON PRINTED BY GILBERT AND EI7INGT0N, LIMITED, ST. JOHN*S HOrSE, CLERKENWELL ROAD. TO W. M. FLINDEKS PETEIE. Ovvofia (Tov fj-aKapuTTov, o NavKparis wSc c^vXa^ei ecrr’ av itj NeiXov vavs l^aAos revdyrj. Posidipjnis ap. Ath. 596 D. 220016 PEEFACE. /\yNy\/> /-v'\/>/-vrvVV When a work begun by one hand is continued by another, an explanation or apology seems to be required. I trust however that in the present case some degree of continuity may be found in the reports of the excavations we have conducted at Naukratis for the Egypt Exploration Fund. For I was by Mr. Petrie’s kindness enabled to study the yet unpublished results of his first season’s exploration, and to have my conclusions incorporated in his report ; and the excavations of the subsequent season were begun under his direction and after his system. The present volume also contains some of his handiwork, in the plates signed with his initials and the description of the articles contained therein. But above all, both in the introductory narrative and in the dedication I have endeavoured to show my sense of obligation to one whose discoveries have led to the results recorded in this volume as well as in Naukratis vol. I. Mr. Griffith also was with us for a short time at Naukratis. The results of his special work will be found in the Appendix of this volume, and in the plate which he has drawn (XXIV.). For another plate (XXIII.) we are indebted to the kindness of Mme. Naville. In preparing the plans of the temples, I have met with the greatest assist- ance from Professor Middleton, by whose advice I have given the m theform in which they now are seen. The general plan (PI. IV.) is reduced from one made by Mr. Chapman of the Fitzwilliam Museum, which was based upon Mr. Petrie’s plan, with such modifications and additions as our later excavations had rendered necessary. The photographic plates have been executed under my supervision by Mr. Griggs ; as they are collotyped directly from photographs of the originals, their accuracy may be relied upon. I have in the narrative of our season’s work referred to the help and kind- PREFACE. iv^ ness with which I have met on every side during my stay in Egypt ; it is my pleasing duty to acknowledge especially the consideration I received from Mr. Maspero and others holding official positions in Egypt, and from Dr. R. S. Poole and other officers of the Egypt Exploration Fund in England. To the University of Cambridge I am indebted for the grant from the Worts Fund which enabled me to undertake the excavations in Egypt. I have there- fore to lay before the University the present memoir as a report of the work for which it equipped me. Last, but not least, I would acknowledge my obligations to my brother, Professor Percy Gardner of Oxford. To him I owe alike my first impulse to the study of archaeology and the first suggestion of the particular piece of excavation of which this volume is the record. I have also been indebted to him for advice and assistance both during my work in Egypt and during my preparation of its results for publication. I therefore feel that his name fittingly concludes the list of acknowledgments with which I would preface this book. ERNEST ARTHUR GARDNER. Cambridge, 1888 . CONTENTS. Chap. I. — General narrative of the season’s work, 1885-6. SECT. page 1. Eesults of the previous season . 9 2. Arrangements for continuance of work ...... 9 3. Arrival at Naukratis . . .10 4. Division of labour . . , .10 5. Discovery of the temple of the Dioscuri . . . . .11 6. Discovery of the cemetery . .11 7. Discovery of the temple of Aphro- dite . . . . . .11 8. Discovery of the temenos of Hera . 13 9. Trial Trenches . . . .'13 10. Sphinx near the Hellenion . .13 11. Great walls on N.E. of city . . 14 12. Alterations in superintendence of work ...... 14 13. Richness of the yield from the temenos of Aphrodite . . .14 14. Trial pits at Neqrash . . .15 15. Houses in the town . . .15 16. Independent diggers . . .16 17. System of work . . . .17 18. Manner of life . . . .17 19. Packing and despatch of antiquities 19 20. Their condition and later fate . .19 21. Previous publication, &c . .20 Chap. II. — The Cemetery. 22. Importance of the cemetery ; its probable site . . . .21 23. Its discovery; monuments above ground . . . . .21 SECT, 24. Nature of graves .... PAGE 23 25. Earthen coffins .... 24 26. Stone and wooden coffins 25 27. Contents of graves .... 26 28. Various objects found . 27 29. Results ...... Chap. III. — Temple of the Dioscuri. 30 30. Plan of temple .... 30 31. Its decoration and painting . 31 32. Objects found .... 31 33. Age of the temple .... 31 Chap. IV. — Temple and Temenos of Aphrodite. 34. Explanation of Plans 33 35. Earliest temple and its appurten- ances ...... 33 36. The contents of the earliest temple. and their fate .... 35 37. Second temple .... 36 38. Third temple ..... 37 39. Chronological relations of the three 37 Chap. V. — Pottery from the temenos of Aphrodite. 40. Circumstances of discovery 38 41. Classification of the pottery . 38 42. Remarks and conclusions 48 Chap. VI. — Statuettes from the temenos of Aphrodite. 43. Circumstances of discovery 55 44. Origin and traditions 55 a VI CONTENTS. SECT. PAGE I 45. Male type and variations . . 56 j 46. Female type and variations . . 57 j 47. Seated figures, &c. . . . .57 48. Their place in the history of sculpture . . . . . 58 | Chap. VII. — Temenos of Hera. * j 49. Temenos and temple . . .60 50. Objects found . . .61 Chap. VIII. — Inscriptions. 51. From the temenos of Aphrodite . 62 52. From the temenos of the Dioscuri . 67 SECT, PAGE 53. From the temenos of Hera . . 67 54. From elsewhere, on pottery, incised, &c. ...... 67 55. From elsewhere, on stone . . 68 Chap. IX. — Conclusion. 56. The age of Naukratis . . .70 57. The inscriptions of Naukratis . .72 58. Results and prospects . . .74 APPENDIX. Kum el Hisn, by F. LI. Griffith . . 77 LIST OF PLATES I. II. III. IV. XII. XIII. XIV. XV. XVI. XVII. XVIII. XIX. XX. XXI. XXII. XXIII. XXIV. Temenos of Aphrodite, Temple of Dioscuri, Plans. Temple of Aphrodite, Plan. Temple of Aphrodite, Sections. Naukratis 1886, Plan. Pottery from the temenos of Aphrodite. Pottery and statuettes from the temenos of Aphrodite. Statuettes, &c., from the temenos of Aphrodite. Statuettes, &c., from the temenos of Aphrodite, and from the town. Selection from the cemetery. Small objects from various sites. Scarabaei, paste and stone. Small objects from various sites. Objects left at Bulak, 1885. Inscriptions from the temenos of Aphrodite. Inscriptions from other sites. Hieroglyphic inscriptions. Klim el Hisn, plan and inscriptions, &c. NAUKEATIS. PART rr. CHAPTER L GENERAL NARRATIVE OF THE SEASON’S WORK. (1) The discovery of Naukratis and the first season’s work upon its site have already been described by Mr. Petrie. But before we enter upon the description of another year’s excava- tions, it seems advisable to pause for a moment, and to realize what was already accomplished when Mr. Petrie left Egypt in the summer of 1885, what still remained to be unearthed by a continuation of his successful labours. In forming such an estimate of the results of exploration on an ancient site, it is well to compare the scholar’s expectations, based upon literature and history, with the material and tangible attainments of the spade. And at Naukratis it is both easy and useful to carry out this comparison. For the literary evidence, if not very extensive, is clear and definite ; and in the process of excavation a fortunate pro- fusion of testimony, and especially the testimony of inscriptions, has removed from the sphere of discussion or doubt the identification of the various sites and buildings that have come to light. From ancient literature we know of five early religious foundations at Naukratis ; four of these are mentioned by Herodotus : first the Great Hellenion, then the precincts separately dedicated by the H^ginetans to Zeus, by the Samians to Hera, and by the Milesians to Apollo. Of a fifth foundation, apparently as early and as important as any of these, we hear from Athenmus, who tells a story, to which we shall have to recur, about the favour shown by Aphrodite to a Naukratite in danger at sea, and about a Cypriote image dedicated by him to the goddess in her temple at Naukratis. Now of these five temples or sacred en- closures, two, the Hellenion and that dedicated by the Milesians to Apollo, were found by Mr. Petrie in his first year’s excavations. One or two fragments of inscribed vases had given reason to hope that two more, those of Aphrodite and Hera, might also come to light ; but no definite conclusions as to their exact sites or identification were possible. Another temenos, not mentioned by classical authors, w'as also discovered, and was proved by the evidence of inscriptions to be sacred to the Dioscuri; but their temple still remained buried. Mr. Petrie had also produced a most careful and elaborate plan, not only of the sites on which he had worked, but of all the streets of the ancient city, so far as it was possible to trace them. (2) Such was the state in which Mr. Petrie B 10 NAUKRATIS. left the excavations, after his discovery of the city of Naukratis, and his first season’s work upon its site. Obviously there was every prospect of good results if that work should be continued. But Mr. Petrie, in his characteristic and disinterested zeal for new discoveries, was anxious to proceed to new fields, if he could find any one to carry on the work he had begun, and to gather in the results of his previous attainments. It was accordingly arranged that I should accompany him to Naukratis at the beginning of the next season ; and that Mr. Petrie himself, when he had started the work and stayed with me long enough to enable me to become familiar with the methods that he had adopted, should again pursue his quest of new sites and unexplored regions. The success that has again attended his search is already known to all: it remains for me to give some account of the continuation of his work at Naukratis, both while he was able to superintend it in person, and after he had left it in my charge. (3) In accordance with the arrangement above mentioned, I accompanied Mr. Petrie to Egypt at the end of November, 1885; and after we had obtained our papers from the govern- ment, and engaged our raises or overseers, we proceeded to Naukratis. There, after a few days, we were joined by Mr. Griffith. We were fortunate in being able to hire for the season the whole of the house — or palace — in which Mr. Petrie and Mr. Griffith had in the previous season occupied a few rooms only. This afterwards proved to be a most convenient acquisition, as it allowed ample storage-room for the numerous antiquities that had to be safely housed. We found the site of the city already some- what altered by the destructive operations of the Arabs, who are continually carrying off the earth from ancient sites to spread it upon their fields. In this way the walls of the Great Temenos or Hellenion had almost disappeared, and the appearance of the mound that takes the place of the ancient city had in several respects been altered. But though, on the one hand, this process is destructive, it is also, on the other, of great service to the excavator, for the digging of the Arabs is constantly laying bare new strata and disclosing new sites, and a careful watching of their work and the objects they find will often supply far more information than large and numerous trial pits or trenches. (4) After we had taken a general survey of the city, and resolved upon the spots which seemed most promising either for thorough exploration or for testing by trial trenches, we at once set to work. At first we had only a few workmen, but their number quickly in- creased, till some 200 were in our employ.^ As many as this were kept in the work while Mr. Petrie was still with me ; after he had left I found 120 to 150 was as large a number as I alone could keep under proper supervision. For though our overseers were excellently trained by Mr. Petrie, and thoroughly trust- worthy, it was of course most undesirable to leave any site where excavation was going on for many hours without a visit. At first, however, as there were three of us to direct the labourers, we were able to make a division of the work. Mr. Petrie, while keeping in his hands the general direction of our plans, de- voted himself more especially to arranging and watching the large trial trenches, by which he was testing various interesting parts of the mound. Mr. Griffith paid special attention to the work that was being done in carefully clearing ascertained sites within the ancient city ; and I directed the workmen who dug ' The numbers, here and elsewhere, include the boys and girls who carried baskets, as well as the men who worked with the mattock. GENERAL NARRATIVE OF THE SEASON’S WORK. 11 some trial pits in the low mounds to the north, where we hoped to find the ancient cemetery, and superintended the excavation of the tombs after our search had proved successful. But though we thus undertook respectively the responsibility of some portion of the excava- tion, our work was really done in common, or at least in consultation with one another. (5) The first work attempted and the first result gained was in the enclosure which Mr. Petrie had in the previous season identified as the temenos of the Dioscuri. Here, after a very httle earth had been removed, some oblong structures of mud-brick were found. These were faced with plaster ; indeed, it was the plaster facing that led to their dis- covery, for without such a facing it is often very difficult, if not impossible, to tell a wall of unbaked mud-brick from the hardened mud that surrounds it. It soon became evident that these structures had been either the bases of pillars, or the lower portion of those pillars themselves. Thus we had the portico in the front of a small temple, facing westwards, as might be expected in the case of a temple dedicated to heroes, and not to a god. After three weeks of excavation, the site of this temple was completely cleared ; but only the northern wall of its cella could be found in addition to the pillars at first dis- covered ; a few fragments of vases, some inscribed with dedications to the Dioscuri, also came to light. (6) The next site to be found was that of the cemetery of the ancient town. It has already been mentioned that we determined to look for this in the low mounds that lie to the north of Naukratis ; one of them is covered by a modern Arab village ; but another smaller one was unoccupied, and offered a convenient opportunity for excavation. Three trenches were cut across it, and in a few hours the middle trench came upon a tile coffin. The coffin contained nothing except some traces of bones; but more coffins and other indications soon proved that we had found the object of our search. Accordingly all the men were drawn into aline, and a deep trench, descending to the water level, was dug along the whole length of one side of the mound. When completed, the trench was made gradually to advance across the mound, so that the whole of the contents of this portion of the cemetery were turned over — a process which occupied sixty work-people for about ten weeks. Of the graves and their contents I shall have to speak in a subsequent chapter, but it is as well to say at once that the results of this search have been rather disappointing. It is at least clear that only a portion of the cemetery of Naukratis has been exhumed — and that por- tion one which dates from a later time, when the city had already passed its most flourishing periods. Some trial pits were also sunk on un- occupied bits of ground in the village, between the small mound to the north and the ancient city itself ; these served to show that the cemetery also underlies this village, but the tombs actually found were precisely similar, both in age and character, to those discovered in the smaller mound. However, it seems a probable supposition that the earliest and most interesting part of the ancient cemetery of Naukratis still lies buried beneath the Arab village. But in the heavy mud soil of the Delta it is not to be hoped that any vases or other breakable objects could be recovered in a complete or uninjured condition. (7) Next, in order of time, though not in order of importance, comes the most interesting and fruitful discovery of the season — the temple of Aphrodite. The story told by Athenaeus, and already referred to, is of a nature to give us a peculiar interest in this temple and its contents ; and last season some fragments of pottery bearing dedications to the 2 12 NAUKRATIS. goddess had been found not far from the region in wliicb her temple and precinct were ultimately identified. Accordingly numerous large trenches were dug in the neighbourhood of the place where these inscriptions had been found ; but though one of these trenches actually cut into the temenos wall, within a yard or two of the temple itself, it was not by such means that the temple was actually found. Thus an excellent illustration was afforded of the comparative uselessness of extensive trial trenches upon such a site. The sides of a trench usually present a uniform black surface; and even if a few courses of mud- brick are here and there distinguishable, it is by no means easy to infer the size and direction of the wall of which they form a part. The ultimate discovery was, as has been already said, made by other means. Some men, digging on their own account, close to the end of our trench, came by chance upon a number of fragments of pottery with incised and painted dedications to Aphrodite. Here at last it was clear that we were in the immediate neighbourhood of the temple ; the particular spot afterwards proved to be within the temenos, just to the north of the temple itself. This building was identified without much trouble after a careful examination of the site. It had been already observed by Mr. Petrie in the previous season, and marked upon his plan ; but a deceptive appearance of a door in one corner had pre- vented his recognizing its true arrangement and purport. As soon as the identification had been made, all the loose rubbish on the sur- face was cleared away from the temple and its neighbourhood. Thereupon an oblong structure emerged in front of the face of the temple towards the east. The corner of this structure had been visible before, but no more. It consisted of thin mud-brick walls, faced with plaster. From its position, the conjecture was obvious that this was the great altar of burnt offering — a conjecture soon changed to a certainty by the discovery of ashes and fragments of bone within the structure, and of steps leading up to it from the temple. But it is perhaps worth while to record a doubt which at first arose, as the possibility of such a doubt is an illustration of the difficulty of distinguishing unburnt brick from mud. It was for a while hard to decide whether the brick, to which the plaster served as a facing, lay within the plaster, or without it, whether the coat of plaster served to cover the inside of the walls of a small chamber, or the outside of a more or less solid structure. Then within the temple itself a wall was found dividing the cella from the opisthodomus. But more than this, it was clear that the divisional wall, and the lower part of the main walls of the temple, which were partly faced with plaster, belonged to an earlier building than the one which was at first visible ; these earlier walls had clearly served as a foundation on which the later ones, which we had at first seen, were built ; and the floor of the latest temple must have been at the level where the earlier walls ceased. When, however, the level of the altar came to be considered, its top was found to be on a level with the floor of the earlier of the two temples we had already found ; and its base was some six or seven feet lower. Now there was no reason for assuming any previous inequality in the level of the ground ; hence it was probable that the altar must belong to a building of much earlier date, which was buried to a con- siderable depth before even the earlier of the other two temples was begun. Of this earliest temple the walls soon came to light in the course of excavation ; they had served as a foundation, on which the walls of the second temple had been built, just as these other walls had, in their turn, supported a later structure. Thus was found the plan of three temples of very different dates ; the details of the GENERAL NARRATIVE OF THE SEASON’S WORK. 13 differences between tbe three must be reserved for a special chapter. Of course all these dis- coveries were -not made at once ; though some indications quickly appeared, three months elapsed before the earliest temple and its pre- cinct were completely cleared. One of the earliest trenches, however, sunkto thenorthof the temple, aroused expectations which were after- wards fully realized ; deep in the earth was found a very rich stratum of fragments of the finest painted and inscribed vases. But in order to reach these without fear of damaging them, of course the plan of sinking narrow trenches had to be given up, and the whole earth was gradually removed in horizontal layers. Thus it was possible to ascertain exactly the level of everything found, and also to extract carefully delicate or breakable objects ; but such a process necessarily caused a long delay before the rich layer of fragments could be reached. (8) Meanwhile work in other parts of the town was progressing. Another fortunate chance led us to the discovery of an important site. A stone was offered us bearing an in- inscription (lepov Jlo? ^ATTorpoTTaLov), which seemed to denote that the place it came from was sacred. Accordingly we ascertained from the finder the exact spot where it had been dis- covered, and began an investigation of the neighbourhood. This proved to be within the large enclosure called the Patestra by Mr. Petrie last year. Very soon one or two frag- ments dedicated to Hera were found. Then the work was left for awhile ; when it was resumed some two months afterwards, a few more such dedications and the foundation of a building were found — enough to prove that this enclo- sure was the temenos once dedicated by the Samians to Hera, and that one of the most promising sites in Naukratis had nothing left to reward the excavator except its name. Here was a case in which the digging of the Arabs had proved most disastrous. The temple had been of stone ; therefore all of it, except oue or two insignificant fragments which were recovered, had been carried off : and the actual dig^ino^' had reached right down to the original ground level of ' the ancient town, in some places even below it. Of course all antiquities found in this process had been thrown away or destroyed. A few fragments were found among the heaps of potsherds and other refuse that had been left behind ; but it was from a few bits of undisturbed earth that there came the best and clearest specimens of dedications to Hera. This fact seems to prove that these dedications really belong to the site upon which they were discovered, and have not wandered there by chance from any other region. (9) Thus we had, in the first few weeks of our stay, discovered two more of the most important temples at Naukratis, whose existence was known from the evidence of literature : but it was not to be expected that our search would in every case prove equally fortunate. Large and deep trial trenches were sunk in various parts of the ancient town which looked promising, and which had not yet been tested. From these not very much further result was obtained ; but the chief discoveries made by their means were at opposite ends of the town, one on the N.E. corner of the mound, the other between the Hellenion and the canal. (10) In this latter place it seemed probable that a great road or avenue must have existed, to serve as an approach from the quay to the principal sanctuary of the city. A broad and long trench was made along the probable direc- tion of this avenue ; and in it was found the lower part of a granite sphinx, of the Ptolemaic period. The sphinx had been split up, in order that the stone might be used for other pur- poses ; even on the part left wedge-holes could be seen, intended to serve for splitting ofi' NAUKRATIS. li another portion. But this destruction must have taken place at some remote period, for the sphinx was about six feet below the present level of the ground. Bj a later search two large portions that had been broken off were ’recovered, and then were restored to their original position — not without some difficulty, for they could only just be lifted by six or seven men. Thus the whole sphinx again stood complete, with the exception of the face only, which could not be found. (11) The discoveries made by means of the trial trenches at the N.E. of the mound were less definite and comprehensible. A large wall was found, which for a while we supposed to be the enclosing wall of the city itself : con- nected with this were various large and very solid