274 THE PULSE OF ASIA
fresher than now, the location of the road becomes highly reasonable.
Two short days' journey northward from Tikkenlik brought us to Ying-pen, the ruins of a little fort and village discovered by Hedin on the ancient trade route west of Lulan. The population was evidently very small, perhaps ten or twenty families. Hedin thought that the water supply came from the Dry River, not far to the south. As the ruins lie sixty feet above the river on the fan at the mouth of the Bujentu valley, the supposition is not tenable. It would require a canal ten or twenty miles long, of which there is no trace; and such a canal would be out of all proportion to the size of the village. The only other alternative is that the water should have come from the Bujentu valley, which leads southward from the Kuruk Tagh or Dry Mountains. Accordingly, I went up the valley, and found a canal, which I followed for two miles to its head. It was simply a ditch dug in gravel, and protected on either side by a row of tamarisks, long since dead. I also found a reservoir with walls of gravel in which the water was stored near two of the shrines called " stupas." Like all the other ancient irrigation systems which I saw, that of Ying-pen differed not a whit from those of to-day. At present, the place is uninhabited. When I questioned some antelope-hunters who often visit the region, they said that where the Bujentu valley leaves the high mountains, there is an old fort called Shindi, on an ancient road, apparently, from Lulan to Kara-Sher. At this point the valley always carries water, which reaches Ying-pen only in flood. There is not enough, however, to support a single family. If there were, some of the people
IIIFカラー高解像度
