§ I.—The interest of the excavation at Phylakopi in 1898 reached its climax just at the close of the season with the discovery in a small room in G 2 of the “Fishermen Vase” (lamp-stand), and with the equally interesting find in a house in G 3 of part of a wall-painting of flying fish (B.S.A. 1897–8, pp. 26–7). The south part of this region had to be left unfinished when excavation came to an end last year and it was desirable that, before we proceeded to any other undertaking, this piece of work should be completed.

On reference to the plan, it will be seen that the N.E. part of G 3 is taken up with an apparently domestic building containing a room with a square pillar near its east wall. This house was the one in which last year were discovered, in the room with the pillar, the frescofragments with flying fish.

The excavations this year proved to be of considerable interest, so far as the architecture was concerned. Besides the Palace several complete private houses were found, and the streets were followed up far enough to enable us to guess at the system on which the town had been laid out. Several houses of the third settlement can be made out, but there are only slight indications of its streets. So far as they can be traced, however, they seem to coincide roughly with those of the fourth settlement. With the fourth settlement itself the opposite is the case—the streets can be traced, but there are not quite so many complete houses as in the third settlement. In a few cases, perhaps, a street or house of the earlier period continued in use in the later; but generally speaking this is not the case, and where a later wall is built on the top of an earlier one, it was probably in order to obtain a more solid foundation.

In a report which appeared in the last number of the British School Annual I gave a short description, in chronological order, of the main classes of pottery found at Phylakopi, and tried to indicate their place in the history of Aegean art. I shall not attempt in the present paper to write a fuller and more accurate account on the same lines. The proper place for that will be in the final publication. All I propose to do now is to make a few desultory notes of a general character on the finds of this last season.

The supply of pottery was as abundant this year as ever. To give an idea of how closely the soil is packed with it, I find on a rough calculation that an average day's work yielded somewhere between ten thousand and twenty thousand fragments. The experience gained in the preceding season made this large daily harvest much more easy to deal with.

The object of this paper is to endeavour to attach their right names to the ancient sites which lie close to the Peneios on either side of the river in its course from Trikkala to Larissa.

The first site which we reach in travelling down stream is that of Palaeo-Gardiki, halfway between Trikkala and Zarkos, close to the high road, on the left bank of the river, which is here some six kilometres distant. On a small spur of the mountains, with which it is only connected by a narrow neck or spur which slopes down to the plain, lie the remains of an ancient acropolis with walls built of both polygonal and ashlar masonry. In front of the acropolis in the plain the remains of a city of no small extent are visible. The walls of ashlar masonry remain to a height of four or five courses, and the foundations of houses and public buildings are clearly marked. The name Palaeo-Gardiki comes from a Byzantine village, Gardiki, which once stood upon the acropolis, but of which no traces now remain except the ruined church. In front of the city lies a marsh, which, combined with the mountain behind, makes the position very strong.

Mr. Ernest Gardner, summing up in 1889 the results of the excavations at Naukratis which Mr. Flinders Petrie had begun in 1884, was of opinion that something still remained to be done on the site. This opinion was shared by Mr. Petrie and has been expressed by him more than once, when Greek remains in Egypt were under discussion. A further campaign, therefore, has long been contemplated; and when I heard late in 1898 that very serious encroachments were being made upon the mounds of Gaif, it seemed that it was time to act. Paying a flying visit to the place in December, I found that, compared with Mr. Petrie's map of fourteen years before, the Mounds showed a greatly changed appearance. The “Great Temenos” at the south end had become a cornfield: the “Arab village” shown in the map on the north-east edge of the mound was now divided from the uncultivated land by a broad belt of green, which in the east centre had been pushed far out into the hollow heart of the site by an artificial embankment twelve to fifteen feet high.

§ 1.—All archaeological scholars are familiar with the splendid find of Greek pottery which we owe to Mr. Petrie and Mr. Gardner, and with the incised dedications on which are preserved the handwriting and names of the early inhabitants of Naukratis. Those interested in either subject will find in the following pages some additional material selected from the results of the recent excavation.

The sandstone relief reproduced on Pl. IX. represents a helmeted warrior marching to the right with spear in his right hand and shield on his left arm. It is the inside of the shield that is turned towards the spectator. A procession of similar figures is a very common subject in early Greek art, from the Mycenaean age onwards; Nauk. ii. Pl. V. 6 is a fragment of just such a scene. The tread of the warrior on our relief recalls, and is in fact the prototype of, that mincing step which is one of the mannerisms of the archaistic style (cf. Hauser, Neu.-att. Rel. p. 165; Émanuel, La Danse Gr. pp. 48 ff).

The terracottas, which were recovered during the spring of 1899 from the site of Naukratis, are both more numerous and more important than those obtained during the two previous campaigns. They number nearly four hundred, the specimens ranging in date from the sixth century B.C. to the second century of our era. A number of these are of types akin to those already known at Naukratis, and are present in such quantity as to make it possible to trace the probable development of their various peculiarities. Besides these, and thirty or forty small female heads of “Tanagra” type and other figures of comparatively late date, we have the remains of nearly a score of female heads of unusual size and beauty, which are to be ascribed chiefly to the fifth and fourth centuries B.C.