buildings, constructed partly of mud-brick, partly of stone; the latter material had of course disappeared, leaving only its founda- tions to be traced. The purpose or use of these buildings could not be conjectured either from their plan or from the objects found in them ; but they did not appear to date from a time before the Ptolemaic period. (12) Almost all the excavations already referred to were begun within the first three weeks after our arrival : in order to avoid con- fusion, I have in many cases stated at once facts which did not become clear till weeks or months more had elapsed. Mr. Petrie left Naukratis to proceed to other work, farther to the east, on the 5th of January, 1886 ; Mr. Griffith had also gone a few days earlier. I accordingly assumed the charge of the work, and continued it for the next three months. For a week (11 — 18 January), I was fortunate in having the valuable help of Mr. A. H. Smith, who especially devoted himself to watching the work at the cemetery during that time ; and I was very sorry that his health prevented him from making a longer stay. For the rest of the season I was alone at Naukratis ; and I found my time and attention very fully occu- pied with the excavation of the various sites I have already mentioned. I endeavoured as far as possible never to have my work on more than two important sites at a time, in order that I might be able both to watch it con- stantly myself, and to keep it under the eyes of my two excellent and trust- worthy overseers, Mahajub and Said Abu Baud. Of course many smaller outlying pieces of digging had to be done ; but these were, as a rule, carried out by contract work. The temple of the Dioscuri was finished in the beginning of January ; for the next two months the temple of Aphrodite and the cemetery formed the two chief centres of excavation ; later the temples of Aphrodite and of Hera. (13) The chief feature of the work during the whole of these months was the richness of the temple of Aphrodite and its precinct. I have already referred to the layer of fragments whose existence we had ascertained, but which had to be approached gradually. Even before the level was reached at which these fragments had pre- viously been found, numerous pieces of pottery and statuettes made their appearance ; and that too not only on the north side of the temple, but also on the east and on the south. Here was no trench like that in which Mr. Petrie had in the first season found the bowls discarded from the service of the Milesian Apollo, but a continuous layer of refuse, bounded only by the walls of the temenos itself, and sometimes even scattered beyond them. How this refuse came into the position where it was found is, and must remain, a matter of conjecture ; but it is hard to believe that any event in the peaceful administration of the temple could have produced such a result. It seems hardly rash in this connexion to think of the disturbances which accompanied the Persian invasion of Egypt, especially as the period preceding that invasion would well suit the style and character of the various vases and GENERAL NARRATIVE OF THE SEASON’S WORK. 15 statuettes that have been recovered. Nor is the consistency of the layer of refuse other than one might expect to have been formed if the temple were destroyed by violence. It was made up of the fragments of hundreds, or even thousands of vases, all mixed up together in the utmost confusion, and of numerous archaic statuettes, or fragments of statuettes. Evidently these had all been votive offerings in the temple of Aphrodite, as was indeed attested by the dedi- cations with which many of them were described. Mingled with them here and there were ashes or pieces of bone. Such was the extent and rich- ness of this layer, that even after we had gra- dually cleared off the earth from above it, and left it free to be worked, nearly two months elapsed before the whole of it could be gathere(^ in. Meanwhile, my resources for storing this great quantity, and keeping the product of dif- ferent parts separate, were somewhat strained. At first we had attempted to keep all the frag- ments which we found in paper bags, which had been provided by the foresight of Mr. Petrie ; thus each bag could be labelled, and the sites distinguished by an easy system of numbering. But this manner of proceeding soon proved inadequate ; then the baskets used for carrying the earth in digging were pressed into the service, and the number of these baskets of potsherds brought in every day came to be considerable ; on one day there were nine, on several seven, from four to six on others. Thus it was easy to bring the frag- ments in separate baskets up to the house, and there they were emptied, at first, into boxes devoted to various portions of the site. Soon all my boxes were filled, and then I felt the benefit of the space at my disposal, for I was able to devote one room to the reception of these potsherds, and to partition off its floor in such a way as to keep the various piles distinct. After the first few days, it also became impos- sible to sort, or even select these fragments as they came in. This is obvious, when one con- | siders that sometimes perhaps as many as 5000 potsherds were found in a single day, and the work of sorting and selecting can be properly done only by daylight. I accordingly had to content myself with glancing through the con- tents of each basket, to make sure that it really consisted of painted fragments of vases, and had no rubbish put in to fill up. Though I after- wards rejected a certain amount, I was obliged to wait till I had worked through the whole in England, before making a final selection. For if I had at random thrown away any pieces that seemed in themselves unimportant, I should have run the risk of spoiling the completeness of a vase whose other fragments had been re- covered. If I had kept too much, the mistake admitted of an easy remedy ; it was not so if I had kept too little. (14) Towards the end of the season, when work was slackening a little on the site of the ancient city, I managed to make trial of a spot of which Mr. Petrie entertained some hopes. It seemed not improbable that we might there find another and an earlier part of the old ceme- tery. This was on the mound of Neqrash, a vil- lage which, as Mr. Griffith ^ has pointed out, still preserves the name of Naukratis itself. It is situated about a mile away towards the east. Here I sank several large and deep pits, but I found nothing whatever, except a bit of wall in one place, apparently of the Ptolemaic period. This at least proves that here there was no cemetery like the one we found on the north of the city ; for there the graves were so closely packed that even very small trenches were sure to find something, and those I made at Neqrash were of considerable dimensions. (15) Another and an even more promising attempt met with an equally disappointing re- sult. At different times during the season, as occasion offered, various private dwelling See page 80. 16 NAUKRATIS, houses in the ancient town were cleared. But whether because this attempt is not of much use on such a site, or because the houses were unfortunately chosen, very little was gained by the clearance. Of one house, indeed, apian was recovered, but as there were no doors whatever, it was clearly only a basement or cellar plan, and though something might be inferred from the walls as to the position of the dwelling rooms above, none of this upper portion was left standing. In other cases even less success was attained. One or two houses look very promising to the excavator in the conjectural restoration of their arrangement, given in the plan made in the preceding season by Mr. Petrie. This plan was made with the help of various indications, such as small extant bits of wall and the like. And as these bits of wall still stood up amidst the wil- derness of rough or late pottery, it was natural to suppose that if their base were cleared the system of which they formed a part would at once become apparent. But this was not the case ; when the excavation had been carried down to the most ancient level of the ground, the arrangement of the houses remained as much a matter of con j ecture as before. And very little in the way of antiquities was found during this work. Doubtless the houses, built entirely of mud-brick, had gradually subsided and served as a foundation for the buildings of a later generation : and in this way little either of the walls themselves or of the objects they once contained remains to be found by excavation. (16) But on such a site as that of Naukratis the explorer does not rely only upon the work- men in his employ and under his direction for the discovery either of sites or of antiquities. Mention has already been made of a practice of the natives of the Delta which is of the higfhest importance to archaeologists, both for what it brings to light and for what it destroys. But for this practice the site of Naukratis would not yet have been discovered ; yet it must have led to the destruction of many valuable and in- teresting objects that would else be still safely hidden under twenty feet of soil. Some years ago the discovery was made that the earth from a mound in the Delta — that is to say, from the site of an ancient city — forms an excellent “ top dressing” to spread over the fields : and the knowledge of this fact spread all over the fertile districts, and led to the digging away of all the mounds they contain. Within, all these mounds now present the appearance of a sponge or a dry honeycomb : for many of the walls and most of the street lines have been left as useless, while the mud that surrounded them has been carried away. At the proper season for the work, in the spring months, hundreds of men fill the mounds, and trains of donkeys and camels bear to the fields the earth that is dug out. Naturally in the course of the digging the men find antiquities also. But these are, upon an unwatched site, liable to destruc- tion. Some are rescued by Arab dealers, who are well aware of the process and its results ; but of course they pay no attention to frag- ments ; and when any object has come into their hands, the place and circumstances of its discovery are almost always lost. But on this last point one must not speak too strongly, since it was from information supplied by one of these Arab dealers that the site of Naukratis was dis- covered. However, with them we are not now concerned ; for Mr. Petrie’s vigorous measures against them in the previous season deterred any of them from making their appearance last year. When a site where the earth is being dug out is constantly watched, a most effectual and inexpensive manner of excavation offers itself. All objects, however insignificant, are sure to be brought and offered for sale : and by judicious management it is possible to secure all that are worth having. In this way many interesting things were obtained ; but, above all, in this way the sites of the temples GENEEAL NAREATIVE OF THE SEASON’S WORK. 17 of both Aphrodite and Hera were discovered. When any such site was found, it was of course necessary to avail oneself of the authority sup- plied by the official authorization of our work, and to stop all other digging upon the temple so discovered ; but by taking the original finder himself on to the work, it was easy to avoid the danger of his being induced by the hope of further discoveries to conceal the place where he had found what he brought. To watch these independent workers, and to collect what they find, is a most necessary part of the excavator’s duty; and an account of the season’s work would be incomplete if it omitted to mention so important a help to acquisition and discoverv. (17) If I here add a few words about the system on which the work was conducted, it must be borne in mind that the system is Mr. Petrie’s, and that I obtained it from him and preserved it in every essential particular. For his successful experience had already proved it to be practically advantageous ; and the over- seers and many of the workmen were already trained by him to follow it. If we set on one side all independent work on the part of the natives which had only to be watched for the sake of observing or acquiring any results it might produce, all our excavations may be divided into two classes. These may be roughly defined as piece-work and time-work. The system of piece-work was necessarily adopted wherever the constant presence of the overseers was impossible. A certain space for a pit or a trench was measured out, and a certain depth ordered : when the hole was completed its con- tents in cubic metres was measured, and pay- ment was made accordingly : thus any man’s slowness or laziness affected no one but himself, since the time he spent made no difference in the payment he received. But since this sys- tem, however convenient in many ways, encou- raged rapid and careless digging, it obviously could not be adopted in any place where there I was a probability of the discovery of delicate or fragile objects, or of walls which must be cleared and left in their original position. In such cases therefore — in any case, indeed, where a site had been identified and was being tho- roughly excavated — the second system had to be followed, and all the workpeople had to be paid by the day. This of course necessitated a daily register of their attendance. In receiving from workmen the antiquities they found in their work, we always allowed them just the same amount as we should have given for the same object to any one who had offered it for sale — a price better than any Arab dealer would be likely to give if he came that way. Thus there was no inducement for them either to conceal anything found in the hope of selling it at a better price in another quarter, or to give a false account of the spot where any object had been discovered, whether the discovery was made within our work or outside it. For if falsehood offers no material advantage over truth, and at the same time involves the risk of detection and punishment, it is hardly likely to be preferred for its own sake, except by in- corrigible offenders. (18) It may be worth while to give a sketch of the way in which the day was generally spent at Haukratis, especially as such a sketch may serve to illustrate the manner in which work has been carried on there. Here again I shall be describing a plan which is, in all essen- tial points, derived from Mr. Petrie’s example : but I would refer rather to the later part of the time during which I was alone at Naukratis, when I had found out by experience the most convenient disposition and organization of my work. By half-an-hour after sunrise all those em- ployed by the day were expected to be in their places. Soon after this time 1 went round to the sites where they were working, varying from day to day the order of my visits. On 0 18 XAUKRATIS. each site I entered the names of all who were present for the day’s work, and enrolled, if necessary, new workmen or filled the places of absentees. After seeing all the excavations well started, and giving directions for their continuation, I usually returned to the house at about nine or ten o’clock. Each time when I returned during the day, I had to unload myself of an accumulation of small antiquities, mostly bought up from those not actually in my employ : and to pay for these I filled my pockets, each time I went out, with a truly Spartan weight of money ; this weight may be realized by considering the fact that 5/. worth of the copper coinage of the country, which was sent down to me from Cairo for the sake of making these small purchases, weighed about 1 90 lbs. Hence it was obviously desirable to revisit the house several times in the day, both to leave my acquisitions and to obtain fresh supplies of copper. After breakfast, I went out again to look after my men till noon ; then I gave, from some central point, the signal for the hour of mid-day rest and food : as the days grew longer and the heat greater, this interval was naturally a little lengthened also. Wlien it was past I again gave the signal for the renewal of work : and the uninterrupted time that remained before sunset afforded the best opportunity for carrying out any piece of excavation that required continuous personal care or supervision. In other cases, after I had seen the afternoon work started, I generally spent the next hour or two in visiting the parts of the mound where no excavation of my own was going on, in measuring the trenches finished by men who were digging by the metre, and in laying out new ones for them to go on with. As the hour of , sunset ap- proached, it became necessary to visit my work- people again, and to sort, select, and roughly record all that they had found during the day. The value of anything which I kept was then entered to the account of its finder. On a site like the cemetery, where separate graves were constantly being discovered, it was best to take at once the contents of each grave as they were found : these could then be laid aside and brought in together at sunset ; in the same way all the larger or heavier articles obtained from men who were not in my employ were labelled, if necessary, or enclosed in a wrappage, and left in charge of one of my overseers till evening. But on a site like the temenos of Aphrodite, where fragments, for the most part insignificant in themselves, were continuously being discovered, the only practicable way to deal with them was to allow all to accumulate till the end of the day : then, in the last few minutes, I had to inspect and to register whatever each had to offer as the result of his day’s work ; though here too I could of course at any hour take over and lay aside any article of peculiar interest or fragility. When I saw the sun touch the horizon, I again gave the signal, and all returned home. Towards the end of my stay a usage sprang up which formed a most interest- ing and characteristic feature in our daily programme. Whatever may have been the interest or appreciation felt by my workpeople for the objects they were instrumental in dis- covering, they at least appreciated the material benefit which they themselves reaped from the large quantities in which these were found. Accordingly their rejoicing at the good fortune we had met with was unfeigned, and it spon- taneously found for itself a means of expression. When the signal of sunset had been given, the people, instead of hurrying away to their homes, formed themselves every night into a kind of festal procession ; they had what the Arabs called a “ fantasiyeh,” this one was regularly known as the “ fantasiyeh of the pot- sherds.” Its order was usually after this fashion. The principal part of it started from the temple of Aphrodite. Several of the girls, selected by the overseer, raised upon their heads the baskets containing the potsherds and other GENEEAL NAERATIVE OF THE SEASON’S WORK. 19 small antiquities found during the day, and led off the procession ; at their head went a piper, who came to meet them from the village, helped sometimes by one or two amateurs among the company. Then followed, in solemn state, my chief overseer, a splendid- looking Arab, and the rest of the workpeople in due order. At some point upon their homeward route they were joined by my other overseer and the people from his work, who fell into their places and swelled the procession, which, thanks to the flowing Arab dress, really had an impressive effect. I did not usually wait to accompany it, but went on in front to my house ; there I distributed the contents of the various baskets among the heaps to which they belonged. In the evening not much was to be done beyond a rough sorting of what had been brought in, and recording as far as was possible at the time the results of the da)^ Five days in every week were thus spent ; and it is worthy of notice that during the whole of the four months, from December to March, there was not a single day on which work had to be suspended or even modified on account of the weather. All this time the shade temperature was moderate, usually keeping between 60° and 75° (Fahrenheit) in the day- time, though occasionally in March the thermometer went above 80° ; at night it was cool, the extreme registered generally being between 40° and 50°. Thus it will be seen that the climate of the Delta in the winter will not seem at all unusual to one accustomed to that of England — except, indeed, for the duration and the intensity of the sunshine. (19) When the season’s work was drawing to a close, it became necessary to consider the packing of the antiquities that had been found, their exportation from Egypt, and their trans- port to England. In this matter, as indeed in every other, it is a pleasure to me to be able to acknowledge the considerate help and co-opera- c tion of Dr. R. S. Poole, then Secretary to the Egypt Exploration Fund. He not only at once responded to any request on my part, but often foresaw and provided against difficulties or delays that were likely to arise. Before our work began, we had come to an understanding with the authorities that we should be allowed to export most of the antiquities we discovered, on the condition that a certain portion of them should be selected to be kept at the Bulak Museum. At the end of the season, I made application to M. Maspero, in order that this condition might be fulfilled ; and I have to acknowledge the great courtesy with which he al- lowed meto bring onlyafew representative speci- mens to Bulak, and his care to select nothing for his museum which by its separation from the rest might injure the scientific completeness of any series of objects discovered atNaukratis. Meanwhile I had packed all the antiquities found during the season into about eighty cases ; a somewhat deceptive bulk, owing, as has already been explained, to the impossibility of selecting at once the fragments of pottery that were worth keeping. Now I had only to bring these to Alexandria, whence the com- mittee of the Fund had already provided for their transport by an Indian troop-ship. But there is one more acknowledgment that I must make before leaving this part of my account. From Mr. Cookson, H.B.M. Consul at Alexan- dria, I met with the greatest kindness and assistance in passing my boxes through the formalities that were still necessary before they could be shipped ; an assistance which was most efficiently and willingly rendered to me by Mr. Harris, the chief constable of the Consulate, in whose care I ultimately left the boxes to await their despatch. (20) Here ends the account of the discovery of the last season’s antiquities at Naukratis, and of their exportation from Egypt. Before pro- ceeding to a more detailed account of the 2 20 NAUKRATIS. antiquities themselves, and of the various sites upon which they were found, it remains for me to say a few words about their condition and treatment after their safe arrival in England. I had all the more important cases brought to Cambridge, in order to work at them with plenty of room at my disposal: for this room I have lo thank, first, the authorities of Gonville and Caius College, and later the director of the Eitzwilliam Museum. Most of my work at Cambridge calls for no notice beyond the publication of its results ; but of the pottery from the temenos of Aphrodite a little may here be said. By gradual sorting and compar- ing, I have recovered some vases in an almost complete state, and about one half of others : many are represented by considerable pieces ; others only by insignificant or isolated frag- ments. The total number of vases that have contributed their fragments to this vast layer of pieces must have been very large : and they must have been broken up before their fragments were cast out ; for parts of the same vase were often found scattered to the east, the south, and the north of the temple. Indeed, in conse- quence of this confusion, I found it useless to preserve in sorting the distinctions of place that had been so far kept. As regards the condition of the vases, and the number of- pieces into which they had been broken, the following figures will speak : of one large bowl, now nearly complete, about 70 fragments were found ; of others again, that have been recovered only in part, I have counted the pieces of half-a-dozen of the first that came ; these amounted to 47, 46, 77, 60, 31, and 17 respectively. As these were, as I have said, completely mixed up together with the remains of other similar vases, it will be clear that the task of separating them was one which required a considerable amount of time. (21) The work of sorting and of mending has now been almost completed. In the British Museum will be found a representative selection of the various objects that have been recovered ; the rest have been distributed among various museums, where it is hoped they may be seen and studied. Fora specimen of the style of the pottery I would refer those who cannot see the vases themselves to the excellent coloured Plate in the Journal of Hellenic Studies, PI. Ixxxix (1887), as well as to the direct photographic reproductions in thus volume. 