66
W. SCHOFIELD
small plateau averaging some 10 to 12 metres above sea level. This is a plane of marine denudation dating from a time when the sea level stood 12 metres higher than now, perhaps during the last great inter-glacial period (Riss-Würm), some 100,000 years ago. From this lower area much less clay could be washed than from the higher and steeper hill to the north; and the gentler wave-action on the west beach, normally on the lee of the island, made it about five times as long as that on the east of the isthmus, with very few large boulders. Somewhere on the west side of the 12 metres terrace, between about 1100 and 1500 A.D., there was at one or more times a small settlement, perhaps no more than one or two fishermen's huts; for at this point on the west beach are found pieces of Sung and even Ming pottery lying on the beach and in the cliff, which here is largely built up of coarse rainwash from the hill behind. There is, however, no modern settlement and no cultivation, and the island appears to be used only by boat-people, either for fishing or for burial of their dead; for on one visit Prof. Shellshear, who was with me, discovered a human skeleton of recent date two feet below the top of the sand cliff.
METHOD OF INVESTIGATION
The site was first discovered and investigated by Dr. Heanley and Prof. Shellshear, who worked together from about 1925 in looking for sites showing early human occupation. Much of what they found lay on the surface of the beaches, but wherever possible they noted the depth from the soil surface of objects found in the sand cliffs. Part of their material was presented later to the British Museum, and some to Mr. Eumorfopoulos and others, but the rest seems to have disappeared during the war in 1941 when the Hong Kong University was wrecked. Their code number for the site was 123, which points to a comparatively later discovery: the Tai Wan site on Lamma, for example, is numbered 83.
The technique employed by the writer at Tung Kwu was as follows. Objects not found in situ were collected and the initials of the site were painted on them in Chinese ink. If a single object was found in situ, its depth from the surface was measured in inches or centimetres; it was extracted; the depth and initials of the site were written on it or its wrapping paper, and later were recorded in Chinese ink on the specimen. In 1935, by which