In the spring of 1910 the work which had engaged the School at Sparta since 1906 was brought to an end, and it was decided to devote a season to a supplementary excavation at the prehistoric site of Phylakopi in Melos, the scene of the School's activity from 1896 to 1899. These dates are of some significance; the work in Crete, which has revolutionised our ideas of the Bronze Age in the Aegean, was only begun in 1900 with the first work at Knossos, and on the mainland of Greece the recent work of the German Institute at Tiryns, the Greek and German work at Chaironea and Orchomenos in Boiotia, and the prehistoric researches in Thessaly and North Greece, first of Tsountas at Dimini and Sesklo and later, over a much enlarged field, of Mr. Wace and his colleagues, had none of them been even begun when the first excavation at Phylakopi was closed in 1899. It was in fact the first important prehistoric excavation in Greece since the death of Schliemann. Few comparisons with other sites were therefore possible; the results of the work of Schliemann and of his successors at Mycenae and the researches in the Cyclades published by Tsountas were the main guides for Mr. Edgar in his pioneer work on the pottery from Phylakopi.

The work began on the 30th March and lasted, with the usual interval for Easter until the 30th of May. The maximum number of workmen employed was about forty. The party consisted of the Director, Mr. J. P. Droop, Miss Lorimer, and Mr. W. R. Halliday, who were present throughout, Mr. L. B. Tillard, who remained until Easter, and Miss Lamb and Miss Tennant, who came after Easter, and stayed until the end. Mr. M. S. Thompson paid a visit towards the end of the season, when Mr. W. J. P. Jones also came as architect, to draw the plans which accompany this report. He also made a series of water-colour drawings of vases, some of which are reproduced on Pls. II. and III. By the kindness of the Athenian authorities of the German Archaeological Institute, we had the services of their veteran foreman Evangelis Kosmopoulos; as mender we again had Ioannes Katsarakis from Palaikastro. The photographs were taken by Herr Röhrer, the photographer of the German Archaeological Institute. It is a pleasure to thank the Greek Archaeological authorities for their support; in Athens the Ephor Dr. Leonardos, and in Melos Dr. Stavropoulos the Ephor of the Cyclades, and the local representative of the government, Dr. Aristides Papa-Ioannou, Scholarch of Melos, whose firmness in dealing with the claims of the proprietors of the site was of the greatest assistance. The objects found are to remain in Melos and form the nucleus of a local Museum in the town.

In several places near the city remains of numerous rock-cut tombs exist, and it would appear from the prevalence of the local Geometric among the fragments of pottery picked up near them during the old excavation, that they belong in great measure to the end of the First and the beginning of the Second City. These tombs had all been robbed, and therefore no bones were found; if they had been undisturbed, it is now safe to say that at least all those with Geometric vases would have proved to be the tombs of adults, for a series of burials of children of this period has now been found, which shews a curious difference between the methods of disposing of the adult and of the infant dead.

On the surface of the rock, upon which the foundations of the houses of the lowest stratum rest, there were found a number of shallow basin-like depressions, in which large pithoi were lying. Probably owing to the hardness of the rock these holes were not more than at most a foot deep, and the pithoi are consequently only partially sunk in them.

The trade connexion between Melos and Crete, which is proved as early as the Neolithic period by the presence of Melian obsidian in Crete, is not indicated by the presence of Cretan vases at Phylakopi before the beginning of the Middle Minoan age. The suggestion that the lustrous glaze Geometric ware was imported and not native, has not been supported by later evidence, and of the Cretan pottery with light-on-dark geometric patterns of the Early Minoan III period, which was first observed at Vasiliki near Gournià, only one sherd was found, and even this is not entirely certain. The finds of polychrome Kamares ware on the other hand though small in comparison with the mass of local pottery, were frequent and well distributed over the site. About seventy per cent, of the total amount found came from levels in which pottery of the native Geometric style predominated; of the rest, the greater part was found with the pre-Mycenaean Melian ware, as we prefer to call the Early Mycenaean with curvilinear designs in matt black of Mr. Edgar's classification, whilst a certain amount was found amongst later fabrics in strata which had probably been disturbed.