21 CHAPTER 11. THE CEMETERY. (22) One of the most important sites to be looked for by the excavator of an early Greek city must always be the cemetery. For it was ever the custom to bury together with the dead various objects intended for their use or delectation. Such at least must originally have been the intention of this custom, even if in historical times its practice had survived the primitive beliefs out of which it arose. Hence it follows that from the imaginary needs of the dead we may learn much as to the real needs of the living ; and the articles of use or of ornament that were adapted to fulfil those needs will be in both cases alike. The cemetery, then, must always be sought ; but the clues that may lead to its discovery are not in all cases the same. In the case of most Greek sites two indications, perhaps, must be followed most carefully. Sometimes a rocky formation naturally lends itself to the excavation of caves and mottoes o to receive the dead ; in this case the nature of the ground will often guide the explorer to his goal. Sometimes considerations of convenience are outweighed by the desire that the tombs may be conspicuous, and may record the burial of the deceased for the observation of all that pass by; if so, obviously the most fitting position will be beside the most frequented roads that lead to the city. Often, again, both objects will be combined, if a convenient face of rock is found beside a road. At first glance it might seem that considerations such as these could hardly lead to the discovery of the cemetery of Naukratis. The level surface of the delta is broken only by the low mounds that testify to the long-continued settlement of towns or villages, whose accumulating debris has gradually risen above the even plain. The ancient roads cannot be traced, nor if they could, would they lend us much help : for tombs could hardly be built beside them in the fields yearly overflowed by the Nile. Yet some mound, natural or artificial, must have been found wherein to bury the dead ; and perhaps it is easier to find the ancient thoroughfare near which the people of Naukratis placed their tombs than one might at first be disposed to imagine. Their traffic with the interior of the country may have been partly by land : but their communications with their own people, with those whose sympathy or remembrance were to be excited by conspicuous monuments, must have been exclusively by water. The canal therefore was the high-road beside which the graves of Naukratis were to be sought. Now as some low mounds lie to the north of the city, that is, in the direction of Greece, and near to the present bed of the canal, nearer still to what was probably its ancient course, these mounds seemed to be indicated by all circumstances as a likely site for the Greek cemetery. (23) These anticipations, as has already been stated, were soon proved to be correct. The small mound to the north of the Arab village proved to be full of graves of Hellenic or Hellenistic period: but of monuments above the surface of the ground very little was discovered : if, as is most probable, tliey were of stone, there was little chance of their being 22 NAUKRATIS. left undisturbed. They seem not to have escaped this fate even in ancient times ; for the only two sculptured stelae of the ordinary type that come from Naukratis were found within the limits of the ancient city. As burials cannot have been allowed to take place within these limits, it must follow that they were in ancient times converted from monu- mental to practical use ; one of them with its face downward formed part of a later pavement. To these must be added the curious stele found in the previous year in or near ^ the temenos of the Dioscuri, bearing the inscription Teaw elfxl crrjfxa,^ “ I am the grave- stone of Teaos.” The two sculptured stelae found last year have no inscription upon them. The first is in very low relief, of an execution too good for any period but the fourth century ; in its subject and treatment it resembles the well-known Attic grave- stones of the time. Though a portion of it is gone, one figure, that of a boy, is still complete, and is very graceful in pose and character ; two other figures can also be distinguished. The whole seems once to have represented one of those scenes of greeting or departure that are so common upon gravestones of the finest period. The other stele, though far inferior in execu- tion, is interesting from its subject. It is a good and very complete representation of the scene so often recurring upon Greek sepulchral reliefs, the scene known commonly as the funeral banquet. Here we see a man with a somewhat cumbrous garland around his head, reclining upon a couch and supported on his left elbow : his breast is bare, and a mantle is wrapped round his legs; his left hand holds a cup, his right rests on the edge of the couch, just above the table that is before it. By his feet sits his wife, her feet resting on a foot- ' It is hard to define the exact boundary of the temenos in the region where this stele was found. * Nauk. I., PL XXX. 15 ; p. 62. stool : in her left hand she holds a shallow cup, out of which drinks a serpent, which coils round over her left shoulder : her head has been carefully chiselled away. Two boys wait on the pair ; the one on the left is clothed in a mantle which he supports with his left arm ; his right hand is raised as if to beckon. The other boy holds a jug (oenochoe) in his left hand ; with his right he offers a cup to his master. The dress of this boy is very peculiar ; it seems to consist of a single close-fitting garment of some smooth hard material, terminating on the arms and legs in short sleeves and drawers. An ample feast is provided ; on the table are cakes and fruit, as well as a cup : on a low stand by the head of the couch is a large bowl, doubtless for mixing the wine ; on its rim and on a stand behind are several smaller vases. All the furniture is also sumptuous ; the couch has elaborately-turned legs, and is provided with tasselled hangings ; the table and footstool are of ornamental construction, their legs ter- minating in lions’ feet : even the large mixing bowl has a design upon it. Above this scene, in the corner to the left, is a square window, through which is seen a horse’s head, ready bridled, as if to take its master upon a journey. The whole is set in an architectural frame ; on each side is an Ionic column : on them rests a narrow entablature, its frieze divided as if into small metopes ; above is a pediment with curved top, and a round disc like a shield in its centre ; a tone end remains an acroterion of palmetto design. The subject of this relief need not long detain us : it is of a well-known type, which has often been discussed ; I need only refer to Prof. Percy Gardner’s paper in the Journal of Hellenic Studies for 1885, where other authorities are quoted. It seems that the feast typifies the enjoyment by the dead of the offerings made by survivors at his tomb ; with this thought is mingled one of his life in another world. Two somewhat incon- sistent symbols are here, as often, introduced ; I THE CEMETERY. 23 the serpent, in whose form the departed is often looked upon as accepting the food offered to him ; and the horse, the symbol of departure to a far country. The style of the relief calls for more attention. So far as the art is Greek, it is of small interest, for it is obviously of a lateanddegenerate period : but many characteristics of the treatment are not Greek. These are most apparent in the small attendant on the left of the scene. The peculiar treatment of the head, especially of the eye, which is represented in full though the face is in profile, the awkward and stiff position of the arms, the backward curve of the fingers, all are mannerisms copied from Egyptian models. The same peculiarities recur elsewhere, for instance in the right hand of the reclining man, and the right arm and hand of the attendant on the right, whose dress also is not Greek. The relief must date from Ptolemaic or possibly Roman times : and here we see the desfen- erate and failing art of Greece taking to itself just those mannerisms and defects of an alien style which it would soonest have rejected in its earlier and better days. The archaic Greek artists may have learnt something from Egypt : but what they sought was help in the difficulty of material expression, not meaningless and lifeless conventionalities. I have said that no sepulchral reliefs were found during the excavations in the cemetery it- self ; but one monument of a similar character appeared. This was a small slab of stone, of about the same size and form as the relief just described ; but it was surrounded only by a plain raised border, and surmounted by a triangular pediment. The plain surface thus left within a sort of frame had evidently once been decorated by painting ; but though I thought, upon a long and careful examination of the stone, I could detect some traces of the design that it once bore, those traces were not clear enough to indicate either the colours used or the subject represented. Another stone monument, of a much more complicated character, was discovered in the cemetery. The fragments that came to light were some portions of Ionic or Corinthian columns, probably once decorating the face of a heroum : they clearly adorned the front of a wall as they are fluted only on one side. One drum was 9^ inches high, and 8^ in diameter ; a base was 6 ^ inches in height, its diameter at the bottom was 14 inches, at the top 11. From these measurements it is possible to gain some notion of the dimensions of the monument to which the columns belonged : it is the only tomb of any architectural pretensions of which any trace has been found at Naukratis. (24) So far we have been considering only such grave-buildings as were originally above the surface of the ground ; many such must probably have once existed, though they have now disappeared. But structures beneath the earth would probably still remain, if they had ever been made ; hence it is more significant in their case to observe how little of them came to light. No built tomb- chambers or even graves were found, except one or two graves of burnt brick, dating from Roman times. In one case some remains of fresco painting, of a blue colour, were found ; this colour seems to have been applied directly to the mud-brick wall, now indistinguishable from the mud that surrounds it; in another instance a grave was lined with a thin coating of stucco. But these were exceptional instances ; as a rule there was no sign that any elaborate prepara- tions had been made for the burial, except mak- ing a new pit to receive the coffin. For the body seems always to have been buried in a coffin of some sort ; I saw no traces of the practice of cremation.^ Nor, on the other hand, do the Greeks of Naukratis seem to have been influenced at all by the Egyptian custom of embalming their dead. The body had 3 Except, perhaps, in one instance ; .see below, p. 27. 24 NAUKEATIS. always completely disappeared ; even the bones could not often be distinguished, and in no case were they well enough preserved to admit of their being kept. It was not uncommon, especially in the case of the richer graves, to place a layer of white sand beneath the body or coffin ; a proceeding most convenient for the excavator, as such a layer can easily be dis- tinguished in the black mud: and thus it is made possible to clear a grave carefully by itself. In one case, also, I found a thin pave- ment of plaster beneath a coffin of terra-cotta ; but this coffin was exceptional in shape ; thus it, as well as the bed on which it is laid, seems not to have been in accordance with the ordinary customs of Naukratis. (25) The coffins themselves call for more particular notice; they may be classified ac- cording to the material of which they were made, stone, terra-cotta, or wood. Only one of stone was found in its original position in the cemetery ; this was a rough sarcophagus, devoid of all ornamentation. It was of massive pro- portions and only roughly finished. The body of the sarcophagus was 8 feet in length, 3 feet 8 inches in breadth at the top, 2 feet 9 inches in depth, externally : its sides and bottom were 8 inches thick. The lid was a flat slab, 18 inches high, with its edges bevelled on the upper side. Within this sarco- phagus was nothing but a few small bronze nails, that seemed to come from a decayed wooden coffin, and the bones of the occupant, now barely distinguishable. I saw one or two other plain sarcophagi in the neighbourhood of Naukratis, which had probably been found under the Arab village on the north mound ; I do not think the small mound I cleared had been disturbed before. The two other materials are far commoner. The coffins of terra-cotta, of about the same consistency as an ordinary tile, were in every case crushed in by the pressure of the earth ; but it was easy enough to see what their shape and construction must have been. They were usually made in two pieces, one to contain the head, the other the feet of the corpse ; and one of the ends usually, but not always, had a pro- jecting flange at the junction. Each half was usually about 35 or 36 inches in length, 20 to 12 in breadth, and about 16 to 9 in depth, the last two dimensions being largest in the middle of the coffin and diminish- ing towards its ends. The coffin was quite plain, and had no ornamentation of any sort. But one or two exceptions to this rule called for notice. I have already mentioned the coffin of peculiar shape that lay upon a thin bed of plaster. This coffin was shaped at the top to fit the head and shoulders of the corpse, so as to present a trefoil outline ; from the shoulders to the feet it narrowed in even lines. I found in one case a hand in rough terra-cotta, the size and shape of a child’s ; this seemed to belong to an anthropoid coffin, with a representation on the lid of the form of the deceased ; but no other traces of such coffins were found. I may also mention here two beasts’ paws of terra- cotta, though their connection with a coffin or anything else likely to have been buried in this cemetery cannot well be traced. While we are speaking of earthenware coffins, another similar method of interment may be mentioned : it seems to have been used mostly in the case of young children, as is indicated by the bones discovered. A coffin was not made, but an amphora was used for the purpose; its top was broken off to allow the body to be inserted ; in one case I found the top lying across the neck, in undisturbed earth, so that it must have been buried so. To judge from the shape and fabric of the amphorae, this practice must have been common at all times, from the sixth or fifth century down to the Ptolemaic or even the Roman Period. In one case were found the handles of an amphora of the shape reproduced in Naukratis I., PI. xvii. 17 ; later forms were more common. THE CEMETERY. 25 (26) From the third material that was used for coffins, wood, one might at first thought expect the least remains. This expectation is indeed fulfilled as far as the coffins themselves are concerned ; the wood has completely dis- appeared in every case. But the beautiful terra-cotta ornaments with which they were once decorated have been found in great numbers : these were gorgoneia, of which about eighty-five were found in almost perfect preservation, gryphons, and rosettes of various sizes and designs. The coffins themselves were probably of the same shape as one in the Bulak Museum, pointed out to me by M. Maspero : it is of Greek workmanship, but comes from a drier soil; and consequently the wood of which it is made has remained. This coffin consists of an oblong box, and a gable-shaped lid, which presents a high-pitched pediment at each end. These pediments, as well as the rest of the coffin, are divided into panels, and ornamented with painting and with terra-cotta reliefs, which are not, however, of so fine work as those found at Naukratis: there are small gor- goneia affixed to the extremities of the round beams that run from end to end of the coffin, along its angles. A similar arrangement would account for the numerous small terra- cotta ornaments that were found in the ceme- tery of Naukratis wherever a wooden coffin had once been buried. Some specimeus of these ornaments have been reproduced upon Plate XVI. 1 — 6. I have called the small terra- cotta masks gorgoneia ; it will be observed that some of these have a pair of sm'all wings springing from the forehead. Mr. Cecil Smith has suggested to me that they should rather be regarded as heads of Hypnos ; when we remember the close relationship between Sleep and Death, such figures seem particularly appropriate upon a coffin ; and the winged type in particular approaches very closely to that sometimes assumed by the god of sleep. But in spite of these arguments, I am still inclined to retain the name gorgoneia.'* The coffins were of essentially architectural construction ; and these ornaments seem to correspond to the terra-cotta antefixes which we usually find in real buildings. A very common form taken by these antefixes is that of the gorgon’s head, originally of the grinning and hideous type that is in early times proper to the monster : thus they resemble in character, and perhaps in inten- tion, the monstrous gargoyles of the medimval cathedral. But the gorgon type, as is well known, lost in later times all its hideous or terrible character ; and it is to the type best known in the “ Medusa Kondanini ” at Munich that our examples most nearly conform . In the suppression of characteristic attributes, they have gone yet farther ; for none of them pre- serve even the conventional knot of snakes be- neath the chin ; the bow above the forehead may be a reminiscence of this, but it is a mere blue riband as here represented. The terra-cotta masks have in many cases preserved traces of colour : the face is white, the hair red, and the riband that ties it blue ; sometimes the eyes also are coloured, the irisbeing red, the outlines darker : gilding is also used for ornamental parts in some instances. Beside the masks of terra-cotta, which were all of the same type, though varying in their size and execution, I also found two or three of white plaster, of a slightly different shape (PI. XVI. 5). These also had traces of colour, especially in the eyes. It is possible to fix with some precision the date of the coffins on which the gorgoneia were used. From the style, which, though already tending to mere prettiness, generally shows good and careful work, they would seem to belong to the end of the fourth century, or the earlier years of the Ptolemaic period. For- tunately, however, we have other evidence by which to test this view : with one of the masks * In this view I am confirmed by M. J. Six, whose trea- tise, de GorgonCy is the standard work on the subject. 1 > 26 NAUKRATIS. was found a bronze coin, having on the obverse tbe head of Alexander with the elephant’s skin, on the reverse the eagle — the well-known Ptolemaic types. This coin seemed from its style to belong to one of the earliest members of the dynasty; thus it fully confirms the date that we were led by other considerations to assign to these gorgoneia. Other ornaments, such as gryphons, bucra- nia, and rosettes of various t 3 /pes and sizes, need not long detain us ; specimens of these are given on Plate XVI. 8 — 14. They also were enriched by gilding and painting, and probably once served to ornament coffins of a similar period. (27) We must now turn to the contents of the graves, the objects of use or ornament that were buried with the dead. Here we meet with greater variety both in the date of the burials and in the nature of the objects recovered : but unfortunately graves from the sixth century, the time when Xaukratis was in its greatest prosperity, are still almost, entirely lacking. Here perhaps the disappointment is greater than in any other case ; for we might have hoped for some complete or even unbroken specimens of the magnificent vases that we know to have been made at Naukratis. This hope has in no way been fulfilled, whether because these vases were for the temple and not for the tomb, or because the cemetery of the sixth century has not been found. Early graves are, indeed, very rare ; the richest ones all date irom the fourth or third century b.c. I have been able to make a catalogue of the contents of about seventy -five different graves ; but it does not seem worth while to reproduce it entirely here ; a few selections will give an adequate notion of its nature. To explain my notation I must state that I called the five trenches, running north and south, ABODE, beginning from the east ; and also divided the mound by cross lines into sections which I numbered 1 — 5, from south to north. But I was not able to draw much distinction between the graves found in different parts of the mound. With this explanation, a few extracts from my catalogue will be intelligible. Grave 0 2, 3 metres deep. 1. Rough red terra-cotta head of animal. 2. Bronze bangle. 3. Iron bangle. 4. Iron comb. 5. Alabastron. Grave C 3, 3|^ metres deep. 1. Black vase, ornamented with projecting dots. 2. Iron and bronze strigil. Grave C 3, 3^ metres deep. 1. Bronze mirror. 2. Minute gorgoneion. 3. Lecythus, with diamond pattern of black lines, and white spots, on red ground. 4. Minute black jug, with small neck. Grave 0 4, 3 metres deep. 1. Bottom of bowl. 2. Cowries. 3. Bone beads. 4. Shells. 5. Piece of lead. 6. Claw of lobster. Grave 0 4, 3 metres deep. 1. Alabastron. 2. Lecythi, plain. 3. Lecythus, diamond pattern, with white spots. 4. Black lamp. 5. Lecythus with head, and white touches added. 6. Black and red bowl. Grave D 1, 1 metre deep. 1. (Vases tapering towards both ends, rough 2 S red ware. 3 . ? /f'Alabastra. 4. > 5. Black vase, with white lines. 6. Smoothed tridacna shell. THE CEMETERY. 27 7. Shell. 8. Gryphons. 9. Bacrania. 10. Terra-cotta horns. 11. Gorgoneia. Grave D 2, 2 metres deep. 1. Large alabastron. 2. Lecythus, ornamented with projecting dots below, and painted with figures in white blue and gold, with relief (PI. XVI. 20). Examples like these might be indefinitely multiplied ; but enough have already been given to show the usual contents of a grave. They are disappointing, indeed, when compared with the expectations we might have had ; yet perhaps they are not without some value, as dating from a period of which very few remains were found in the town, and tending to show that it is not by a mere accident that almost all the things found at Xaukratis date from an earlier time. For those who could not afford to bury more than this with their dead, are hardly likely to have left behind them in their life many objects that would repay the search of the excavator. (28) It only remains for us to notice briefly any of the things found in the excavation of the cemetery that seem worthy of individual attention ; these we may classify under various heads; — 1. Vases; 2. Articles for personal use or adornment ; 3. Other furniture of the tomb. 1. Vases. It has already been stated that the cemetery yielded little or nothing to en- lighten us as to the great Naukratite manufac- ture of vases in the sixth century. Hardly any individual fragments of pottery have been found whifjh can with certainty be assigned to so early a date ; much less any perfect specimens of the vases of that period. One or two small vases were found which seem to go back at least into the fifth century; but even these were exceptions, most of the graves are not D much earlier than the beginning of the Pto- lemies. Among the vases or fragments which seem distinctly early were a small vase with red lines on a yellow ground ; some Cyrenaic fragments ; two large rough plates of yellow ware, of the same description as those found last year in the lower strata of the trench in the temenos of Apollo ; and a vase of very peculiar shape, like an askion with a vertical hole through the middle (PI. XVI. 19). A small urn with four handles and red ornamentation on a yellow ground also seems to be of early style ; it was found together with the contents of a later grave, but this was probably a mere accident. The urn seemed to contain ashes ; if so, it yields the only trace I found of the practice of crema- tion — a fact of some interest in one of the earliest interments. But the facts are not certain enough to be insisted upon. The urn was broken when discovered, and its contents may not have been human remains. Vases of the ordinary red ware, either com- pletely covered with black varnish, or with circles of the red ground visible, were common enough — among them some drinking cups of graceful shapes, canthari and cylices. A not unusual shape was that of a plain bowl, with two horizontal handles just beneath the rim. Three of these are of interest, two from in- scriptions incised on their bottoms, one for the figures upon its sides. The first of these bears the graffito A AT I on one part of the bottom, opposite it Aim, the other has HAAAA, and in the middle of the bottom, AAP : here it is obvions that we have numbers, and H seems the sign for drachmas, h, retrograde ; thus wo may read the first 16 drachmas, 14; the second 40 drachmas, 25.