the best proof that the water had only recently reached that locality; and if it stays, it will not be long before vegetation grows up there as vigorous and as dense as that in the Kara-koschun. The flowing water will carry it into all the northern basins; the plants we already observed were the forerunners of its propagation in that quarter. On the other hand, dead tamarisks, with or without mounds, as also old kamisch-stubble, were now becoming more and more plentiful.
At first the sand was not more than 5 m. high, and as a rule the ground between the dunes consisted of hard clay, with jardangs, but it was increasing, and the dunes had already met in some places. The only localities in which the ground was level and bare was close beside the new lakes, where we were travelling. Although these lakes were only a couple of years old, and in their present size only one year old, the effect they had on the distribution of the sand was evident. As, generally speaking, the lakes make their way from south to north, they put themselves as an insurmountable barrier in the way of the drift-sand coming from northeast and east-north-east. The dunes which stand on the west side of the lake receive no further augmentation, but are diminished and transported onwards by every fresh storm, so that in this way they gradually move away from the lake-side, which soon becomes swept altogether free from sand. If the new lakes go on increasing in size, and spreading farther and farther to the north, as they probably will do, and if they remain long enough stationary, they will in the course of a few decennia totally alter the character of the south-west part of the Desert of Lop in which the dunes now are. These will continue their migration towards the south-west and west-south-west, while the belt of sand-free ground beside the lakes will continue to increase in breadth. Nevertheless the mass of the sand will always advance along the same route as that in which it is now moving, until it comes to the bank of the Tarim and the Tokus-tarim. As for the stream of drift-sand which continues to pour in uninterruptedly from the north-east, it will, even as happened to the Kara-koschun during a long period, help to fill up the new lakes, and then, after they too have become overgrown with kamisch, they will be forced to leave their beds and seek lower depressions elsewhere. It is this incessant circulation, which has unquestionably been going on for thousands of years, that not only has levelled up this part of Central Asia, but also explains how it is that the whole of the Desert of Lop and the region of the Kara-koschun lie virtually at the same level, the difference between the highest and the lowest point not exceeding 8 m. Beside the lake where we discovered the canoe and the huts, we fancied we detected some effort towards providing a natural protection against the drift-sand. The little belt of tamarisks which accompany that lake had, at any rate to some extent, arrested the drift-sand, causing it to form a low rampart, which its roots bind and hold together. But even though this rampart, which will increase in height as time goes on, does arrest a part of the drift-sand, the fate of the new lakes is nevertheless sealed, for large quantities of sand and drifting dust do all the same find their way into them.
During the second half of the day's march we passed the same belt as we did the year before. The sand, after once more increasing in height, suddenly passed over into scattered dunes. Then we crossed the zone of dead tamarisk-forest, where the stems of the bushes were more than one dm. in diameter, and after that made our way through
IIIFカラー高解像度