A small quantity of the ware generally called Urfirniss was found lying on the rock, chiefly in the south-eastern part of the excavation, that is to say in rooms 1 and 2. From its position it would seem to belong to the earlier part of the First City, although the extreme scarceness, in the deposit, of the incised fabrics, warns us against a too early date. It was so much broken that nothing can be said of the shapes, but the characteristic coating of reddish-brown slightly lustrous paint was quite recognisable. It is of some importance as an early link with the mainland, as similar pottery has recently been found in great abundance in the lowest stratum at Tiryns by the German excavators, who consider that this Tirynthian Urfirniss is allied to early Cycladic pottery. These Melian pieces may well be of Cycladic origin, and an instance of this connexion.

Work carried out in the course of the last ten years on prehistoric sites has given an interest to another class of pottery which it did not yet possess when Phylakopi was first excavated. Thus it came about that only ten lines, and that in the section devoted to ‘odds and ends,’ were given to the very considerable finds of that ware to which Schliemann gave the name of Minyan. This name will be used here without any discussion of its propriety; it is convenient and has been generally adopted. This much may be said for it; so much of this pottery has been found at Orchomenos and in the neighbourhood of Drachmani (Elatea) that it is reasonable to believe that it was made in Boiotia and Phokis.

Pottery with a finely burnished surface, brown to deep black or less commonly red, was plentiful among the Geometric deposits, and not scarce in the higher levels in which the pre-Mycenaean Melian predominated. The sherds found in still later deposits were isolated and must be regarded as out of place; whilst, close to the rock, which in the area excavated this year means in the middle period of the First City, they were also scarce. It belongs therefore to the early period of the Second City, and is contemporary with the Cretan Middle Minoan.

Evidence of the local use of bronze or copper appeared amongst the deposits in H 5 1 in the shape of several pieces of small clay crucibles. None of them were complete, but the two best examples, shewn on Plate VII give the shape. They were spoon-shaped vessels about ·24 m. long by ·16 m. wide, made of coarse dark clay with a spout at one end; this is seen in the left-hand example on the plate. They stood on three short legs so arranged that the crucible could be tilted forward and cause the molten metal to flow out of the spout into the mould. That they had seen some service was proved by the remains of metal which clung to their inner surface. A fragment of such a crucible was found in the old excavation and two broken stone moulds, one for a Double Axe. Such a Double Axe, shewn on Pl. XIV, 59, was found this year together with a bronze spear-head, in a cleft between two stones almost immediately above the rock underneath a floor-level in H 4 15. They had the appearance of being in a sort of cache and probably belonged to a later period than that of the First City, to which their actual level would assign them. The only other bronzes found were the arrowhead and knife shewn on Pl. XIV (Nos. 58 and 60).

In the course of the excavations carried on by the School from 1906 to 1910 at Sparta, a considerable quantity was found of the mediæval glazed pottery that is usually called Byzantine. As specimens of this fabric are not common in museums, and its date and general relations are still matters of some doubt, it seems desirable to publish all the pieces of any interest as material for further study, although the writers of this paper have no claim to speak as experts on the subject. The fragments were found in the numerous trial pits that were made on and around the Acropolis, and were especially abundant outside the east end of the Late Roman fortifications. The exactly similar ware from Pergamon and Constantinople now in the Kaiser Friedrich Museum at Berlin is there officially ascribed to the thirteenth and fourteenth centuries, and other pieces from old Cairo in the same museum of very similar make, to the eleventh and twelfth. Two almost complete bowls from Pergamon, on the other hand, plainly of our classes I and III, are assigned by W. Altmann to ‘the first centuries of Byzantine art’ The position in which the sherds were found at Sparta gives some slight indication as to date in the form of a probable terminus ante quem, for the building of the fortress of Misithra, or Mistrá, by Guillaume de Villehardouin, in the middle of the thirteenth century, was quickly followed by the decay of the Byzantine city situated round the Acropolis of Sparta.

Fabricius' view, based on archaeological evidence, that the lower town of Thebes extended over the high hills East and West of the Kadmeia (Pl. XIX. A) has in general been accepted by subsequent scholars: it has only been modified by the theory of Kalopais and Soteriádes, which makes the town extend yet further eastwards.

How weak this archaeological evidence is, was shown by the criticisms of Wilamowitz and Frazer; and the literary evidence suggests quite a different view. It is to this that I wish to draw attention. Any theory based on such evidence is of course liable to be upset at any moment by fresh archaeological discoveries. But in the present uncertainty it may be useful to see to what theory this evidence seems to lead us.

