THE SAND-BURIED RUINS OF CHIRA 173
valent of modern Dumuka. The southern site, called by my guides Derevzeh Dung, is unimportant. It occupies nearly the same location as Stein's small Aktaz, with which it is probably identical. At the next site, Kuk Jigdeh (Green Eleagnus Tree), as well as at Kushkusteh Dung, the one farthest to the north, we found some little clay heads of Buddha and some plaques with typical Buddhist figures, which show that the sites antedated the Mohammedan conquest in the tenth century. The other site, Khadaluk, appears to have been the centre of the old town. In two places we found abundant votive tablets with heads of Buddha, and many fragments of painted plaster and gilded stucco, evidently the remnants of an old lamasery or temple. Apparently here, as at Dandan-Uilik, the most permanent structures, and probably the ones last to be abandoned, were of a religious character. We also found several Chinese coins, dating from early in our era, some fragments of paper bearing records in Brahmi script, and two pieces of wood covered with the characteristic Kharosthi script of the first three centuries of the Christian era. One of these (B, in cut, page 204) bears on the reverse side paintings of a camel and other objects. Evidently, the four sites just described are parts of an agricultural district at least four or five miles long, and quite as large as modern Dumuka. The final abandonment of the ruins certainly took place before the Mohammedan conquest in 1000 A. D., and perhaps earlier.
About eight miles north-northwest of Khadaluk, at Payet-Begning-Ilesi or Tetti-Gerim, I discovered another little site, with remains of tamarisk-walled houses, abundant crude pottery, a few beads, and a bit of blue glass, but
IIIFカラー高解像度