® The meaning of the figures can- not as easily be conjectured. Plain bowls like this could hardly have cost more than a drachma a piece, so that the number can hardly refer to * An Attic -writer, wishing to exjjress 16 draciimas, would have written API-. 2 ^from coffin. 28 NAUKRATIS. the lot of vases for which the sum was paid — unless, indeed, that lot contained some larger or more elaborate vases. Mr. Petrie made the in- genious suggestion that we may here have a record of the undertaker’s bill. Graffiti of this kind are not very common, and it is hardly pos- sible to determine their meaning in all cases ; often, perhaps, it was not meant to be under- stood by any but the writer. The third of these bowls is ornamented with two scenes, in the ordinary red -figured style of the fourth century : under each handle is a palmetto design ; on one side is a seated Satyr, holding a thyrsus ; before him stands a draped female, holding a cup ; on the other side is a hunter standing, with one foot supported on a rock; he holds two spears ; in front of him is another male figure, now much damaged, but seemingly of comic or grotesque character; the two suggest a resemblance to two figures on the Ficoroni cista, derived perhaps from the same source. Small lecythi were of course abundant ; these were either plain, or ornamented with a palmetto or other design : this was often a male or female head, in one case a crouching hare. Another common ornamentation was a diamond pattern of black lines crossing, with dots of white on the intersections. One lecythus (PI. XVI. 20) calls for especial mention ; it is the one already referred to in the catalogue of various graves. Unfortunately the mouth and handle are lost, but the rest is perfect : its body tapers towards the top, and its lower part, which is thicker, is studded with projecting dots, so as to give the whole a resemblance in appearance to an elongated acorn ; on the smooth part is a rich design ; beneath the handle is an elaborate palmetto, but in front is a scene painted with all the richness of polychromy. In the middle is Eros,® white, with gold and blue wings, mount- ® This scene seems to represent the gathering of tlie incense, according to a suggestion of Dr. Furtwiingler quoted in M. Froehner’s catalogue of the Exhibition of the Bur- ing a ladder, and holding a small censer in his hand ; on each side of him is a female figure, one seated, one standing ; their skin is white, their drapery blue and red, and gold ornaments are added here and there in relief. Various plain or rough vases were also found, some tapering from the middle towards both ends ; also some miniature amphoraB, one care- fully worked to imitate an almond in shape and appearance (PI. XVI. 15). 2. Articles for personal use or adornment. Here must be mentioned various articles of toilet; several bronze mirrors were found, and one mirror case, but all were either quite plain or had only the simplest ornamentation ; some combs came to light, both of iron and bronze, and two or three toilet pots, one of lead (?), two of pottery ; one of these still contained its rouge, quite fresh and ready for use ; on its lid was painted a tripod; strigils both of bronze and of iron were also pretty common ; there were also some long and narrow bars of bronze, terminating in a minute spoon-bowl, apparently medical implements (PI. XVI. 17). Among per- sonal ornaments the commonest were bangles, both of iron and bronze, usually of very small size, and some rings ; most of the latter were either plain, or too much damaged to retain any design ; but there was one exception. This was a bronze ring plated with gold ; on it was a beautifully executed intaglio, representing Eros crouching, and apparently holding a wreath on a stick — doubtless a scene from some popular game (PI. XVII. 7). Some remains of silver ornaments were also found, but in very bad condition ; beads were very rare ; here perhaps we may also mention some small bronze bells, with iron clappers (PI. XVI. 7) ; several of these. were found, but it is not clear whether they were attached to the dress, or buried with the deceased for other purposes. lington Fine Arts Club, 1888, p. 18. There two or three other instances of the same scene are quoted. THE CEMETERY. 29 3. Other furniture of the tomb. Alabastra were often found, both of alabaster, often so corroded by damp as to be mere skeletons, and of other materials made in imitation, whether terra-cotta or faience. In one grave were found two massive and shallow bowls of alabaster, narrowing at the top ; two or three lecythi of alabaster came to light. A leaden bottle too was discovered, and several pieces of lead ; also several large shells which must have been used for some practical purpose, probably to hold food or other necessaries. Two or three of these were plain tridacna shells. Here may be mentioned some curious minute saucers, sometimes with two handles, and lamps of various periods and shapes, including most of the prevalent Greek types. In one case an iron spear-head was found, but this was exceptional. Two or three graves yielded small amulets in bone or faience, representing the god Bes. Many large iron and bronze nails had doubtless once served to hold together the wooden coffins. In only two or three cases terra-cotta statuettes were found. The two best examples are re- produced upon PI. XVI. 16, 18. One is a single figure of a boy ; the other a group of Eros and Psyche, worked as in relief, and hollow at the back. On this some traces of gilding and colour remained. (29) But it would be tedious to carry this enumeration into further details. Enough has been said to indicate the nature of the objects discovered, and to show that they date mostly from a time when Naukratis was already declin- ing. But though we may not have gained from the cemetery, or this part of it, the results that might have been hoped for, it may perhap.s have yielded some results that are not without their interest. ao NAUKRATIS. CHAPTER III. TEMPLE OF THE DIOSCURI (30) The discovery of the temple of the Dios- curi has already been mentioned in the first chap- ter. Its remaining walls aod pillars, as well as its probable plan, are indicated upon PL I. The hatched parts represent the brickwork that was still extant, the outlines give what is an all but certain restoration. Walls and pillars alike were built of unbaked mud-brick — a dark structure, hardly now to be distinguished from the dark mud that surrounds it, but for the divisions often visible between the bricks and courses. This brickwork must always have been unsightly, and so we find it to have been covered by a coat of plaster. On the pillars in front, this plaster was still in its place, and presented a plain white surface to the view : the stucco had, however, peeled off the inner walls of the cella of the temple, but was found in considerable quantities at their foot, lying upon the floor of the building. To this stucco we must again refer, because of the colour that remained upon it : but first it is necessary to say a few words as to the plan of the temple. A glance at the plate will show that it consisted of a single chamber, and was a variety, though a peculiar one, of the ordinary temple “ in antis.” Only a small portion of the cella walls remained ; that on the north was almost perfect, being only broken by a gap formed in sinking a well at a later period. The east wall could not so easily be traced ; the portion marked as extant upon the plan was extremely difficult to distinguish from tho surrounding mud, and may possibly be inserted by a mistake. No trace of a door or of a second chamber behind was to be found, but this too is a fact that must not be too strongly insisted on, in consideration of the difficulty of distinguishing wall from mud in this region. The southern wall has entirely disappeared, and has been assumed to correspond exactly to the northern one. Of the pillars on the west three remain in part; and the extant portions exactly correspond in their size and position to the requirements of a symmetrical portico of four pillars. If the pillars themselves were 34* inches by 17, and the intercolumniations five times the breadth of the pillars, or 85 inches — measui’ements which are as nearly as possible those of the extant parts — we obtain a total width for the colonnade which is almost precisely that actually found by the most careful measurement.^ Hence it follows that all the essential parts of the temple must have been as indicated in the plan ; and though this plan is a very peculiar one, it does not admit of much doubt as to its characteristic features. The temple is, as has been already said, “ in antis; ” but the anta3, or ends of the cella walls, are not, as is usual, in a line with the front row of pillars, but project considerably beyond them. Again, these are oblong pillars, not round columns — a peculiarity due doubtless to the nature of the material that is used ; then the arrangement also of the pillars is curious ; the one at each end has its outer side contiguous to the cella wall. Thus the projecting antse, doubt- less covered by a gable roof, enclosed a sort of ' Or perhaps 28| inches at first ; in the only complete one, a layer of plaster was visible at this distance from the front. ^ The measurements and levels of the temple were taken by Mr. Petrie before he left ; I also verified them afterwards. TEMPLE OF THE DIOSCURI. 31 pronaos, separated by the row of pillars from the cella inside. It is worthy of notice also that the temple faces westward, as was usual in the case of buildings dedicated to demigods or heroes like the Dioscuri : while the temples of the gods usually opened towards the east. (31) The fragments of stucco from the inner walls of the cella were of great interest, because they bore upon their surface, in wonderfully good preservation, the fresco painting that once adorned them. The colours used were very brilliant — red and blue on a white ground here and there, perhaps, a little yellow ; but this may be due merely to the discolouring of the stucco. The designs consisted exclusively of decorative patterns ; and though no fragment was found that was large enough to admit of a complete restoration of the patterns used, it was clear that most of the pieces came from a mseander design, varied apparently with stars within squares — a practice not uncommon in early Greek decorative work. (32) Beyond the plan of the temple itself, the second season’s excavations made but a small addition to the number of the objects found in the temenos of the Dioscuri. No more fragments of the fine dedicated bowl, found in the previous season, and reproduced in Naukratis I., PI. VI. 6, were recovered. Nor were many other important pieces of pottery to be seen ; one, however, is interest- ing ; it represents men seated in a boat upon a sea of white waves, and above is a piece of a wing; it seems probable that the subject of the representation was the sirens singing to * I am indebted to Prof. Middleton for the following facts as to the pigments used, &c.: “ The blue is made of powdered glass coloured with some salt of copper and then mixed with lime. The red is a pure oxide of iron. The stucco is the most wonderful stuff I ever saw — very much harder than English Portland stone. It is made of about .3 parts of lime to 1 of finely-ground silica or quartz — probably local sand ; it has been made with extremes! skill and care. None Like it could be made now.” Odysseus. The design is black on red ; but the types of the faces are curious ; the treat- ment of the waves with a wash of thick, white pigment in wavy lines, is most peculiar. So far as I know, this fragment is quite isolated in style, and it merits more attention than can here be bestowed upon it ; it is to be seen at the British Museum. The style of this piece can hardly be later than the sixth century ; other fragments of the same period came to light ; but none of them call for especial notice. Several more inscriptions record dedications to the Dioscuri, thus proving, if more proof be needed, that the temple and temenos do belong to those deities. One other inscription, which was incised upon a piece of rough pottery, is of interest, as it contains a portion of a dedication to Apollo. It was evidently apiece of a broken vessel that had wandered a short distance from the neighbouring temenos of the Milesian Apollo ; but as it was buried in undisturbed earth, at a depth of some eighteen inches below the bottom of the pillar that was nearest to it, it clearly must have been buried there before the temple of the Dioscuri was built. (33) This fact leads us to the last question we need consider in connection with this temple — that of its age. There is very little definite evidence afforded by the objects dis- covered ; many of them may not have belonged to the actual building which was found ; the frescoes can hardly supply any indication of period. There is, however, no reason for supposing this temple to be of a very high antiquity. The fragment from the temenos 6f Apollo just referred to seems to indicate that the temple of the Dioscuri is later at least than the earliest one dedicated to Apollo. The plan and construction of the temple, again, though it is peculiar, is not necessarily of very early date. On the whole, there is not evidence enough to outweigh that afforded by the level of the foundation, which can be compared with 32 NAUKRATIS. that of the Temple of Apollo and other neigh- bouring buildings ; the ground on which they all were built was probably quite level origin- ally. The bases of the pillars vary from 310 to 316^ inches above Mr. Petrie’s datum, the north wall was 303 inches, but its foundation may have been a little deepened for support. It would thus appear that the temple of the Dios- curi, as we found it, was nearly contemporary with the second temple and temenos-wall of Apollo,^ If this be true, it dates from about the middle of the fifth century b.c. ; from the time, that is, when Naukratis began to revive from the calamity that seems to have befallen it near the end of the sixth century. The archaic pottery and the earlier dedications must in that case be regarded as survivals from an earlier temple that once occupied the same site. Thus, though the evidence here is scantier, we may suppose that the Dioscuri as well as other divinities at Naukratis had in their temenos successive temples, built in different periods of the city’s growth and prosperity. ^ Nauk. T., p. 16, PI. IV. CHAPTER IV. TEMPLE AND TEMENOS OP APHRODITE. (34) In the first chapter an account has already been given of the way in which the site dedicated to Aphrodite at Naukratis was discovered ; how it was first identified by the fortuitous discovery of inscriptions, how the walls of the Ptolemaic temple were then recog- nized, and how finally two earlier foundations were found beneath them, and the boundaries of the temenos itself could be traced. We must now reverse the order of our description, and endeavour to sketch briefly the architectural history of this ancient sanctuary, beginning from the lowest and earliest level. Fortunately the materials for such a sketch are in this case exceptionally complete ; upon Plates I., II., III., they will be found in a convenient form for reference. Plate I., in its upper part, contains a general plan of the temple and temenos. In this plan the various levels of the foundations of’ the various walls are distinguished by different manners of hatching, and thus it is easy to see at a glance what parts of the building are of the same level, and, therefore, probably contemporary ; walls reconstructed at various levels are indicated as belonging to the lowest level at which they occur. Plate II. is a plan on double the scale, showing in detail the various levels of construction that are found in the temple itself. Plate III., on the same scale as Plate II., gives the sections along the two lines drawn on Plates I. and II., from W. to E., and from S. to N. On this plate are indicated, beside the walls, the places where the stratum of pottery and other fragments was found. In it is also added, at the sides, a scale of levels ; these are taken from the arbitrary datum fixed, as a convenient one, by Mr. Petrie last year — a point 500 inches below a certain conspicuous platform of Roman brick on the east of the cleared part of the ancient town. Thus Plate III. may be compared with Plate XLVI. in Naukratis, Part I. ; but in comparing the actual levels we must take into account certain considerations that will afterwards be noticed. After so much explanation of our plates, we may now proceed to the history of the temple itself. (35) The earliest temple of Aphrodite and the great altar that stood in front of its eastern door were founded upon the hard mud that everywhere underlies the town of Naukratis. This mud is always reached by the excavator as soon as he has cleared away all the artificial accumulation that generations of occupants have heaped above it. Its surface is repre- sented on the section by a thick black line. If its level, as here indicated, be compared with that of the original surface of the ground in the temenos of Apollo, Naukratis I., PI. XLVI., a difference of about twenty inches will be observed ; but this is hardly more than may be due to an accident, though the ground was probably almost perfectly level before the city was built ; for the two sites are nearly a quarter of a mile apart. We have, then, every reason for believing that the temple of Aphrodite was one of the earliest religious foundations at Naukratis, or at least that it was the first building to occupy the site on which it stands ; and that site is so central that it can hardly have been left bare in the prosperous days of o4 NAUKRATIS. the city — certainly not when the Hellenion was founded, for it lies between that and the temenos of Apollo, in the most populous part of the town. We may, however, see reason hereafter to suppose that the southern part of the town, where the Hellenion lies, was of later date than the more northerly neigh- bourhood of the temple of Apollo ; but in any case the scarab factory and its surroundings, which are close to the temple of Aphrodite, are anterior to the reign of Amasis, when the Hellenion was founded. But these chronologi- cal questions must be considered together in a later chapter. Here we need only note that the only literary reference to the Aphrodite of Naukratis and her temple, that of Athenmus, mentions as a date the 23rd Olympiad (688 B.C.), and though this seems impossibly early, the whole story that he tells clearly does belong to an early period in Greek history. If we now turn to the extant remains them- selves, we may notice first that there are no dug-out foundations. Such are rarely if ever found at Naukratis for structures of mud-brick such as this temple. And the test of time has fully justified a proceeding that might at first sight appear to show lack of care in the builder. The earliest temple is a simple and regular oblong building, with a door in its eastern wall ; within, it is divided into two compartments, the cella or naos proper and the opisthodomus, by a smaller wall, having a door for communication in the midst. The position of this doorway, as well as of that in the east wall of the temple, could only be traced by carefully following''with a knife the edge of the plaster floor ; the earth that had filled it could not be distinguished from the mud-brick wall on either side. The walls were covered within by a thin coat of plaster ; on this no signs of decoration, either by painting or otherwise, could be seen : but the plaster itself remained over a considerable portion of the surface of the walls. A similar but somewhat thicker coat of plaster formed the floor of the temple. At some period not very long after the first building it was found necessary to raise this floor ; and accordingly a second horizontal layer of plaster is found at a level about a foot above the first. This later floor extends also over the part of the precinct in front of the temple, and between the temple and the great altar. This altar consists of a thin case of mud- brick walls, filled inside with ashes, doubtless those of victims that had been burnt in sacrifice to the Goddess Aphrodite. Thus it reminds us of the great altar of Zeus at Olympia, which was also constructed of the ashes of victims ; it is cased outside with a double layer of plaster, dating from two differ- ent periods. In front of the altar is a flight of steps leading up from the temple, and on each side is a small wing ; these are all of mud-brick faced with plaster; the altar itself is sur- mounted by a small cornice. But the altar and its appurtenances cannot be thus simply described. For we may observe various stages in its history and construction which, if not very interesting in themselves, are of import- ance as throwing some light on the history of the temple and its precinct. It is clear from the plan that the altar was completely buried before the second temple was built, but we may notice four stages in its previous history. 1. The altar itself and a flight of three steps leading up to it were built, resting on the basal mud ; at the same level is also a founda- tion of the south wing : the whole was coated with plaster. 2. Fifteen inches higher (level 315) ; when the temple had a new floor given to it, that floor extended over the precinct in front also ; and on it was founded a new flight of four steps leading up to the altar. 3. Seven inches higher still (level 322) ; when the lowest of the four new steps was buried in the accumulation, the south wing was TEMPLE AND TEMENOS OF APHRODITE. 