fore Rawak was abandoned, about 300 A. D., the cultivated area was but little, if any, smaller than that of modern Chira, which is estimated to have a population of ten thousand souls. Moreover, the ruins lie in the midst of a great expanse of dead vegetation, extending a dozen miles to the east, and an unknown distance in other directions. I traversed beds of dead reeds and poplars for six miles south of Dandan-Uilik, and could see dead trees extending indefinitely to the north and west. The reeds were broken off close to the ground, and were largely covered with sand. One scarcely noticed them; the poplars, on the other hand, stood up boldly. Near the ruins, dead tamarisks were numerous. A few stood on mounds; but the majority were on the level plain, just as they grow to-day near rivers or in the moister parts of the zone of vegetation, especially in the region immediately around the modern oases. I found no indication that the Keriya River ever flowed to DandanUilik. Nevertheless, in ancient times the place received a supply of water large enough not only to irrigate the oasis, but to support abundant vegetation over the plain on every side. Apparently, in those days Dandan-Uilik stood among surroundings practically identical with those prevailing to-day around the villages of the zone of vegetation.
It has often been assumed by writers on the Lop basin and on deserts in general, that the encroachment of sand is the chief cause of the death of vegetation and of the abandonment of oases. Their theory holds that deserts by their very nature devour the regions around them. Doubtless this occasionally happens, but in the Lop basin, at least, it is the exception, not the rule. Usually the vegetation dies
IIIFカラー高解像度