Thebes is situate towards the East end of the long range of low, cultivated hills, running eastwards from Helikon as far as Mount Sorós, and dividing the Aonian plain on the North from that of Leuktra and Plataia on the South. Here is a small group of hills, none of them rising much above the general height of the range, divided by the three streams flowing from. South to North, the Plakiótissa (identified with Dirke), a small and nameless brook, and the H. Joánnes (the ancient Ismenos)

In May 1911 I made a systematic examination of the fortifications of Phokis. My attention was primarily directed to the fortifications quâ fortifications, but I have been led to some identifications which may help to add more certainty to the topography of Phokis (Fig. 1). Briefly, my results so far as concerns the plain of the Kephissos are as follows:—

(i) Tithorea occupied the site of the earlier Neon, that is to say approximately the site of the modern village Velítza. Historically the town was of no importance; the military remains are among the finest and best preserved in Greece.

(ii) Erochos was situated near the village of Kato-Souvála.

(iii) Charadra is to be placed at Mariolátes.

(iv) A Kastro, marked on the French Map as ‘Psili Kastro,’ may be identified as the Patronis of Plutarch.

(v) Near some Hellenic remains, which have been conjectured to mark the position of the Phokikon in the valley of the Platanias, is a Kastro hitherto ignored or undiscovered by travellers. The presence of this fort—it was not a town—lends colour to the identification of these Hellenic remains as the Phokikon.

The remainder of the article deals with the fortifications of Phokis, which with the exception of Abai, Hyampolis and the remains at Modi, are of a uniform type and date from the restoration after the battle of Chaironeia in 338 B.C.

In the rather uninteresting landscape of the Thracian plains artificial mounds are almost everywhere a conspicuous feature. These mounds have long attracted the attention of archaeologists, but, on the Turkish side of the frontier at least, few have been scientifically explored. In Eastern Rumelia, where considerably more has been done, excavation has shewn that a large proportion of the mounds are places of burial, others representing, like the Thessalian magoules, prehistoric settlements, and others again sites of ancient watch-towers or even of Turkish camps.

The neighbourhood of Kirk Kilisse, some thirty-five miles east of Adrianople, boasts its full share of these monuments. About half an hour south of the town is a group of three tumuli in a line running W.S.W. and E.N.E. All are said to have been excavated by the Russians during the campaigns of 1828 and 1878. The southernmost tomb of the group remains open, shewing the arrangement of the tomb-chamber and its approach.

Had I been aware, when I published the reconstruction of the great group of Damophon at Lycosura in volume xiii. of this Annual, that there was even then lying under lock and key in the National Museum of Athens a coin bearing on its reverse a representation of that very group, such misgivings as I had about the correctness of my reconstruction would have been greatly intensified.

The lucky re-discovery of this coin (Fig. 1) and its publication by M. Staïs in the last number of the Journal International d'Archéologie Numismatique afford an embarrassing opportunity for testing the value of the reconstruction, in which M. Staïs has treated me with great consideration. Before discussing his analysis of the coin, I should like to express to him my gratitude for acquainting me so promptly with its discovery and providing me with a cast.

The monument of the charioteer Porphyrios which stands in the Atrium of the church of St. Irene at Constantinople will no doubt be an object of interest now that the church has been thrown open to the public as a Military Museum. The description of it by Mordtmann which appeared in 1880 is still the best available, and except for some further remarks by the same writer nothing seems to have been published subsequently concerning it until the appearance last year of a short paper by M. J. Ebersolt who treats the sculptured reliefs from the artistic standpoint, and discusses the place which they occupy in the history of Byzantine Art.

The notes which I publish here consist of a few comments on the texts of the inscriptions as printed by Mordtmann, and small corrections in them, together with a suggested interpretation, which seems new, of one of the scenes sculptured on the stele, and a short account, kindly supplied by Professor J. B. Bury, of the language and metre of the inscriptions in popular Greek which appear on two of the four sides of the stele.

The inscriptions in popular Greek on A and B are highly interesting as the earliest examples of the metrical cries of the Factions of the Hippodrome. Hitherto the earliest known example dated from about half a century later, in A.D. 561 (see Theophanes, ed. De Boor, 236). Then we have several acclamations in the record of A.D. 602. Phocas was greeted by the chant:—

ὕπαγε, μάθε τὴν κατάστασιν, ‘Away, understand the situation,

ὁ Μαυρίκιος οὐκ ἀπέθανεν, Maurice is not dead,

<ὔπαγε> μάθε τὴν ἀλήθειαν (Away), learn the truth!’