35 rebuilt, and a north wing also was added. ^ At the same time the altar was covered with a new coat of plaster above the old one. 4. On both wings later constructions of mud-brick were added, to raise the level of their tops. These facts seem to show that the accumula- tion of the earth in the temenos round the altar took place gradually, and that it had to be altered from time to time in consequence. The walls of the temenos, as we shall see, point to a similar conclusion. The question is a difficult one, and can hardly be settled apart from a consideration of the various strata found in excavating the site. The shape of the temenos is irregular, being probably determined by the boundaries of other properties or by roads. On the north, south and west, and also on the northern part of the east side, the wall that bounds it seems to- have been built at the same time as the temple itself, and to have remained through the later periods ; this may clearly be seen by a glance at the plan (PI. I.), and sections (PI. III.). Another part of the enclosure wall on the eastern side was built at a later time, but apparently before the foundation of the second temple. In the northern part of the enclosure were two wells, constructed of cylinders of pottery, ten inches in height ; these wells descended to a depth of twenty feet and twenty-three feet respectively : there were also in each case holes in the cylinders, to facilitate descent. In the first well, which was thirty-five inches in diameter, there were four such holes in the top cylinder, two in the third and fifth, and so on. The second was thirty-three inches in diameter, and had two holes, about twenty inches apart, in each cylinder, but these were not arranged in ’ This north wing is hatched in PI. I. as level 300 — 315 ; its foundation is really 7 inches higher; hut it did not seem worth while to introduce a fresh notation for so small a difference, no others of less than twenty inches being indi- cated. E any vertical succession above one another. It was clear from the level of the wells that they both belonged to the first temple, and that in later times they were forgotten or lost, and buried in earth and refuse. Yet the second well as soon as it was cleared out yielded a plentiful supply of good water, which was used all the rest of the season both by my own workpeople and by others in the neighbourhood. (36) The problem that next meets us is one of great difficulty. On the section are marked the places where “fragments of pottery, &c.,” were found. They were mingled with. statuettes and parts of statuettes, ashes and bones, all in the utmost confusion. This strange mixture lay almost entirely in two layers, one just above the level of the earliest floor ; another just below the level of the floor of the second temple ; between the two layers was a thick stratum of dark sand ; the lower was muchthicker and richer than the upper. Both layers extended over the whole surface of the temenos, on the north, south, and east of the temple ; behind it, on the west, only a few insignificant fragments were found. I have already mentioned the indications of a gradual accumulation of earth that could be seen in the remains of the altar and the boundary walls of the temple precinct ; but the facts now before us can be explained on no such hypothesis. It is, in the first place, all but impossible that such a thick accumulation of rubbish of all kinds should be allowed within the temenos. Then no stratification, chronological or other, can be observed within the layers ; and no distinction in style or age of pottery and inscriptions can be drawn between the upper and lower layer. Again, fragments of the same vase or statuette were often found, some on the north, some on the south, some on the east of the temple. Such a scattering and confusion can hardly be explained by any other supposition than this ; the contents of the first temple must have been violently broken and 2 36 NAUKRATIS. thrown out into the temenos around it. Mr. | Petrie lias suggested to me an ingenious ex- planation of the sand and the upper layer of fragments ; perhaps when the second temple was to be built, the site was first artificially raised to the level of the surrounding district ; then all fragments that had remained within the temple on its destruction were cast forth, and mixed with the highest stratum of the filling in. The only objection to this theory is to be found in the indications of gradual accumulation we have just seen in the case of the altar and the precinct walls. But both processes, both the gradual and the sudden filling up of the site, must be assumed to some degree in different parts. It is in any case probable that the pre- cinct of a temple would be better preserved from gradual accumulation than the land sur- rounding it. For this accumulation arises mostly from the washing down of mud-brick walls — a process that would hardly be allowed free scope in such a place. The road that runs just in front shows by its regular stratifica- tion that it was constantly necessary to put on fresh materials, as the district beside it was thus gradually rising in level. As to the time when the first temple was destroyed and its contents scattered, no certain evidence can be produced. But it seems extremely probable that here, as in the temenos of Apollo, the Persian invasion marks the later limit of what is found, and that some calamity befell the city of Naukratis at that time from which it never completely recovered. (37) The town did, however, recover to some extent, as is proved by clear enough indications both in this temenos of Aphrodite and in other parts of the site. Here, iipon the top of the walls of the first building, partly perhaps ruined, certainly afterwards levelled for the purpose, a new temple was founded. This new temple was not of exactly the same dimensions as the old one, nor did it occupy exactly the same site. Its length and breadth were both alike in creased. Thus it projected both in front, towards the east, and more still behind, towards the west, beyond the lines of the original building. Though the breadth also was in- creased, the wall of the new temple was beyond that of the old one only on the north side ; on the south it was a little within the old limit. This was doubtless in order to bring the temple nearer to the middle of its temenos ; now that the walls had disappeared, there was no longer any reason to consult other considerations than those of symmetry. The division between the cella and the opisthodomus also underwent a modification. It was extant, when found, from the south wall to a point some two-thirds of the distance across the building, and there were no signs of a door in the middle ; hence the door must probably have been close to the north wall, where it is marked in the plan. The outside wall seems to have been continuous all round the back of the temple, and so the door- way marked in the plan is in the only place where means of access to the opisthodomus can have existed. Parts of the plaster that had once covered the walls and floor of the second temple were still visible, both inside and outside the building. Upon the floor, in the south-west corner of the cella, were found some accumulations of rubbish, including a disc of gold, some fragments of bronze and iron orna- ments or implements, and some pieces of blue paint, fallen from the walls. It will be seen that the great altar was com- pletely underground at the time when the second temple was built. Nor is there any sign that another permanent structure was built to take its place. This alteration seems to have involved some modification in the boundary wall on the east side of the temenos, two portions of which may be seen to belong to the same date as the walls of the second temple. One of these portions ends in a : square pillar, which most probably formed one TEMPLE AND TEMENOS OF APHEODITE. 37 side of the gateway that led up to the temple during this period. (38) The third temple, which rested ^ upon the walls of the second, differed in arrange- ment from its two predecessors; it only contained a single chamber, occupying the whole space within the outer walls. Thus its cella was larger than that of the two earlier temples, though its entire measurements are somewhat smaller than those of the second one ; it is, indeed, intermediate between the second and first temples in size. Some pieces of wall to the west of the sacred precinct seem to belong to the same period as the third temple, but they are hardly massive enough to be a part of the boundary wall. At the back of the temple it is hard to say whether that boundary wall still existed, but on the north and south the earliest wall still rises here and there above the level of the ground on which the third temple was built. In the temenos were found fragments of a plaster floor, some belonging, from their level, to the period of the second temple, some to that of the third. They seem to show that the ground all round the temple rose gradually, and that the floor of the temenos had to be repeatedly renewed. A gradual accumulation of the ground during the later period also is unmistakably shown by the road, which appears on the extreme east of the upper section in PI. III. It consists entirely of uniformly stratified layers, and must constantly have been raised, in order to be kept at the same level as the district around it, which at every shower received a fresh deposit of mud washed down from the mud-brick walls of the houses. (39) Now that we have reviewed successively the three temples of Aphrodite and their ^ On the N.E., the face of the wall of the third temple is continued in front of that of the second temple, and appa- rently helow what was then the ground level ; it is not easy to see the reason for this arrangement. appurtenances, it remains for us briefly to consider their relations to one another, especially from the chronological point of view. We have already seen reason for believing that the first temple of Aphrodite was among the earliest built at Naukratis. If we think of other periods that we know to have been marked by activity in the rebuilding of old sanctuaries and the founding of new ones, two occur apparently to us as the most prominent — the close of the fifth century, when the temple of Apollo was rebuilt and that of the Dioscuri was probably constructed in its present form ; and the period of the earliest Ptolemies, when the great repairs and altera- tions in the Hellenion and its neigbourhood were made. These dates would then at once suggest themselves as probable for the founda- tions of the second and third temples respectively in the temenos of Aphrodite. And the relation of the various levels is at least not inconsistent with the suggestion. Even if the rate of accumulation in the temenos itself was not regular, that of the district around may well have been so, as is shown, for instance, by the road; and each temple would in all probability be adapted to the level of the ground around it at the time when it was built. Now at a rate of accumu- lation of about forty inches a century, we obtain an interval of about 200 years between the first and second temples, of 100 between the second and third. If, then, we assign the foundation of the first temple to about 600 B.C., that of the second to 400 b.c., and that of the third to 300 b.o., we shall probably not be very far wrong. Of course these dates must be regarded as only approxi- mate ; there are not sufficient data to preserve us from an error of a few years on the one side or the other ; but they will be a help to our memory in the endeavour to realize the various periods through which the city and its temples have passed. :38 NAUKEATIS. CHAPTER V. FOTTBRY FROM THE TEMENOS OP APHRODITE. (40) The site upon wliicli the pottery was discovered has been described in the last chapter; and upon the section, PL TIL, are marked the various strata in which it was found. The circumstances under which these numerous fragments of fine pottery and other objects once dedicated in the temple came to be buried in the place where they were found, have already received some discussion. A few words may be added as to the condition of the vases or of their fragments. All were com- pletely broken to pieces, and the pieces were in almost every case scattered amongst all the other refuse over the whole surface of the temenos. Thus the fragments belonging to each individual vase could only be recovered by gradually sorting the whole mass according to style and fabric. Many of the vases possessed, fortunately, a sufficient individuality of treat- ment or colour to make it possible to recognize their pieces at a glance. But this was not always the case, and therefore a systematic sorting was necessary. By this process were separated more than 150 vases, each repre- sented by a number of fragments varying from two or three to seventy or eighty ; in this total fragments are not included that remain isolated, and are only of value as fragments. To take one or two examples : the large bowl repre- sented on PI. XL 3, is made up of sixty or more pieces ; the plate, 1 and 2 of PI. XI., of 46 ; the bowl with a lotus pattern, PI. VII. 5, of 47. These are instances taken quite at random, and are fairly representative of the condition in which all the vases were found. Fortunately the colouring has in most cases suffered but little, and thus it is possible to restore in imagination the appearance of these vases, though they are so much mutilated. (41) The following classification of the various styles and fabrics of pottery found in the temenos of Aphi’odite at Naukratis cannot claim to be exhaustive ; if it were so, it would necessarily be so overloaded with detail as to miss its intention, which is to give a general outline of the various kinds of pottery and their distinctions and characteristics. Even in the classification as now given I fear that some assignments may seem somewhat arbitrary ; I may even in some cases have been led by characteristics that are accidental and not essential to the vases in which they occur. But my divisions are at least the result of a long and careful study of the pottery brought from Naukratis, and a familiarity gained by constantly sorting and turning over the fragments ; and the difficulty of m.y task must serve as an excuse for any errors or inadequacies that may occur in my attempt to carry it out. After so much apology, I will proceed at once to the classification itself, which may be followed more easily with the help of the table at the end of this chapter. A. White-glaze type (Naukratite). All examples of this class are dis- tinguished by the pure creamy colour of their glaze, and also by its fragility ; it will peel off in flakes ; on this are painted ornamental designs in brown, varying to light red ; purple and sometimes white POTTERY FROM THE TEMENOS OF APHRODITE. 39 touches are added in the more developed examples; once or twice, too, a pink flesh colour for the skin of men ; as a rule, the drawing is in fine outline, and no incised lines are used ; two exceptions will be noticed below. a. Bowls of the shape indicated in Nauk. I. PI. X. 1 and 3, a rounded body and a long conical rim, which is, indeed, the chief field for ornament. The body is decorated outside by plain brown bands, and at the top by a line of guillauche pattern. These vases are always black inside with lotus patterns, rosettes, &c., in red and white, varying in gorgeous- ness and complexity, but always of the same nature (see Nauk. I. PI. V. 1 — 10. Journ. of Hell. Stud., 1887, PI. LXXIX.). This ware is undoubtedly of local manu- facture, as is proved not only by the quantity of it found here and nowhere else,* but by inscriptions painted on it before firing, showing it to have been specially made for dedication to Aphro- dite. 1. Small, only ornamented with concentric bands and geometrical patterns ; inside as described. (Nauk. 1. PI. V. 29, 30). 2. Usually somewhat larger, with animals and figures (PL V. 2 — 6 ; Nauk. I. PI. V. 11 — 28), Style similar to next class. 3. Similar, but very fine ; some examples must have been fourteen inches or more in diameter at the top when perfect. Thus they approximate to the ordinary crater in shape and size (PI. V. 7 ; Journ. of Hell. Stud., 1887, ' PI. LXXIX., where the inside and out- i side of one of the finest fragments are * There are two .small examples in the Louvre, from Rhodes ; we might expect some exports thither, since so many vases were imported from that island. excellently reproduced in the colours of the original). The empty spaces of the field are always varied, but not filled up, with geometrical and other ornaments. The colouring is wonder- fully rich in its decorative effect. The drawing, on the other hand, though often very delicate and care- ful, seems to lack the vigour that marks some of the other types of pottery, and sometimes sinks to a purely conventional treatment. Thus muscles of animals tend to become meaningless spirals. So too the lion’s head on PI. V. 7, however fine, looks weak when contrasted with the powerful jaws of the beast on a different ware (B. h.) depicted also in PI. LXXIX. of the Hellenic Journal. Yet for beauty and richness of decorative effect one can hardly deny that this white glazed ware surpasses all others. The easy freedom of the brush seems, however, to have enervated its art, while the concise vinour of the incised line has left >=> , more strength to the drawing which we shall meet in class B. 4. Similar to 2 ; but generally smaller ; with incised lines. This treatment was especially used for grotesque figures of negroes, &c., (see Nauk. I. PL V. 34, 35, 40—42, &c.) If these figrures have sometimes more rude vigour of drawing, they entirely lack the fineness and delicacy that marks classes 2 and 3. In one or two exceptional cases we find incised lines combined with careful colouring, e.g. in the dark red heads, Nauk. I. PL V. 35 (one or two other similar examples have been found). A. h. Similar, but much larger and coarser vases ; only a few fragments have been 40 NAUKRATIS. found, whicli are not sufficient to justify j any conjecture as to their shape. The | ground is again white ; the drawings, often apparently of scenes of combat, ' including human figures, in brown out- lines, with the use, sometimes, of pink flesh colour. c. Small and delicate vases of various shapes, minute cups, &c. The ornamentation is similar to that of a 1 and 2. d. This is one of the most magnificent examples of ancient pottery found at Naukratis; one complete bowl has been recovered, and there are fragments of I a second (PI. VI. two views). The j bowl is of an open, basin-like shape ; it has two triple handles, each ! terminating in a human face at each end ; and between the handles on each side is a boss with two faces back to back. The glaze is a pure creamy white ; the design in light red or brown, with touches of purple. The outline drawing is similar in style and execution to that which we meet in A. /). This reason, as well as the fact that the glaze, colouring, and decorative effect are similar, have led me to assign this class to the type A. And it is hardly a rash conclusion that these bowls also are the work of the potters of Naukratis, and are among the master- pieces of their craft. It is to be observed that the designs on the fragmentary bowl seem simpler than those on the complete one ; and that the forms of the letters in its inscription of dedication are certainl}^ j earlier (see PI. XXL 701-705). It seems probable that it must have been an earlier experiment in the style of which we are fortunate enough to possess an almost perfect example in its later development : perhaps the dedicators of the two are identical ; but it is hardly probable that the dedicator was the potter himself, for he is most unlikely to have so disfigured his work by a carelessly incised inscription. e. Similar ware, with designs of similar style, is also found in other shapes : of particularly delicate work is the askion represented in PL Y. 1 ; we can hardly err in assigning this to the same local class of pottery. /. Of open bowls, slightly narrowing towards the neck, and with a flat horizontal rim (the shape of the bowl represented on PI. X.), some very fragmentary remains were found in this same white glazed ware ; the colouring and drawing were similar to that of d; among the animals on it we may especially observe a leopard, drawn only in outline, and his spots indicated by circles in outline, — a good instance of the preference found in this style for outline work. On the horizontal rim is often a guillauche pattern — a favourite design throughout this type. g. The next class bears a strong resemblance in many ways to a. 4, which we have already considered ; its representative is the cover apparently of some vase, figured on PI. VII. 2. This and a. 4 are distinguished from all other examples of this type by the use of incised lines. Together with them we also find a vigour aud humour in the treatment of the subject represented, here sphinxes and lions ; the little lion playing with the sphinx’s tail can hardly be anything but a humorous touch. In this example the material seems closer and the glaze more durable than in most other specimens of this pottery ; but the style and effect will hardly allow us to separate it from our present type. POTTEEY FEOM THE TEMENOS OF APHEODITE. 41 h. Lamps (PI. VII. 3). Several lamps of this peculiar shape and construction were found : the example selected might almost as well be assigned to some other type ; but since one specimen shows the characteristic white glaze and guillauche pattern, it must certainly be assigned its place here ; and it seems convenient not to separate these peculiar lamps from one another. These lamps show no sign of use, and may have been purely ornamental ; it may even be disputed if they are lamps at all, but such seems the most probable intention to ascribe to their shape. We may most suitably mention here a bowl with a round bottom and horizontal rim, orna- mented only with parallel bands of red on a whitish ground ; on the brim is a pattern of cross lines, as on the “ lamp ” in the plate. The two handles are made of triple rings. B. Eye-bowl type (Naukratite). This type is characterized by the decoration of the interior of the broad basin-like bowls that it comprises. This is in every case, from the smallest and most insignificant to the largest and most sumptuous specimens, precisely similar — only in a few examples further ornamentation is added ; but we never miss the essential part. The whole inside of the bowls is covered with a hard and brilliant giaze, varying according to the firing from red to dark brown in colour, but very different from the dull black of type A. Unlike that, again, it is very durable, and never peels off. On this glaze are painted parallel bands in red or purple and white, at regular intervals from the centre to the edge. The glaze of the outside is of a brownish-yellow colour, only approaching in a few specimens the whiteness of type A. ; it also is quite fast, and never peels off the surface of the vase. In the ornamenta- tion of this outside we may distinguish various classes. a. The eye -bowl proper, so called from the frequent occurrence on vases of this type of a large pair of eyes, often with a conventional arrangement of spirals between them to represent a nose. 1. The shapes may be seen in Nauk. I. PI. X. 11. Another common orna- ment is a set of vertical lines close together (as in the specimen referred to). 2. Similar, but of more complex shape, with double or triple brim, and four handles, one above another on each side (PI. VII. 1). It is not impossible this vase may be a mere accident, or a caprice of the potter’s, who made two of these vases into one. In any case it is of great interest as bearing the dedication of the sculptor Rhoecus to Aphrodite ; a part of this may be seen near the bottom of the vase in the illustration (see also PI. XXI. 778). &. 1. A splendid series of large bowls was found which one might not, at first glance, be inclined to connect with this type. But the similar treatment of the interior seems to indicate that we have here a development, far indeed removed from the simple ori- ginal from which it is derived, but preserving the essential characteristics of the type. These bowls are deco- rated with friezes, partly of animals, partly of vegetable or merely geome- trical forms. The lotus is an especial favourite; of this a very graceful variety, combined with an inverted palmetto, may be seen on PI. VIl. 5. F 42 NAUKRATIS. The peculiar combination of spiral volute and lotus which may be seen beside the handle of all these bowls was even last year described by Mr. Petrie as especially belonging to Naukratis. For last year many frag- ments of these bowls were discovered (Nauk. I. PI VI. 3—5, Xiri. 2, 3). The animals represented are lions, leopards, boars, stags, ibexes, and birds ; sphinxes and human-headed birds (called harpies or sirens in later applications of the form) also occur. For style the best examples are PI. IX. 1. (which though belonging to the next class, is similar to this outside), and the bowl reproduced in the ori- ginal colours on PI. LXXIX. of the Hellenic Journal. The two lions and the stag between them on that plate are a very powerful piece of drawing and composition ; though there may be errors, as in the foreleg of the stag, there is a vigour and force in the conception and execution which forms a great contrast to the delicate but conventional work of type A. And it is a fact not unconnected with the style, that incised lines are always used in the drawing of animals on the upper frieze of these vases, though never upon the purely deco- rative and ornamental designs that occur in the lower ones, even though these include ibexes. Fortunately we possess a clear proof that these large bowls are the work of local artists. On one of them is an inscription, painted before firing in large white letters on the dark glaze inside the rim, recording its dedication to “ the Aphrodite at Xaukratis ” (PI. XXI. 768). It there- fore follows that to type B., as well as to type A., we must assign a local origin ; they seem to show the deve- lopment of two simple indigenous styles of fabric and decoration ; this conclusion is strongly confirmed by the fact that neither type was known * before the excavations at Xankratis, and that to the two belong the greater number of all the fine vases found on the site. 2. This differs only from 1 in the richness of the interior decoration, which is represented on PI. VIII. 2. The principal designs are again drawn from the lotus ; between them are sets of vertical bars which again recall the ornamentation on the out- side of the eye-bowls, and so afford a further confirmation of identity of type in all these bowls, from the smallest to the largest. The bowl on PI. VIII. is no less than twenty inches in diameter, without including the handles. C. Black and red ® type. Here we descend at once from the new and characteristic vases of Naukratis to those which it exhibits in common with all other Greek sites. Some examples may call for special notice ; but there is no reason here for assuming any peculiar local manufac- ture. a. Simple bowls, usually of the cylix shape. 1. Plain. 2. Decorated with a series of concentric bands inside, of black and red alter- - See, however, the exception mentioned under type A, p. 39. 3 I use the word “red ’’here in preference to “buff,” which is used by Mr. Petrie to denote the same colour in Nauk. I., because the names “black-figured” and “ red -figured ” vases are terms of universal use and acceptation. It must be borne in mind that when I speak here of “ black and red ” vases, I mean precisely the same as Mr. Petrie when he speaks of “ black and buff.” POTTEEY FKOM THE TEMENOS OF APHRODITE. 43 Dating. The bands vary in width from an inch or more to the finest lines that can be drawn ; sometimes they are on all except the rim, some- times on the rim only. 3. The rim decorated outside with a con- tinuous wreath, usually of olive, in black on the red ground. 