The Crusader castle (now called ‘Saranda Kolones’) on the ancient site of Paphos was built and occupied c. AD 1192–1222. It overlies and partly truncates a series of ancient features (tombs, cisterns, wells, church remains, etc.). The layers associated with these, excavated at various times between 1957 and 1985, contain rich deposits spanning a period from the 4th century BC to the 8th/9th centuries AD.

Some 410 pottery items from the pre-Castle phases are presented here, mostly in a series of 16 selected deposits arranged in chronological order. These range from early tomb-groups to stratified well fills and an important destruction deposit of c. AD 650. The final ancient occupation (8th–9th centuries) is marked by the appearance of some lead-glazed wares and some imports from the Umayyad orbit.

Individual items of interest from other layers are appended. Some Hellenistic and Roman imports from Phoenicia and elsewhere are here documented in Cyprus for the first time. The later (Medieval) pottery from the site is reserved for publication in the main report on the castle (forthcoming).

Analyses are given of the clay composition of kiln waste from some production sites, and of excavated sherds of some distinctive wares, as a contribution to a possible archive aimed at localizing the origins of different classes of pottery found in archaeological contexts, and at clarifying trade patterns.

Trench V of the 1927 excavations of the British Academy in the Hippodrome of Constantinople was located just beyond the tomb of Ahmet I, in the direction of St. Sophia. With two extensions (VA and VE) it became a considerable excavation, which uncovered a series of massive piers of brick interrupted by occasional courses of stone. These the excavators identified as part of the Baths of Zeuxippus. In 1928 they uncovered more of the building and, to the south-east of it, part of a separate portico with a large exedra. This second building fitted the identification, as part of the connected gymnasium, particularly when two pedestals were found inscribed with the names of Hecuba and Aeschines, both of whom are known to have had statues there.

Objections to the identification have been raised; notably on the grounds that the excavators found no water-tanks or hypocausts. This particular objection was removed in 1952 by the discovery near by, during the laying of a sewer, of connected structures; for these included a plastered cistern and basins. In a recent review of the topography of the approaches to the Great Palace the identification is accepted; nor does it appear to be invalidated by any of the testimonia relating to the Baths of Zeuxippus which have lately been collected.

Strzygowski, in his short account of the ninth-century church of the Dormition at Skripou-Orchomenos, made no mention of the evidence for a contemporary screen. Some of the fragments were visible in his time, for they had previously been recorded, though not published, by Barnsley. It was not until 1932 that the character of the screen was first made known by Maria Soteriou's study of the building, in which three of the marble fragments surviving from it were illustrated. The repertory of ornament used on these fragments was seen to be identical with that employed on the string-courses and cornices of the church itself, which left no doubt that the screen was an integral part of the building erected by the protospatharius Leo in 873/4. More recently, Professor Grabar has contributed an invaluable study of the style of ornament with which both church and screen were enriched.