In this paper the writer examines the largest group of Linear B tablets from Knossos, the great archive of records dealing with sheep. The results of this inquiry provide, it is suggested, a plausible solution to a long-standing problem, the source of the wealth of Knossos in the Late Bronze Age.

The account of the great archive of sheep records from Knossos (Series D) which Ventris and Chadwick present in Documents in Mycenaean Greek remains the most influential treatment of this series taken as a whole, and must provide the starting-point for any further discussion of the texts which it contains. In the course of their account, V.C. consider what the purpose of these records is likely to have been, and rightly reject suggestions that the sheep listed on them are, for instance, hecatombs of sacrificial animals, or merely tokens of exchange, as sheep sometimes were in the ancient Near East.

Of the considerable epigraphic harvest gathered by W. Kubitschek and W. Reichel in Geyre (Aphrodisias) in 1893, during their stay of sixteen days, only a few inscriptions were published in the Anzeiger der kaiserl. Akademie der Wissensch. zu Wien 1893, 100–3. Some of the inscriptions recorded in Kubitschek and Reichel's notebooks were published independently by Th. Reinach from Gaudin's squeezes, in REG xix (1906), 79–150, 205–98. Reinach's nos. 13, 17, 127, 7, 70, 77, 50, 146 bis, 32, and 62 (all published as new) are respectively nos. 1, 3, 4, 7, 8, 9, 10, 11, 12, and 14 of Kubitschek and Reichel's article cited above.

Total success in establishing a text for such a defaced document as the alleged decree of Themistokles can hardly be expected. The left half of the inscribed surface is so deeply scratched as to resemble a butcher's chopping block. There are so many scars, running in so many directions, that it is easy to imagine letters where we can be sure that none exists. For example, in line 8, but much below the inscribed surface, there are four strokes which seem to make a mu. In line 34 there are strokes in the shape of an alpha between the pi and eta of the word hyperesia. In both cases the ‘strokes’ are merely scratches. For this reason, this is one of the most treacherous inscriptions in Greek epigraphy to study by photography alone. And published texts which depend on photographs, of whatever number, should be treated with great caution.

The column drums of the temple at Phanai and the blocks and mouldings of Emporio have attracted a number of occasional scribblings; some of these are cut, many are scratched or punched in rows of single or multiple dots, for others the surface of the stone has only been ‘slightly stunned’. Most are memorial and need no general comment after Rehm's treatment of the Μνήσθη formula. Those that are complete are all of Rehm's ‘motherland’ type and the rest should no doubt be restored on the same pattern.

In general they are of little interest and less importance; indeed the main thing we should like to know about them is their date since in several cases the text has run on from block to block (nos 25, 28 (a), (b), and (c), and 29 (a)) and was therefore inscribed while the blocks were still in their original position. At the moment we have a terminus ante quem in the sixth century A.D. for the destruction of most of the pagan buildings on the site when the basilica church, in which most of the surviving fragments have been found, was built.

The inscriptions here published, which were found in the Anglo-Turkish excavations at Old Smyrna between 1948 and 1953, are mostly fragments from the short, trivial writings of daily life. There are also one or two dedications, and a few inscriptions painted by potters on their wares. No inscriptions on stone were found, but this is not surprising, for the parts uncovered were chiefly private houses and stretches of the city's great wall. One important temple of a goddess, perhaps Artemis, was located and cleared, but it had unfortunately been much eroded. No other public building was reached, nor an agora: so the present epigraphic harvest should more properly be called gleanings, some of it merely owners' or merchants' marks—single letters or signs—on domestic pottery.

These travels occupied us during four weeks of July and August 1962. Our initial intention was not to look for new sites, but to visit those already known, in order to revise the lists in Pendlebury's The Archaeology of Crete (1939).

Crete has perhaps been more thoroughly explored than any other part of Greece. Pashley (Travels in Crete (1837)) and Spratt (Travels and Researches in Crete (1865)) identified many ancient Greek and Roman cities. At the end of the nineteenth century came Evans, and the Italians Halbherr, Mariani, Savignoni, and Taramelli, who began the search for Bronze Age sites continued by Pendlebury, Kirsten, and others between the wars. Much has also been contributed by a series of distinguished Antiquities Officers, I. Hatzidakes, St. Xanthoudides, and Sp. Marinatos; and since the war by N. Platon, whose very full annual reports in Kritika Khronika have provided a valuable record of current excavations and discoveries. In recent years P. Faure has located many new sites in the course of exploring the caves of Crete.

The position and general form of the Stoa by the Harbour at Perachora have long been known. It was discovered and excavated in 1932, the third and fullest year of Payne's excavations at Perachora. Brief notices of its discovery are to be found in the preliminary reports for that year, in BSA xxxii (1931–2) 260–1, and JHS lii (1932) 242, and in AJA xxxvii (1933) 155, AA xlvii (1932) 563, RÉG xlix (1936) 151, and in ILN 8/7/33; of these the last is perhaps the fullest. In Perachora i a single paragraph is devoted to the stoa on pp. 14–15, but rather more information can be gathered from the plan of the Hera Akraia temple, the triglyph altar, and the stoa on plate 138. This shows the position, plan, and state of preservation of the building when it was excavated. Since the publication of Perachora i in 1940 no further work has been done on the stoa, but there is a brief treatment of it in R. Martin, Recherches sur l'agora grecque, 461, 465 n. 4, 486, and in tables iv and v, and a discussion and drawing of the Ionic semi-column capital in G. Roux, L' Architecture dans l'Argolide aux IVe et IIIe siècles avant J.-C., 346 and plate 91.1.

The excavations described here were undertaken for the purpose of obtaining a fuller knowledge of the nature of the Neolithic settlement site which lies below the Minoan palace of Knossos.

The presence of this settlement site was originally recognized by Sir Arthur Evans very soon after he began work at Knossos at the beginning of the century. His numerous soundings established that it was of considerable size, covering the whole of the hill of Kephala below the palace and the surrounding houses (an area of at least 11 acres), and that it must have endured for a long period, since the depth of deposit attained to nearly 10 metres in places. A pottery sequence was worked out by Mackenzie on the basis of the soundings carried out up to 1904, and this was subsequently adopted in essentials by Evans in the final publication. It has subsequently been refined and modified by Dr. A. Furness (Mrs. Ozanne), using Evans's material in the Stratigraphical Museum, in a paper which she published in this journal in 1953.

Whatever absolute dates are to be attached to the L.H. IIIA 1 period, its historical significance is indisputable. At present this significance derives mainly from events outside Greece itself, the fall of Knossos and the disruption in the Aegean area which went before it. On the mainland, although in South Greece no major event can be associated with the period, in Boeotia the fall of the palace at Thebes has been dated to L.H. IIIA 1 and has been a source of much speculation.