4. Various other decorations ; among the most peculiar are small applied astra- gali and ram’s heads (Nauk. I. PL XIIL 1). h. Painted with various animals and human figures, in black on the red ground, in friezes round the vase ; touched with purple and sometimes with white. Of this almost universally found type of pottery two classes are worth especial mention. 1 . Bowls with a flat horizontal rim, usually with no stand, but a spherical bottom ; the rim decorated with rosettes. (PI. IX. 6, 7.) The subjects are men, human-headed birds, leopards, boars, &c. Incised lines are used ; the field is left quite plain. The man holding a staff or caducous, facing one of the human-headed birds, is interesting, as an example of the probable prototype of the group later appropriated to the scene of Oedipus and the Sphinx. Harpy, Sphinx, and Siren had not pro- bably yet assumed the distinct forms in which they are known to later art. Those forms are certainly unknown to early mythology. 2. This class is similar to the last, but that the vases it contains are jugs, of the oenochoe shape, and not bowls : the j friezes of animals are of a similar i nature in both cases. ! c. Of the same colours, but larger and j coarser, and with designs also of a I larger and rougher style. The design is usually in black on a square red panel left clear of the black ground ; the vases are sometimes large bowls, sometimes jugs (PI. XIII. 2 ; IX. 5, with a large ram; this comes from the bottom of one of the wells; cf. Nauk. I. PI. VI. 1, 2, also from the bottom of a well — a likely place to find jugs). d. Small and delicate vases, often of extreme beauty in detail (VII. 4 ; IX. 8). No further classification of these seems to be of much use. The examples given will speak for themselves, especially the first. Throughout, in speaking of a type so common as this, I have made no attempt at an exhaustive enumeration, but have merely selected a few interesting or peculiar examples. D. G-eometrical, of very early style. One ex- ample of this was found at the bottom of a well ; it was a large jug or hydria, of light-grey ware, with only the simplest ormentation in cross-lines in light red. Being of rough and coarse ware, it may well have been used for ordinary pur- poses. But in any case the appearance of a vase of so primitive a character among fine Greek pottery calls for some com- ment ; it would seem to show that such articles may have survived for ordinary rough use, side by side with the more artistic productions of the Greek potters. E. 1. Cyrenaic. Several specimens of this ware have been found at Naukratis during both the seasons of excavation : most noteworthy is the splendid bowl re- produced by Mr. Petrie on PI. VIII and IX. of Naukratis I ; another piece is figured in the same volume, PI. VII. 11, and two or three other vases, more or less fragmentary, were found in the temenos of Aphrodite. In Naukratis I. p. 53, Mr. Cecil Smith made the 41 NAUKRATIS. suggestion that it was not impossible that this peculiar style of pottery should be attributed, not to Gyrene, but to Naukratis/ But he did not then think there was sufficient evidence to justify our setting aside the accepted name and attribution; and we certainly have not now any more reason for re- jecting it. The peculiar fabric of these vases, and the style of their decora- tion, are not at all similar in their nature to those we find upon the vases that we know now to have been made at Naukratis itself The Oyrenaic ware is so peculiar, that any fragment of it can at a glance be picked out from the confused mass ; yet the total number recovered does not imply the existence of more than three or four vases in the temenos of Aphrodite — an impossibly small number, if the ware were really of local manufacture. On two of the Cyrenaic vases were inscriptions (PI. XXI. 766 — 767) ; on one was the name Negomandrus, on the other, probably, Philammon. This last is a name that is obviously fitting for a Cyrenean, when we remember the close relations between Gyrene and the cult of Zeus • Ammon, — “ Oraclum Jovis inter sestuosi Et Batti veteris sacrum sepulcrum.” Negomandrus also has a Gr^co- Egyptian look about it, that seems to imply a connection with either Nau- kratis or Gyrene : in several cases be- sides this there is good reason for supposing that it was customary for dedicators to offer the ware that was * I am sorry to see that this suggestion, which Mr. Smith himself would not now maintain, has been repeated with confidence in the continuation of M. Dumont’s work, “ Les Ceramiques de la Grcce propre.” See below, p. 51. made at their native cities. On the whole, then, while we have no farther reason for connecting this “ Gyrenaic ” ware with Naukratis, but rather evidence for rejecting such a connection, further confirmation is given to the theory which attributes its origin to Gyrene. F. Flat, plate-like, shapes. To make a class like this dependent apparently upon the shape rather than the fabric or style of the pottery, may seem inconsistent ; but the vases comprised in it will be found, for the most part, to have some character- istic in common besides their form. a. Slightly curving up towards the rim with or without stands (Pinaces). 1. Decorated only with lines, or with vegetable forms, especially lotus ; occasionally with a head of an animal in a square, among other decorative designs — the ordinary Asia Minor and Rhodian type ; red on light yellow, often with purple touches (Nauk. 1. PI. VII. 1—6.) 2. Much richer in decoration; especially in one very fine example, painted both sides, with concentric friezes of ani- mals, boars, leopards, ibexes, birds, and sphinxes : a very rich lotus pattern is also added. The style of this plate recalls that of some of the finest Naukratite vases, and suggests that this also may have been made in the town (PI. IX. 1 — 4 both sides of the same plate). h. The field quite flat, surrounded by a raised edge, which serves also as a frame to the designs. 1. Plain. 2. With various designs, mostly represent- ing animals, usually arranged in concentric friezes ; various fabrics and colours are found, sometimes the POTTERY FROM THE TEMENOS OF APHRODITE. 4.'i ordinary black and red, sometimes red figures on a wbitisli ground, some- times a very fine white pottery, from which the ornamentation has almost entirely disappeared . 3. With human figures ; brown on a white ground with purple touches : incised lines are used (PL XL 1, 2). The fio^ures are arranged in concentric friezes, and there are a few beasts, perhaps lions, as well as men. The dress of the men is very peculiar, apparently a mere loin-cloth closely tied in. The whole treatment seems to be more or less humorous, thus one of the men grasps the tail of a lion, which being by the design inverted, presents an incoherent effect. We may indeed compare this plate, though its execution is far superior, with the humorous figures .which we met with in class A. a. 4. And, both in colour- ing and design, it resembles strongly [ the local pottery of Naukratis. It is also similar to a neck of a large vase with a row of dancing female figures (PI. XIII. 1), which seems to be of the same fabric (see type H. h.) 4. In some cases the field is not divided into concentric bands, but is treated as a whole ; in that case the lower portion is sometimes cut off, as the exergue of a coin, and filled with a subsidiary design or merely with ornamentation. A very curious ex- ample of this maybe seen on PL XX., where the fantastic beasts in various positions occupy both parts of the field. But by far the finest specimen discovered was the plate with a single figure of a seated sphinx, PL XII. This is a plaque painting rather than j a vase design. It is executed with the j utmost delicacy and ease, in four I colours — yellow, brown, purple or red, and white ; these are the typical four colours of early painting, and we can hardly doubt that they were the four that characterized the technique of Polygnotus and other early masters. Here, then, we have an example closely approaching to a panel picture, showing us exactly how those colours were used. Perhaps the most re- markable thing of all is the use of touches of white to bring out the high lights. Unfortunately it is hardly possible to see this now, but when the plate was first taken out of the ground, such touches were distinctly visible in some places, especially on the front of the fore-legs and paws. The use of the other colours may be pretty clearly seen on the plate. The outlines are drawn in brown with a brush, but incised lines are also used, especially to indicate the plumage on the breast. Above the head of the sphinx two small holes were bored through the rim of the plate, clearly indicating that it was intended to be hung up, in all probability as a picture to decorate the wall of the temple. If so, we may with yet more certainty regard this plaque as affording us in- valuable information as to the style prevalent in the free paintings of the period — if indeed any existed in the sixth century which were not purely decorative in their subject and treat- ment. G. Ibex type. This title is in many respects unsatisfactory ; on some of the vases included under it the ibex is not found ; and the ibex often occurs on other pottery which, though someAvhat similar, has elsewhere been classified. But the 46 NAUKRATIS. animal is of such frequent occurrence on this type of ware that it seems to afford a characteristic mark by which we may indicate it. The ground is always yellow, the design is light red, shading to brown, sometimes with touches of purple. Sometimes examples of this ware approach very closely to A 6 and other classes, but the drawing seldom shows the same delicacy and finish. The origin of this pottery is very probably Rhodian ; it was well enough known before it was found at Naukratis ; there is one very fine example, a jug, in the Museum at Bulak (cf. PI. XIII. 3). The friezes of animals of this type of pottery are almost precisely similar to the lower friezes upon the large bowls, type B ; both are without incised lines, and so have outline drawing extensively used. a. With tiers of ibexes (PI. XIII. 3) ; this I regard as the typical form. This jug also comes from a well in the temenos ; hence it appears that it, as well as that previously mentioned, was for actual use, not only for ornament. h. The covers of some vessels ornamented with the same animal. These covers are also found with the common orna- mentation of Rhodian pinakes. Hence again it would appear that this type has its home fn Rhodes. H. Large bowls, with a flat horizontal rim ; the ground light yellow or whitish, the designs in red varying to dark brown, according: to the firinof. a. With friezes of animals (PI. X.) ; these are often arranged into splendid groups in the most conspicuous positions ; for in- stance, the serpent and the two cocks on the vase reproduced ; on the other side are two horsemen facing one another, an eagle flying between them. The lower frieze is merely a consecutive band of animals, mostly leopards and stags. Incised lines are used, but the slight lines visible round the contours of the horse on the left are only added in the plate, as the outline is very indistinct upon the vase. This is a type of vase of which I believe this one from Xau- kratis is the first example to reach the British Museum; but the Louvre already possesses a fairly numerous series. These vases are exhibited in Room 5 of the Musee Campana ; and are de- scribed as “ vases of Corinthian origin from Etruria.” On some of the later examples inscriptions in the Corinthian alphabet confirm this attribution. Some are precisely similar to the bowl on our PI. X. ; one has two cocks and a serpent, another two horsemen and a bird, just as this. The subjects are various, friezes of animals and men, scenes from life, banquets, &c., on the upper part, rows of animals below. This series almost merges in its latest examples into the ordinary black and red type. The field is rarely quite free from ornaments till the latest period; in this particular our vase is distinguished from them. It certainly in style and covering ranks among the earlier examples of the series ; and of these it is perhaps the finest, though it has been so much broken. The inscription recording its dedication by Philis adds to its interest (PI. XXI. 780). b. Larger, and richer in colouring, white and purple being freely used (PL XI. 3). This vase was nearly two feet in diameter. The animals, which are very bold and vigorous in their drawing, are lions, bulls, ibexes, human-headed birds, &c. Here, as in class B, the upper friezes have incised lines ; the lower and sub- POTTEEY FEOM THE TEMENOS OF APHEODITE. 47 ordinate ones, of ibexes only, have none ; this affords an argument for the vase being of local fabric. The handles are very peculiar in their arrangement, owing doubtless to the enormous weight and size of the vase they had to support. Similar in colouring, though not in size, and even more interesting in subject, is the neck with a series of dancing women (PI. XIII. 1). The drawing of these figures is in some respects like that of the dancing men on the large flat plate (PI. XL 1, 2), but the vase seems best to find its place here. But I am inclined to think it probable that both this class H. b. and also F. b. 3, were made at Naukratis ; the glaze on both resembles that on type A, which is undoubtedly local. I. A very fine and soft whitish-yellow pottery, with simple ornaments in a bright red glaze. The best specimen is a jug from well 1. The neck was red, and there was red ornamentation round its base, and also the base of the jug. The jug is also remarkable for being circular in the neck, not of the usual trefoil shape. The triple handle was very delicate, made of three pieces, which had broken off, and were each separately recovered. Similar to this jug in fabric and ornamentation was a small urn which was found in the cemetery — apparently containing ashes. J. Here also may be mentioned, though the shape of the vase from which they come is not clear, some peculiar fragments. In these the ground is a black glaze, laid over ordinary red pottery, and the design, whether of plant or animal forms, is painted on this ground in white. TJn- i fortunately it is impossible to add much i more to this brief statement, as the fragments found were very few, and in very bad condition. A large vase, de- corated in a similar manner, may be seen in the British Museum ; it can only be distinguished from “ Polledrara ” ware by careful examination. K. Black throughout, through the whole fabric of the ware, glazed or highly polished outside. a. Bowls with a horizontal rim (shape, pro- bably, as in Pi. IX. 6). b. Large conical stands, with raised bands round the base. c. Tripods and other ornamental pieces, with relief ornamentation. d. Various small bowls, plates, &c. This black ware is of especial interest from the inscriptions that were found on it. (PI. XXI. 786— 793 ; cf. also PI. XXII. 840). As will be seen from the remarks on those inscriptions in Chapter VIII., it is extremely probable that they all record dedications by Mytileneans or Lesbians. If that be the case, it is hardlyrash to assumethatthe ware which they so consistently dedicated was a product of their native island. This is an assumption which can only be tested by excavation in Lesbos itself ; meanwhile we may hope that we have learnt some- thing as to both the pottery and the alphabet that belonged to the island of Sappho. This black ware, as has been pointed out by Mr. Cecil Smith, is similar to the “Polledrara” ware from Rhodes. But the “ Polledrara ” ware has not been found there in sufficient quantity to prove it a local fabric ; in both places it may be imported from Lesbos, Mr. Smith tells me he has found traces of colour on the black ground on this, as on the “ Polledrara ” ware ; but 48 NAUKRATIS. I have not detected them ; compare type J, and see below p. 51. T^. Dark rough earthenware, with a greenish glaze. One small bowl of this nature was found ; it probably is not of Greek make. M. Rough, unpainted pottery. This was of various shapes, mostly adapted to use ; bowls, cups, plates, &c. ; but no speci- mens seem to call for any especial comment except the lamps. Beside the ordinary types described by Mr, Petrie last year (Nauk. I. p. 45), were found several which must have had a very large number of wicks, and one of a peculiar construction. This was a shallow, circular dish, enclosing an open ; space in the middle ; round both the inner and the outer edge of the circular trough thus formed were a very large number of small holes pierced through an overhanging brim, in such a way as to support a continuous series of wicks round both margins. I have seen a j similar arrangement in a lamp preserved in the Acropolis Museum at Athens. Other specimens of the pottery discovered might here be quoted, but they would perhaps overweight an enumeration already sufficiently complicated. (42) We have now considered in detail the various types of pottery that are found at Naukratis, and especially in the ternenos of Aphrodite. We have seen which of them are probably, which certainly, the work of the potters for whom the city was famous. A few words must now be added as to the relation of these types, not only to one another, but also to other kinds of pottery that are found upon early Hellenic sites ; and also as to the probable age of the pottery discovered at Naukratis. Fortunately the first of these questions has already been considered by Mr. Cecil Smith, in his chapter on the pottery in Naukratis 1. But the far more complete and representative series of pottery from Naukratis that we now possess has supplied fuller material than was available for his discussion of the subject. In a subject so extremely complicated as that of the early history of vase-painting, it will be best, before we venture on any general sketch, to notice a few points that seem to throw light upon some difficult problems. It will already have been observed by those who re- member the distinction noticed by Mr. Cecil Smith (Nauk. I. p. 50) between what he then termed the “ Assyrian ” and the “ Egyptian ” classes of the “ Oriental ” style of vase-paint- ing, that our type A roughly corresponds to the latter, B to the former. In the “ Egyptian ” class Mr. Smith observes that “ no incised lines are used, and the artist is in consequence forced to leave portions of his figure in outline.” This is just the characteristic that we observed in class A, the white glazed Naukratite vases. In class B, on the other hand, which, as we have seen, is also of local manufacture, we always find incised lines, and what seems to be a development of the “ Assyrian ” ® style of vase-painting, in the principal friezes ; but in the subordinate or merely ornamental parts we have no incised lines ; sometimes we have decorative friezes of ibexes, on the lower and less seen parts of the bowls (PI. YIII. cf. also PI. XI. 3), which recall at once Mr. Smith’s “Egypt- ian style.” Thus these bowls offer at once a striking analogy to the vase published by him in the Hellenic Journal, VI., p. 186, with its two friezes in the two different styles ; and thus we have a confirmation of his view that such 5 I have adopted here Mr. Cecil Smith’s terms ; the nomenclature of this subject is already sufficiently confusing. Mr. Smith’s terms are, however, only used '‘for convenience,’' and he prohaldy would not insist on their scientific accuracy. “ Dorian ” is by Mr. Murray used as equivalent to “ Egyptian ” in this context, as Mr. Smith points out — a good illus- tration of the uncertainty of the subject. POTTERY FROM THE TEMENOS OE APHRODITE. 49 vases must be attributed to Xaukratis. In any case it is clear that Naukratis was not a centre of exclusively Egyptian influence upon the early art of Greece. "We must rather believe that in this colony the Greeks found many channels by which they derived the technique and subjects of their representations from various Eastern sources. Through their inter- course with the Phoenicians, whom they must have constantly met in their coasting voyages and especially in Cyprus, they must have felt the influence of Assyria as well as that of Egypt upon their art. Here, as in other cases, we shall not find them ready to borrow much directly from the ancient and stereotyped art of Egypt. Rhodes afforded still earlier a great centre of Greek and Phoenician intercourse. The influence of Rhodes upon Naukratis is very great. Rhodian pottery is constantly found in the Egyptian colony ; and it is often extremely hard to say whether a vase is a product of the local potteries, or an import from Rhodes. But Naukratis is far freer alike in subject and in style ; and we certainly find here a variety and an advance in artistic invention and execution which would make it an easy task to select the masterpieces of the potteries of Naukratis from among a number of other products of the earliest Greek vase-painters. As regards the period to which we must assign the vases found in the temenos of Aphrodite, it is not easy to come to a definite decision. But one thing is clear ; we have here no stratification, no gradual accumulation that began at one time, and ended perhaps a century and half later. All these fragments of vases and statuettes must have been cast forth at the same time ; though of course they may have gradually been accumulating in the temple from which they came. Many of the vases and of the inscriptions which they bear seem to belong to the same period as the later vases from the temenos of Apollo ; that is to say, to the latter half of the sixth century B.C. As has already been observed, all the circumstances seem to point to a violent destruction of the temple and its contents, and such a destruction is very likely to have taken place, in this temenos as well as in that of Apollo, at the time of the Persian invasion. If thus we suppose that the vases which have been discovered were for the most part made and dedicated during the half century which preceded the year 520 B.O., we probably shall not be far from the truth. And this conclusion, though it lacks the clearness of successive periods that could be attained in the temenos of Apollo, may yet be of great value to us in showing what styles of pottery were contemporary, and in fixing approximately the date to which they must be assigned. We must now make some attempt to estimate the position of Naukratis in the early history of Greek vase painting. We have here a dated series ; and we may divide the early Greek pottery as to which we may hope to learn something, into three classes. Firstly, we may consider such pottery as is absent from Naukra- tis, for this absence is in itself an important indication ; it implies that the types of pottery in question either were not being made at the time when this accumnlation at Naukratis was taking place, or were by some reason prevented from coming in any quantity to Naukratis. In the second place, we may discuss those types found at Naukratis which we know to have been made elsewhere, and to have been previously found upon other sites ; for we now have data for assigning a fairly definite and certain period to their manufacture. Then lastly must be mentioned those vases that seem to have l)een made at Naukratis itself, whether scattered examples have found their way by export to other sites, or we find in them types new to the student of Greek pottery. It is of course not necessary to enumerate all the primitive types of pottery found upon c NAUKKATIS. 