(See Theoph. Simoc, [304 ed. de Boor], and John of Antioch, 218d, in F.H.G. v. 37.) Each verse has ten syllables and the accent falls on the antepenultimate. In all this popular ‘poetry,’ the metre depends not on quantity but on the number of syllables and the accent.

In a note on the Argive festival of the Hybristika I mentioned the diviners of Skythia as examples of that strange phenomenon familiar among many savage peoples, the magician who dresses as a woman. As the θήλεα νοῦσος has been somewhat of a crux to Herodotean scholars, I may perhaps be allowed to develop the point a little more at length.

It has usually been suggested that the θήλεα νοῦσος must be explained either as a vice, or as a bodily disease, or as a mental affliction. None of these theories, however, accounts for all the features of the evidence which we possess, and can only be tortured into conformity by the Procrustean ingenuity of the scholar. Those, for example, who explain the θήλεα νοῦσος as a vice, have further to show why those who are afflicted with it form a definite class, why they are connected especially with Askalon, and why they are magicians.

John Covel, afterwards Master of Christ's College, Cambridge, visited Athos from Lemnos on his way from Constantinople to England in the spring of 1677. His letter of recommendation from Nointel (which he hoped might be of use if he fell amongst French ‘privateers’) and his circular letter to the monasteries from Manolaki Petrovitzes are preserved in his correspondence.

The Athos notes are bound partly in the Folio MS. (Add. 22, 912) and partly in the Quarto (Add. 22, 914). They are very hastily written and ill arranged. I have therefore taken considerable liberties with the text, revising punctuation, resolving contractions, and (for the convenience of the reader) often transcribing words written wholly or partly in the Greek character; such words are indicated by italics.

The town of Amastra (anc. Amastris, med. Samastro), one of the most important of the ancient Greek colonies on the Euxine, and later a Genoese trading-post, lies on the neck and landward slopes of a steep peninsula forming a safe harbour on either side. The peninsula is connected by a bridge with an adjacent island, on which are a few houses still protected on the harbour side by the mediaeval town-wall. The bridge is defended on either side by a gate-tower, and the city proper and castle, which are on the peninsula, retain their lofty walls and towers, built of squared blocks evidently plundered from Roman ruins (Fig. 1). Despite this plundering and the recent construction of the new road (still unfinished in 1910) to Bartin, extensive ancient remains still survive above ground on the mainland opposite the town, notably the large Roman building of brick and opus reticulatum called the Bezestein, two columns of a temple or mausoleum, many built tombs, and considerable remains of masonry embankment and vaulting about the stream which debouches into the smaller (W.) harbour. Saving the change of ownership, the place has altered but little since Clavijo's description in 1404:—

‘The town of Samastro belongs to the Genoese, and is near the sea…. on a very high hill. Opposite this hill, but more inland, there is another on which the town is built; and from one hill to another there is a great arch, like a bridge, by which people pass, and it has two gates, one at each end.

The two white marble slabs shewn below (Figs. 2, 3), originally in the castle of S. Peter at Smyrna, are now built into the circular court of the prison, high up, facing the entrance from the street. I was first apprised of their whereabouts by the late Dr. Weber, who himself published them in the somewhat inaccessible ῾H μερολόγιον τῆς ᾿Αμαλθεἰας. For access to the prison I am indebted to the kindness of Mr. E. Harris, Consul-General of the United States.

The history of the castle need not long occupy us here. It is sufficient to say that it dated from the capture of Smyrna in 1344 by a combined Christian fleet under Jean de Biandra, Prior of Lombardy, and was held by the Papacy till 1374, when it was transferred to the Knights of S. John (Grand Master de Julliac) by Gregory XI. In 1392 it was repaired for the Order by Domenico d'Allemagna, Admiral of Rhodes, at the charges of G. M. de Heredia, again by G. M. de Naillac in 1398, and finally stormed by Timur in 1399. It was never recovered by the Knights, who are said to have built the new castle of S. Peter (Budrum) to replace it as a mainland stronghold.