5U Greek soil, iu speaking of those that are absent from Nankratis. That absence only becomes interesting when we reach the products of a period not far removed from those at which the temples of Naukratis flourished. Of Mycense vases we of course expect and. find nothing : and the geometrical style that follows them or is contemporaneous with them in Greece and the islands (“ Dipylon ” and “ Island ” style) also lacks any characteristic representative. But here the absence is not quite complete. A vase from a well in the temenos of Aphrodite, with a decoration of crossed lines (Type D), cannot fairly be quoted. It is quite isolated, and may perhaps be an importation from Cyprus, where this style of work lingered later than in Greece. But upon bowls of the eye-bowl class (Type B a. 1.), we sometimes meet with birds in Die panels between the groups of vertical lines almost precisely similar to those we find in the interstices of the geometrical patterns on the “ Dipylon ” and “ Island ” vases. Mr. Cecil Smith® is doubtless right in describing these “brown bowls” as “evidently imitations of the ‘ Geometric ’ style made much later than the archaic period to which they probably belong;” but the mere sur^uval of such forms is of itself enough to show that the interval in time between the two classes is not very great, or at least that no other different types inter- vened. Bor of course at such a period there is no question of conscious antiquarian imitation of earlier ware. We may then definitely fix the period of geometrical ornamentation as pre- ceding, but immediately preceding, the accumu- lation of vases in the temples at Naukratis ; and if we assign that accumulation to the sixth century, the later varieties of the geometrical style will find their place in the seventh. And when we remember that a geometric vase now in the Polytechnic Museum at Athens bears the earliest Attic inscription in- scribed upon it, we shall hardly be disposed to carry the latest examples of that style beyond the seventh century, for as yet we have no reason for assuming the use of the alphabet in Greece during the eighth century. All our facts indeed point to its being still in the transition of adaptation during the earlier part of the seventh century. The result, then, which we obtain in this case from the evidence of exca- vation is in accordance with our expectations, and affords a satisfactory confirmation to the dates we are otherwise led to assign to certain types of Archaic Greek pottery. We must next consider those types of pottery that, though probably made elsewhere, have been found at Naukratis. For we may now go on with some confidence, and assign them to the sixth century. Some of course may only have survived into this period, others may have only begun during its lapse a development to be carried farther in later times : of others, again, it may have contained both the rise and the declme. We need not spend much time over all these types of vases ; a mere glance at the table at the end of this chapter will suffice to show what they are. But one or two styles call for more notice, either because we have now something more to say about their origin, or because some mis- understandings have arisen as to their relation to Naukratis. These styles are the Lesbian, the Cyreuaic, and the Rhodian ; they have already been mentioned iu the classification of the pottery, but it seemed better to reserve for this place a somewhat fuller discussion of the questions to which they give rise. In Naukratis, Part I.,' are mentioned certain fragments in which the clay is black throughout; many more fragments of this ware and some almost perfect vessels were found in the temenos of Aphrodite (Type K). But an important difference must be mentioned; the fragments found in the temenos of Apollo had decorations ® NaTik. I. p. 49. 7 p. 40. POTTEKY FROM THE TEMENOS OF APHRODITE. in “a ricli red colour” and others still visible upon the black ground ; but those found in the temenos of Aphrodite were mostly plain, and in no case were the traces of colour distinct, though Mr. Cecil Smith tells me he has suc- ceeded in discovering them in one or two in- stances. It is probable, however, that in any case the pottery is of the same type, and that it is hardly to be separated from a similar ware found in Ehodes, and called by Mr. Smith “ Polledrara ” ware, because the finest example of black painted ware is a hydria in the British Museum from a tomb at Polledrara. But the style of that hydria seems to be dis- tinctly Etruscan, and to prove a local origin, the ware being the common Etruscan black fabric. If so, I am hardly disposed to adopt this name for the pottery of Rhodes and Naukratis. I have already given reasons which seem almost convincing, for believing that the black pottery found at Naukratis was made in Lesbos. The question arises whether we must suppose that the specimens found in Rhodes were also imported from Lesbos, or that the theory of Lesbian origin is in all cases a mistake, or that there were in early times two factories, one in Rhodes and one in Lesbos. I incline to the first of these three suppositions ; for I do not think the evidence for a Rhodian origin is strong enough to outweigh that of the Lesbian dedications ; nor on the other hand is it probable that there were two distinct factories for a ware so different from any other known iii this region and period. We may then, until we acquire farther evidence such as could best be supplied by excavations on Lesbos itself, regard this black pottery as a fabric peculiar to that island in the sixth century. The next class with which we have to deal will present more difficulties. It is commonly known as Oyrenaic, and I do not believe that we have any reason for abandoning that name. Such new evidence as we possess tends rather to confirm it, as has already been pointed out G in our classification (Type E). But it calls for especial attention both because of the sugges- tion already referred to, that it was made at Haukratis ; and also because of certain misap- prehensions that have arisen from its apparent similarity to the characteristic Naukratite ware, our type A. Since the collection and comparison of known examples made by Dr. Puchstein in the “ Archaeologische Zeitung”of 1881, this style has attracted considerable notice ; and we may in the first place notice that its occurrence at Naukratis is what we might expect from its usually accepted period, the middle of the sixth century b.c. But it is not found in any great profusion. Mr. Petrie discovered one nearl}’ complete example, and four or five others more or less fragmentary have since appeared. These, however, are no more than one would expect to find in a colony so near Gyrene, if the ware were Oyrenaic; on the other hand, they are far too small a number to find if we had really discovered the place where this pottery was made ; for at least thirty examples had already been found on other sites. The fact is that a misapprehension has arisen in some quarters, which is not unnatural in any who have only read descriptions of one of the two kinds of ware, but would at once be corrected by an examination of the vases themselves. All students of vases are now familiar with the characteristics of Oyrenaic ware; the usual, though not invariable, cylix shape, the hard and brilliant white glaze, the peculiar palmettos on the outside, the close and elaborate lotus and pomegranate patterns, the rich metallic ornateness of the whole effect, the frequent figure subjects. Now we do not meet with a single one of these characteristics upon our true Naukratite white-glazed ware. Here the shape is usually that of a diminutive crater with conical base, bowl-like body, and high conical rim, the chief field for painting; the glaze is flaky and crumbling, and of a dead- white colour outside ; inside it is a dull black ; the decora- o NAUKRATIS. tive patterns introduced show great variety in the outside borders, mostly of guillauche or maeander ; inside they are of lotus, in red and white on the black ground, but freely and largely drawn, not closely set together; the whole effect is rich indeed, but with the richness of its colouring, the designs being always sparse and floAving; and the subjects are always animals, more or less decoratively treated.® The only resemblance between the two wares lies in the similarity of colouring, broAvn and red upon a whitish ground. But I venture to think that no fairly trained student of pottery would ever mistake even a minute fragment of white Naukratite ware for Cyrenaic ; and this distinction of style sums up and is beyond all the detailed differences above enumerated. We may then dismiss at once the view that the Cyrenaic and white-glazed Naukratite are only variations of the same type of pottery. But one supposition still remains possible. We have seen that two types so distiuct as our A and B were both undoubtedly of local manufac- ture. May we believe that this “ Cyrenaic ” ware, though a different fabric, was also made at Naukratis ? I think that we must answer this question in the negative, for the reason I have mentioned above, the small proportion borne by the specimens discovered at Naukratis to the total number known. Till excavations can be made at Cyrene — perhaps the most difficult and most promising site still waiting to be thoroughly explored — we cannot finally determine that the name Cyrenaic is correct ; but in any case there is no reason for confusing * It is necessary to dwell upon a distinction apparently so oLvious, becairse tlie two are treated as similar, if not identical, in so valuable a work as Dumont and Chaplain’s “Ceramiques de la Grece propre.” This portion, written after M. Dumont’s lamented death, is the work of M. Dottier, who gathers his information from the descriptions of Mr. Petrie and Mr. Cecil Smith. Doubtless he rvould have written differently had he been able to examine the Naukra- tite ware himself. our nomenclature and our ideas by regarding this ware as another type of Naukratite pottery.® A few words must suffice here for the Rhodian pottery, or Camirus style, as it is sometimes called. This type (G) is often very difficult to dis- tinguish from the white-glazed Naukratite (type A). This difficulty is what we should expect, if, as has already been stated, the influence of this Rhodian style was paramount in the work- shops of Naukratis. I have already touched upon the difference between the two, which lies mostly in a certain refinement and delicacy of style that may be recognized with the eye more easily than described. The mention of the Rhodian pottery, of which that of Naukratis seems to be a develop- ment, leads us on to the last question here to be considered : the position of Naukratis in the history of Greek vase-painting. Here, of course, I would speak only of those types of pottery which were without doubt made upon the spot. It is only in very few respects that these form a link in the regular succession of development, joined to other derived types that follow, as well as to the earlier styles from which they originate. Mr. Cecil Smith ^ has already suggested that we must regard the vases with black figures on a white ground as probably due originally to the influence of Naukratis, and has assigned a similar origin to certain other characteristics associated with the name of the artist Nikosthenes. But in any case this influence only affects some details of ornament or technique ; it can hardly be traced in the style and charac- ter of the vases in question. On some sites ® asrain have been found later vases with an internal ornamentation of white and sometimes 9 In confirmation of the above views, it may be added that some fragments have been found at Naukratis which seem to be a local imitation of Cyrenaic pottery. '■ Nauk. I. pp. 51, 52. - E. g. on the Acropolis at Athens. POTTERY FROM THE TEMENOS OF APHRODITE. red also on a dark ground, that must be derived from Kaukratis. This is, however, again only an isolated trait. On the whole, we must acknowledge that the pottery of Naukratis does not so much represent a stage in that tran- sition from Oriental to purely Hellenic forms which is really the main subject of the early history of Greek ceramic art, as the most perfect and complete development of the decorative “oriental” style. From it we maylearn what this style would become in the hands of artists who were gifted with an exquisite delicacy of touch, and a wonderful feeling for colour and decora- tive effect, but who lacked that freshness and originality both of conception and execution without which Greece would only have imitated and carried to that perfection of which it was capable the art inherited from the East. We see at ISTaukratis a style of vase-painting not only beautiful in itself — far more beautiful OO than the uncouth attempts at originality that were later to produce such wonderful results — but also interesting alike in what it attains and in what it lacks. Had all Greek art of the same period been like that of the Haukratite pottery, we should indeed possess many attrac- tive works of a time that now has given to us more than we can find interesting for its O promise than admire for its intrinsic beauty ; but, on the other hand, without the uncouth originality that we here miss, the art of the sixth century could only have refined itself away into decorative detail in the fifth, instead of developing into the greatest that the world had ever seen. We may then be thankful for what is left us of the masterpieces of Naukra- tite pottery ; but we must also be thankful that they are not fairly representative of all the tendencies of Greek art during the period to which they belong. (For TuUe of Potterij, see next page.) TABLE OF CLASSIFICATIOK OF POTTERY Specimens reproduced. a. Crater-sIiaiDed bowls. ^ ^ 1 . Small, only decorative patterns ; lotus red and white on black inside. 2. Larger, with animals, &c., inside, as 1. 3. Very fine, similar. 4. Incised, men, &c., carica- A. Naiikratite , rvhite-glaze / Larger and coarser ware, figures, human and other, on the outside c. Small and delicate, various shapes. d. Elaborate bowls, painted on white glaze inside and out with animals ... e. Painting red on white, other shapes, askion, &c. f. Eowls with flat rims, similar designs. g. Harder glaze, light red, &c., on white, incised .A r>. Naukratite ; large bowls Lamps C?) and bowl, decorative patterns only . . . a. Eye-bowls, inside dark fi c- i glaze, with rings of white and red. More elaborate, but similar C. Black and red type IJ, Geometrical E. Cyrenaic... , T , , . , . , 1. Inside as B a h. Large bowls animals mi 3^ friezes outside. 1. Plain. 2. With concentric circles. 3. With ■wreaths and other ornaments on rim. 4. With raised ornaments. la. Common cylices r 1. Bowls with flat rim, black S insii I?;. With friezes of animals^ inside [2. Jugs c. Similar colours, large, designs in panel... ' d. Very small and delicate ... ... ' . . . a. Pinaces. E. Pinaces, &c. I h. Flat plates, with raised rim : ' 1. Plain, or simply decorated \ with floral and other de- signs 2. With ibexes, &c., on both sides / 1. Decorative designs, i 2. Figures and animals I 3, Brown on white glaze ; ' dances '' 4. Field rvholly occupied by single design, or divided G. “ Ibex ” type, Rhodian 11. Large vases with two handles horizontally, not concen- V trically ... ( a. Jugs ( h. Covers of vases. a. Corinthian, red and brown on yellowish ground, vaiying to black on red ... h. Very large, white and purple freely u.sed I. Pine bright red glaze on yellow ground, only decorative designs. •(. White painting on black ground. d a. Bowls. K . Black fabric throughout, ' b. Conical stands. probably Lesbian. } c. Tripods &c., with relief ornament. \ d. Various smaller pieces. L. Rough ware with greenish glaze. M. Rough red pottery, lamps, &c. Xauk. I. Plates. Xauk. II. Plates. V. 29, 30. V. 11-28. Y. 2-G. AL 7. V. 34, 35, 40, 41, 42, 54. AM. 1, 2. V. 1. V. 5G. AMI. 2. AMI. 3. X. 11. A"n. 1. AM. 3-5. AMI. 5. AMII. 1, 2 XIII. 1. vi. 1, 2. IX. G. IX. 7. IX. 5 ; XIII AMI. 4 ; IX. AMI. 11, VIII., IX. IX. 1-4. IV. 3. XI. 1, 2. xn.; XX. Xin. 3. X. XI. 3 : Xin. 1 55 CHAPTER VI. STATUETTES FROM THE TEMENOS OF APHRODITE. (43) All that has been said as to the circum- stances under which the fragments of pottery were discovered in the temenos of Aphrodite applies equally to the statuettes. These were scattered about in quite as much confusion as the vases themselves, though, from the nature of their material, they were not broken into so many fragments. But portions of the same figure were often found in distant parts of the sacred precinct. The statuettes show even more indications than the vases of an intentional destruction ; almost all the fractures, upon careful examination, show some signs of a severe blow upon one side. Hence it appears that they must have been violently broken up and then scattered — and by some enemy not only of the Greeks, but also of their religion. Such a proceeding is in accordance with what we know of the Persians ; and hence yet further confirmation is gained by the con- jecture, already well-founded, that it was the Persian invasion that led to a destruction of the temples at Naukratis, and gave a serious check to the prosperity of the city. (44) But at present we are not concerned with the destruction of the statuettes, but with their origin, and with the history and progress of the art that produced them. And it will be well at this point to quote a passage of Athenseus, already referred to more than once, which has a direct bearing upon the origin of the statuettes discovered in the temenos of Aphrodite at Haukratis. Athenseus, himself a native of Haukratis, quotes from a work of his felloAV-citizenPolycharmus, “ Ou Aphrodite,’' the following story. “ In the twenty-third olympiad, Herostratus, a fellow -citizen of ours, was on a journey ; and having sailed round many lauds, he touched also at Paphos in Cyprus ; there he bought a statuette of Aphrodite, a sjian high, of archaic style, and went off with it to Naukratis. Now, when his shipAvas near Egypt, a storm suddenly came on, and they could not see whereabouts they wmre ; so all of them took refuge by the image of Aphrodite, praying her to save them. And the goddess, with her wonted favour to the people of Nau- kratis, suddenly filled all the region about her with green myrtle, and made the ship full of the sweetest odour, when the crew had now given up hope in their severe sickness. They were at once freely relieved, and the sun shone forth ; so they made out their land-marks and reached Naukratis. Then Herostratus, rushing from the ship with the image, and also Avith the green myrtle boughs that had suddenly come forth, dedicated them in the temple of Aphrodite. And having sacrificed to the god- dess, and dedicated the image to Aphrodite, he called his friends and relations to a banquet in the temple itself, and gave to each of them also a garland of myrtle, to which he thereupon gave the name Naukratitc.” Such is Polychar- mus’ story, which is quoted to prove that the garlands called Naukratite were made of myrtle. This question does not immediately concern us at present, but neither the miracle recorded nor the probably erroneous date need distract our attention from the tradition that the tale 56 NAUKRATIS. embodies. Evidently the writer bad beard that “ statuettes of Aphrodite, a span bigb, of archaic style,” bad been brought from Cyprus and dedicated in her temple at Naukratis. Now this description exactly fits a large number of the statuettes that were found in the temenos of Aphrodite, and precisely similar statuettes are very frequently discovered in Cyprus. They seem, indeed, to have been originally the products of Phoenician art, and especially of that branch of Phoenician art which was established in Cyprus. But, although we may thus see the origin of the type of statuette we have now before us, neither the subject represented nor the style of the re- presentation can be thus summarily dismissed. For all who have studied the earliest stages of Greek art know that whenever the Greeks bor- rowed a form, they always infused into it new life and meaning. Brunn’s famous analogy ^ cannot be too often quoted ; “ from the Phoe- nicians the Greeks borrowed only the alphabet of art, as they did also of letters ; in both alike they spoke from the first in their own language.” But the tradition preserved by Polycharmus is of great interest, as indicating a fact that we learn also from a study of the statuettes them- selves; — the fact that even at Naukratis the Greeks did not borrow, as a rule, directly from the Egyptians, but that even there the influence of Cyprus was of the highest importance. To this question we must return after some more detailed examination of the statuettes them- selves. As to subject also, a simple and comprehen- sive statement is obviously impossible. It will perhaps be best to begin by distinguishing three types, the standing male type, the standing female type, and the seated type. Each of these will require a separate discussion, first of what seem the representative specimens, whether earlier or later in period, and afterwards of the many variations to which each has given rise. ' Dio Kunst bei Hoiuer. (45) The standing male type is represented by some specimens reproduced upon PI. I. in Naukratis I. Other similar statues were found during the later excavations, and may be seen in the plates of this volume (XV. 13). The type is well known, especially in Cyprus and Rhodes, where examples of it are found in abundance. But the fact is perhaps hardly yet clearly recognized, that from these primitive figures up to the magnificent athletes of perfect Greek art, we can trace an unbroken succession of type. This type is known by many names ; the small figures now before us are variously interpreted ; the life-size statues of the archaic period commonly go by the title of Apollo ; in the finest period they are generally regarded as athletes ; but it can hardly be disputed that the last is produced from the first by the long process of artistic evolution. But we are anticipating. The statuettes we are now concerned with show hardly any promise of such a future. Yet even among them we may perhaps distinguish various stages of de- velopment. Not that these stages represent of necessity a chronological sequence ; an earlier type often survives side by side with a later one which has been produced from it. Sometimes these figures are draped, some- times nude. In the former case they often have hardly any indication of the forms of the body ; the lines of the drapery are sometimes mere cuts at the place where any garment ends. This is an arrangement often found in Cyprus and Rhodes ; a specimen that exemplifies it well was left last year at Bulak (so too PI. XIV. I.). But such figures can hardly be said to belong to Greek art at all, though they show, perhaps, the material, without form and void, which the spirit of Greek art was afterwards to fill with life. We may see the beginnings of this influ- ence in two different classes of statuettes. In the one the drapery is discarded, and thus we have at once before us the type that afterwards was developed into the whole series of Greek STATUETTES FEOM THE TEMENOS OF APHRODITE. 57 Apollos and athletes (PL XIV, 13) : in the other, though the drapery is retained, the forms of the body are clearly felt and indicated. In PI. XIV. 14 may be seen a figure which shows at least some signs of this tendency. In the position of these figures and the objects which they carry we may observe a great variety even at this early period. Sometimes we find a musician, with harp (XIV. 14) ; we may compare this with the primitive figures of flute players that have been found in abundance, made either of glazed ware (PI. XVII. 4 ; cf. Nauk. I. PL II. 7) or of alabaster. Another common type is a hunter, who is represented as bearing his weapons and his quarry. This is clearly the meaning of the figure represented on PL XIII. 7 ; he holds in his hands his bow and his hunting knife, and over his shoulders are slung two hares and two young boars. On his thigh was once a dedica- tion, now unfortunately illegible, which doubt- less recorded his name (KaXXi. . . ) and dedica- tion to the goddess of himself and his spoils. It is perhaps doubtful whether we are to regard as similar in its intention such a figure as that re- presented in PL XIV. 10. Here the man, whose figure is very conventionally executed, without much feeling for nature, holds by its hind legs and tail a lion or leopard, which rests its fore-paws on his feet. A hunter may thus dedicate his spoil ; but perhaps this figure may have a mythological significance (cf, also Nauk. I. PL I. 1). (46) We must now pass on to the other main type, which is, as we might expect in this temenos, of most frequent occurrence. This is a draped female figure, holding some object in front of her breast with one hand, with the other supporting her drapery. Such is the normal arrangement ; of course many varia- tions in detail are to be met with. These female statuettes bear the same relation to the female statues in elaborate drapery found on almost all early Greek sites, as the male figures bear to the later ‘Apollos ’ and athletes. Of the other type, without drapery, no early example was found in the temenos of Aphrodite. In no instance is there any marked advance on the primitive statuettes of this simply draped class. Above the waist there is hardly ever any clear indication of the drapery, except sometimes round the neck. Beneath the waist there is sometimes a division between an upper and lower garment, usually in a curved fine ; often there are a number of vertical folds, either at equal distances all round the body, or drawn together into one broad fold, de- scending from the hand that supports it down the front. Examples may be seen in PL XIV. ; the last arrangement most clearly in 4. Hardly a sign is given in any case of the way in which the drapery has been put on ; it is no more than a thickness of some material, enveloping and obscuring to some extent the forms of the body, but with no form or independent existence of its own. These figures from Naukratis cannot indeed be distinguished in any way from the similar ones frequently found in Cyprus. Any of them might well be the “ statuette of Aphrodite, a span high, of archaic style,” which, as we have seen, Herostratus brought over from Paphos. Whether they really are statuettes of Aphro- dite is a question that admits of some discus- sion; but of this we must say more here- after. These figures in almost all cases hold some object in front of their breast with one hand. Sometimes it is a flower (XIV. 11 ; XV. 5); often an animal ; thus XIV. 12 holds a bird, XIV. 8 a small goat, XV. 1 (which was not, however, found by my own workmen in this temenos, but was brought me from another part of the town) holds a large goat with both hands in front of her body. (47) Various other types will be observed upon our plates, some closely allied to those n 58 NAUKEATIS. we have just been describing, some widel diverging from them. The most interesting of them are perhaps the seated figures, which again remind us, on a smaller scale, of certain statues that have been found upon Greek soil. One, XIV. 3, bears a strong resemblance to the Branchidse statues, “ si parva licet componere magnis.” But this seated male figure can hardly be the product of a different art from that to which we must assign the female figure with her child, XIV. 7. Here, however, the resem- blance to the well-known Egyptian Isis and Horus type is one that could not under any circumstances be overlooked — least of all in the case of a statuette found upon Egyptian soil. But whether even here the influence of Egypt is direct or not may be open to some doubt. The Egyptian origin of the type is, however, beyond dispute ; it is perhaps the most persis- tent type known. Its similarity is throughout obvious, from Isis to the Madonna of Christian art ; and a valuable link in the series is offered by this rough little statuette. More distinctly Egyptian in style is the little kneeling figure, XIV. 2 ; it is worked in low relief upon both sides. Two or three animals are worth notic- ing ; for instance, the limestone lion, XIV. 7, and the aquatic bird, XIV. 1. This is of terra cotta, painted with brown and purple, like an archaic vase. It resembles very closely in colouring the swans or geese which so frequently occur upon those vases ; but in its general proportions it certainly seems nearer to a duck. Some statuettes also served the purpose of vases ; XIV. 11 is merely an alabas- tron, shaped at the top to resemble a human breast and head. Head-shaped vases, such as that represented in XIV. 5, are well known in Cyprus and elsewhere. But no sign of a mouth or spout can be seen on this head, though it is made in two parts so that the top will lift off. The parts may, however, have been originally intended to be fastened together in some way. (48) If we now turn from these details to take a more general view of the statuettes that have been discovered at Xaukratis, there are two questions which at once meet us. In the first place, what is the meaning and intention of all these statuettes, male and female, dedicated in the temple of Aphrodite ? And then, how far are they to be considered as the products of Greek art, or as directly connected with it, and what place do they occupy in the great series of which we have already seen them to form a part ? The first of these questions cannot yet find a definite and conclusive answer. These statuettes cannot be disconnected from those found in the temples of Cyprus on the one hand, on the other on many of the most important sites in Greece, Branchidse and Delos and Athens. But one thing is clear. All these male figures and “ Apollo s ” and athletes, all’ these female statuettes and early draped statues of Delos and Athens and other sites, all these seated figures that have been found in or near various temples, must for the solution of this problem be considered as a single series. Various varieties and developments of meaning or purpose may belong to various regions and periods ; but these must not be regarded as independent, but as deviations from their common prototypes. So long as this great fact is lost to sight, no explanation or theory can be more than partial and inadequate. Without attempting to solve a question which has never yet found a complete and satisfactory answer, we may note one or two pieces of evidence that may guide us a little nearer to its ultimate decision. First we may observe that all these statuettes are dedicated to Aphrodite, yet many of them are male ; one is certainly a hunter and his spoils, and pro- bably bore his name, just as one of the Bran- chidffi statues states that it represents its dedicator. On the other hand, one of the female statuettes was dedicated by a man STATUETTES FROM THE TEMENOS OF APHRODITE. 59 Polyhermus (PI. XXL 794) ; and, if we trust the evidence quoted by Athengeus, “ span-high statuettes of Aphrodite ” were dedicated in early times. Thus we seem still as far from a conclusion as ever. Our evidence seems to show that statuettes were dedicated sometimes representing the dedicator, and denoting perhaps his devotion to the service of the divinity ; some- times representing the goddess herself. But we have as yet nothing whatever to guide us in any attempt to draw a broad distinction between the two classes ; and where the dedica- tor and the divinity are both of the same sex, the confusion in their representation is not as yet unravelled. As regards the style of the statuettes, it is clear at a glance that the influence of Cyprus is paramount ; and thus literary tradition is con- firmed by the testimony of excavation. But, as was to be expected at a colony on Egyptian soil, those Cypriote types which are derived from Egypt are the most readily adopted. It is hard to say whether we must allow that there was any direct influence exercised by Egyptian art upon the artists of Naukratis ; but in any case it was modified by the hybrid Phoenician character of Cypriote work. It is instructive to observe how far the art of statuary lags behind the sister art of vase-painting ; while there are many vases from Naukratis that for beauty of decorative effect have never been surpassed, no statuette has been found which shows more than a small advance upon the models from which it was derived. This fact is most interesting when we remember that it was from Egypt ^ that Rhoecus and Theodoras of Samos are said to have learnt the innovations which they introduced into the still primitive art of Greece ; and that Naukratis was the only town of Egypt which was in their time open to Greeks. Whether they studied the Egypto- Cypriote style prevalent among the artists of the colony, or went straight to the technical perfection of Egyptian statuary, is a question which can only be answered by con- jecture ; but in any case we now possess a most interesting reminiscence of their sojourn. In the temenos of Aphrodite at Naukratis was discovered a vase which had been dedicated to the goddess by Rhoecus — in all probability no other than this very artist, then pursuing in Egypt the studies that were afterwards to influence the sculpture of Greece. * cf. Mitchell, Hist. Anc. Sculp, p. 199. 60 NAUKRATIS. CHAPTER VII. TEMENOS OF HEKA. (49) The temenos of Hera, and the accident that led to its discovery, have already been mentioned. Herodotus had spoken of the temenos dedicated by the Samians to their goddess Hera as if it were as important as that of the Milesian Apollo. The two are side by side in the city, and we might reasonably have hoped that each would yield an equal treasure. Indeed, when we remembered the friendship of Amasis with Poly crates of Samos, and the offerings that he made at the Hergeum in that island, we might well have regarded the temenos of Hera as the most promising of all sites at Naukratis. But all such expecta- tions, as has been related, proved deceptive, owing partly to the digging of the Arabs, partly perhaps to other causes which we cannot now even conjecture. But it is unprofitable to discuss hopes that were not fulfilled ; all that can now be done is to give some account of the results that actually did proceed from a care- ful exploration of the site. The walls of the temenos, which is of very considerable extent (see general plan, PI. lY.), had been completely traced by Mr. Petrie in the previous season ; but he had then no evidence to lead to the true identification of the enclosure which they surrounded. He had indeed conjectured that it might be the Palaestra, because an inscription,* found else- where, recorded the dedication of the Palsestra to Apollo. Hence it seemed probable that the large enclosure adjoining the temenos of Apollo might be the Palestra ; but Mr. Petrie himself See Nauk. I. PI. XXX. p. 63. only regarded this as a conjecture, not of course to be insisted on in the face of definite evidence. Such evidence has now been found, in the vases bearing incised or painted inscriptions of dedi- cation to Hera ; and accordingly the name of this great enclosure has been altered in the map. In clearing the surface of the temenos, some traces of buildings were found. The most im- portant of these presented an oblong plan, and its internal measurements were 56 feet from N. to S., 18 feet 10 inches from E. to W. It was surrounded by a low wall of mud brick, about twenty-six inches thick — much too thin to be the outer wall of a building of such dimensions. The whole space within this wall was carefully levelled, and had been covered with sand. Hence it seems clear that what is now left is only the foundation of an edifice once built of stone, surrounded, as in other cases at Haukratis, by a retaining wall of mud brick. The levelled surface is at a level of about 300 inches above the datum : thus this building ranks among the earliest at Naukratis. Was it the temple of Hera ? This is a question which cannot now be answered decisively; but its posi- tion and dimensions make such an identifica- tion extremely probable. The chief objection is that its narrower ends face north and south, not east and west, as is usual in the case of temples. But this is not a fatal objection ; instances of such a situation are well known, as, for instance, in the case of the temple at Bassse. But, on the other hand, there are here no exi- gencies of the ground to necessitate any such arrangement. On the whole, however, we TEMENOS OF HEEA. 01 seem hardly justified in asserting that a build- ing of such dimensions, situated in the temenos of Hera, cannot have been the temple of the goddess. One other structure calls for mention, though its intention cannot now be discovered. This structure appears to be the lower part of two pillars, with a plaster floor between them ; the level of their foundation is about a foot above that of the building just described. They are in a line with its east wall, at a distance of about nine feet from its southern extremity. But whether they formed a part of another building, or are an independent construction of which we cannot now conjecture the use, must remain in uncertainty. This is but a meagre and unsatisfactory account to give of the architectural remains found in situ upon a temenos of which so much had been expected ; but it is all that can be stated without entering the realm of baseless conjectures. One other architectural relic may here be noted, a piece of egg-moulding in limestone, probably from the earliest temple of Hera ; but its forms are not remarkable enough to be worth detailed description. (50) If we turn next to the enumeration of the various objects found upon this site, we shall not find in them any richness to com- pensate for the poverty of its architectural remains. The most important are some vases and fragments of vases with dedications to Hera incised upon them ; these will be found in their due place on PI. XXII. The vases themselves are not, in most cases, of much interest. A favourite type, however, in this temenos is a small cup with one handle — a shape which we find indeed elsewhere, in the temenos of Aphrodite for example, but which in no other place is so common as it is here. These cups are sometimes made of rough red ware, some- times with a slight glaze, of a yellow or brown colour ; they also bear sometimes the name HPH in large letters. Other articles that were found in the loose refuse scattered over this temenos need not be described here, as they do not properly belong to the sanctuary of the Samian Hera, but had only found their way into it by accident. They may either have been thrown here from other sites in the neighbourhood, or from higher strata that in later times had occupied the same ground, and had been successively dug away by the Arabs. It is possible even that the in- scription, ’lepov .id to? ’ArroTpoTTaLov, to which we have already referred as giving the first clue that led to the identification of the site, may also have come here in this way. In any case its presence was fortunate, since it led to the identification of a temenos which is rich in associations though poor in remains, and which, if still unknown, might have led to a deceptive estimate of the treasures still buried at Nau- kratis. 62 NAUKRATIS. CHAPTER VIII. INSCRIPTIONS. (51) It will be as well to state at once that the inscriptions found in 1885-6 are not, in importance and antiquity, comparable to those produced by the previous season’s excavations : and that they do not throw any light upon the questions that have given rise to some contro- versy. That controversy cannot be ignored or passed over; but since no new facts are forthcom- ing which bear directly upon it, to deal with it here would only complicate its issues, and confuse the discussion of the new matter now before us. In the final chapter a few words will have to be said as to the details of the epigraphic evidence, and as to the light they throw on the age of the Greek colony at Haukratis. But at present we are only concerned with the interpretation and the transcription of the inscriptions found in the course of the last season. Here again the temenos of Aphrodite has yielded by far the most interesting and numerous collection. The ^circumstances of their discovery and the pot- tery on which they are incised have already been described. The inscriptions themselves will be found upon the large folding plate (PI. XXI. ).^ They are mostly in the Ionic alpha- bet; the only exception that calls for special notice is that of the inscriptions 86-93, which are, as we shall see, in all probability Lesbian. If so, they are of great interest, and are by far the most important epigraphical discovery of the year at Naukratis : for hitherto no Lesbian inscription has been known previous to the * The inscriptions have, with a few exceptions mentioned, been traced directly from the originals, and then copied by a photographic process. The forms of the letters may there- fore be relied on as accurate. general adoption of the Ionic alphabet. Hence both the letters and the dialect are important. But all such matters can best be discussed in immediate connection with the inscriptions to which they refer ; I will therefore proceed at once to the transcription of these, adding such notes as may seem necessary or desirable in each case. I do not think I have omitted any inscriptions that are of interest either from their substance or the forms of their letters. Mere repetitions of avidrjKev and ry ’ A(f>poSLTrj were obviously not worth recording. INSCRIPTIONS PROM THE TEMENOS OP APHRODITE. Ornate howls, painted both inside and out with animals in red on a white ground. (PI. VI.) 701-705. 701.^ ^dxTTpaTos p,’ dv€0r)Kev T'gfjypoBiTr). The inscription is also visible on PI. VI. We find that between the article and noun hiatus, crasis, and prodelision are all allowed; here the second is found (TH) ; we might equally well have Trj ’AcfipoSCrr) (THIA<1>), or Trj ’cfypo- SiTr} (THI). I shall not again notice these varieties when they occur. 702-5. These may all be parts of a dedication precisely similar to 701 ; they are all from dif- ferent fragments of the same bowl, of which only a few scattered pieces were found. 705 is on the outer rim, the rest inside. 702 has much older forms than 701 ; but the bowl was similar, and the name of the dedicator seems to have been the same in both cases. But we cannot “ I begin with this number in order that these inscriptions may be consecutive with the series in Nauk. I., numbered 1-700. INSCRIPTIONS. r>3 believe that the same man dedicated both at the same period. The complete bowl, which is more careful and ornate in its style, may have been a later offering from the same man, to supersede his earlier gift. White-glazed Nauhratite ware ; inscriptions incised on the body of the vase. 706-738. 706. ... 7]^ 6 The Chians were among those who shared the privileges granted by Amasis to Naukratis. Her. ii. 178 ; cf. 757. 707. . . . t0e(cr)Tos avi9rjK[^e rjp ’A<^poS[iT]p[i 708. . . . (p,)atos avedrjKe 709. ... 05 fi’avidrjKe ’.4^po8tTp] ini Tfj. . . . 710. 6 Selva dvedrf\Ke rg 'A(f)poSLTr) 711. Part of a name? 712. a[ve6r}Ke. As that of a daughter of Peitho (see Pape) this name may have been assumed by a Naukratite hetgera. 713. 714. Parts of names? 715. ^[. . . a]ve9rjKev Trj ' A^poSiTYj. On the bottom of a lamp of the central tube pattern, of white Naukratite ware. 716. ‘Epp6[TLpopoSiTr}'\L ©T 7 crar[ 8 po 5 . The name is new, but not impossible ; or we may restore AL07Te^Ldr)<; dv[^edr)Kev. 741. AlyvTrTL\_o' 5 . This is interesting, if in- dicating the nationality of the dedicator ; cf. the second suggestion for 737. There is no authority for restoring ’^^^ 081777 ] AlyvirTilg. ; the epithet is, however, applied to Zeus and Dionysus, but only late. See Pape. 742. . . . 17 X 05 p dveOrjKev. 743. ZwiXo 5 or Z(t)uko<; ; the reading can hardly be doubtful, but the spelling is peculiar : cf. Trpo^ or T-gdo^, Nauk. I. PI. XXXV. No. 700 ; the first t must represent a kind of y sound preceding the vowel. 744. 'O Selva e^SoiKe. Such a form of dedi- cation is here unusual ; but it seems the simplest restoration. 04 NAUKRATIS. 745. Mt/cls auedr]Kev. This feminine name will bear the same relation to Mlkojv and Mlklcjv as, e.g., ria/jjaerts to Uapfievcov and Ilapfievicjv. 746. . . . iTTi? a\^ve9'r]Kev 74<7. ra^\r]i’L(o\_v or r^\rjVLO)[v Nauhratite vases ; inscriptions incised on bottom. 748-761. 74S-753 (except 751) are copied, not traced, since the curvature of the bases made it almost impossible to employ the latter process with satisfactory results. 748. ‘Eppricn(f)dvr)<; p dvidpKev T'q(f)poSLTrj. This coils right round the base ; a double stroke is placed between v and e, where the second line comes beneath the first. 749. Trjcfypo^iTr) OopriiXo? dve[^dr]K€v. 750. Ty ’A(l)pohL^Tr)‘Eppaya9lpo^p^oin€d[r]K€v ; in the second a the third stroke is probably a mere accident. The first part of a similar name may be seen in 762. 751. AipKy^ ry ’A\_(f>poSiTy ; cf. AepKvXos, &c. 752. Trj 772. This and 773 are on a similar light- coloured ware, and may be parts of the same plate. Thus we may read : 'EppoKpdT[7]<; dvdd]r)Ke Tr}(j)pohlT[rj. The writer seems to have made a false start with the first two letters. 774. These letters are painted in light red on the bottom of a plate. 775. KjXeohrjpos pe d[ve]d7}Ke rfj ’A[(f)poSLTr). 776. Xdpp[^r]^ pe dve9r]K]e T^[t ^A^pohirrj ev\xdvr]<; dve6[r]Kev] 6 Nav(TLTe[\ov<;. 785. 'O SeiraJ p‘dv[e9r]Ke rfj ’Afp^odlry. Written from right to left. A very peculiar ivare, blade throughout, and very highly polished on the surface ; decorated tri- pods, fc., large conical bases, and bowls with a flat rim are the most usual forms; inscriptions incised. 786-793 ; cf. also 840. 786. 'O Selva d MjaXdet? ’Jo. . . . See after 793. Maloeis is an epithet, perhaps local, of Apollo in Lesbos ; may it be also used of one who lived by his sanctuary at the place named after him ? 787. Td ' A^poSira 788. 'O Selva Kd9~\9gKe ra 'A(f)poSLTa 6 Mvtl- XijvaLos 789. 'O Selvdi] Upe Kd)\9r]Ke d '[MvTiK’gvaios {lpdve)j 790. '0 Selva p]e Kd99g[_Ke~] 6 MvT[fK'QvaLo<; 791. 'O Selva K^d99r][^Ke 792. . . . \r] ep[e Kaddyne 793. . . . (jjv e. . . .? In this series of 8 inscriptions, distinguished from all the rest by the ware they are incised upon, wo notice at once some striking pecu- liarities. Two are almost certainly dedicated by MytilensBans ; a third, 786, by a man of MaXdets,^ a harbour in Lesbos with a temple of I Apollo Maloeis, as we learn from Thucydides. I All the inscriptions seem to bo in the same alphabet and dialect ; a dialect which in every j 3 'XJie first letter of this word is liardly doubtful ; it must ‘ be either /x or A, from the traces left. I XAUKRATIS. 6 () way corresponds to what we know of the yEolic of Lesbos, Twice also a peculiar for- mula of dedication, KaddrjKe, is found. Putting all these facts together, we may without rash- ness conclude that we have here specimens of the Lesbian dialect and alphabet. These are of the greatest importance, as no inscriptions were before known from that island of a date prior to the prevalence of the Ionic alphabet. We may now thus record the Lesbian alphabet ; it seems to have no peculiar forms. a S € r) 9 i K \ IJ. V Ou> (> (T r A..AEi..0IKNM/V.O.P5T. 0 ^ As regards dialect, the most important fact to notice is that a double aspirate is used wherever possible; thus we should probably write in Sappho’s poems Sacfxfxo (or Wacfxjico), Kayyeerai, &c. For these inscriptions are, per- haps, within 50 years of the date of Sappho’s writing, and so supply by far the most trust- worthy evidence now to be found as to her orthography. It is very remarkable that no f appears; we should certainly have expected one in MaXoets; cf. Maleventum. Stone; inscriptions incised. 794, 795. 794 is on the back of a limestone statuette, representing a female figure. noXveppoq [xdv^^edrjKe] Trj' 'A(f>poSiTr]. 795 is on the edsre of a limestone dish, found o y just outside the temenos on the west; the third line is inside the dish. JStJ? Na\_v]KpaTLV [a(f)LK6p.ei'^o<; ^'A(j)poSiT7f\i. Kat 9 o[9 duedrjKev. This is of great importance, from the scarcity of inscriptions that mention the name of the town. BlacJc and red holds, mostly plain ; inscriptions incised. 796-826. 796. ev 797. T7}(^poSi[rp 798. Aojph (f)ikT\^pov (?) 'A(f)pO^LTp. 799. ’/2yt\o? (?) pddveOrjKe. It is doubtful whether the name is complete or not ; in any case it is curious. 800. . . . SpaTCJu pe K[d9r)Ke. This is not Ionic, if rightly read; it resembles the My- tilensean we recently met with. 801. '£JXe|[crty8i09. 802. ’A(^po]8tTat. Again non-ionic. 803. Part of a name ? 804. 'J4payope[us] | Trj(f)pofBLT] \ p 6. . . This was found near the top of the excavations, and probably does not belong to the remains of the first temple; but it can hardly, from the forms of the letters, be later than the fifth century, B.o. If so, it is interesting, as showing, per- haps, the earliest known example of true liga- tures. (Those of Amorgos are, at least, doubtful, and even if the connecting lines be part of the inscription as intended, they do not produce true ligatures, since the letters combined do not become integral parts of one another.) 805. TeL(Td[_pevo<; 806. 'Jcr]Tiat05 d^yediqKev . . .? 807. ’A^poStJro. 6 M. . . . Non-Ionic. 808. ^i;]XXt9 dv\^edrjKev. 809. Part of a name ? 810. , . . Xvra d[^i>€d7)K€v. 811. Part of a name ? 812. ’A(f)poSLT[rj 813. . . , TTOS AlO. . . . 814. ’A(f)podi]LTa o. Ke[. . . . — not o Ketos ; the dialect precludes such a restoration. 814*. . . . 6vr]^10?] TOJTToWcxiVL 877. nvp(p)o^ pe dveOrjKe. On the rim of a black and red bowl. 878. TwTToXXwro? [etpi. On a bowl with birds. 879. ’AnoWcovoq dpL. On a bowl with birds. 880. ^AttoWo). ... In a red band of a black and red bowl. 881. ’^TToXXojros elpL. On a scarlet and reel (?) eye-bowl. The second X has been at first omitted, and afterwards inserted. 882. On PI. xvii. 3, will be found a curious caricature, scratched on the bottom of a vase ; round it runs the inscription ’ AneWapoveiov and nOAKO MIMHMAE TnOdKov plprjp ’E. . . . Thus, the satyr-like head, the “ image of an ape,” is doubtless described as the likeness of some one whom we may suppose to be Apellamou. (55.) Inscriptions from various sites ; on stone. 12 and 13.'^ Two sides of one thin marble tablet. The first is much earlier than the second, and it seems that the tablet may have been reversed and used a second time in the This numbering, as well as that of the inscriptions on pottery, continues that of Nauk. I. fifth century, n.c., the older inscription being no longer wanted. The inscription was found about 100 yards east of the temenos of Aphro- dite ; but a search in the neighbourhood revealed no further indications of a sacred enclosure. 13. ... o? 1 ... . heKd[T7)v; the first letter of the second line must be X, p, or a ; probably the last ; if so, we may read dvedrjKa. 14. ‘lepov zlt[o 5 1 dTTOTpoTT[aLOV. This has already been referred to as being found in the temenos of Hera. 15. . . . ’A^ppcjviov. . . . iu crJuroSm ^ap^aOtKyj (?). .... Kal(japo<; (J)apeva>6 IG. . . . 7ro\vp6p(j)[(p. -dav eSa)\r)v Apparently from the basis of a statue, perhaps of Artemis Hecate, whom the epithet fits. 18. jfTroXjep.raot' | ^iXoTrarJopa $t[Xa§€XANEPONTOIZEniriNO- MENOIZ ’Er^aSe MeXi[av Karejyet ttovXv^o- retpa TrjLov, 09 7Tda"r) ecrrt tovto to Tepepo?, Kal TTpocTTctra? Tov ipTTopiov avTat al TrdXte? etcrt at Trapeyoucrat* ocrai Se dWai TrdXte? peTairoLevvTaL, ovBeu cT(j)L peTeov peTaTTOLevPTaL. yojpt? Se AlyLvrjTai iir iwvTwv ISpvaavTo repet'o? -dtd?, fcat aXXo Xdpioi 'Hpr]