230 THE DESERT OF LOP.
The sand has no doubt been prevented from accumulating over certain areas by the gullies which existed prior to its advent, and these areas are surrounded by dunes. Nevertheless I have been unable to hit upon the precise cause of the relation to which .I have alluded. At all events the jardangs in this part of the desert are quite insignificant, though their step-like terrace formation comes to light again in the bajirs. The bottom of each bajir is covered with a layer of fine dust an inch thick, excessively soft, pôwdery, and perfectly dry. It requires but a very slight puff of wind to set it in motion, so that it appears to settle afresh after every storm, only to be driven on farther by the next storm that comes. Possibly the presence of these bajirs depends solely upon trifling differences in the surface relations, and they indicate nothing more than places which are avoided by the drift-sand. At all events they do not owe their existence to the same cause as the bajirs of the Desert of Tschertschen, and their outlines are less regularly drawn. Their regional distribution is also very irregular. Sometimes we would travel a good distance without encountering anything but sand, then again the bajirs would lie close together. Farther south the bottom of some of them was overgrown with kamisch, of course withered and cut down to stubble centuries ago; it lay over towards the south-west, as though it had been brushed down in that direction. This kamisch would hardly seem likely to have anything to do with the origination of the bajirs, for dead vegetation is rather a hindrance to the sand and more likely to initiate the formation of dunes than of bajirs. Several of these small bajirs are however beginning to fill up with sand; in which respect again they are very different from the large bajirs in the northern part of the Desert of Tschertschen. Upon climbing to the summit of one of the higher dunes, we saw that the sand grew higher and higher towards the
. south-west, but lower and lower towards the east and south-east; and indeed on our more easterly route in 1901 we encountered only very small quantities of sand.
Towards the close of the day's march the survivals of ancient vegetation became rarer and rarer; and although the mollusc shells still occurred, they were fewer than formerly. Of tamarisks we saw only two living bushes, both small.
. On the 31st March the desert grew more desolate, and the sand higher in the
direction in which we were travelling. The bottom of the bajirs was now covered with sharp-edged crystals of gypsum, agglomerated together, and with tiny cylinders of ' sand and lime cemented together, which had been formed round the stalks of reeds. Pieces of dry timber were excessively rare, and of living vegetation only two small tamarisks; and apart from these not a single wind-driven leaf. It would have been useless to dig for water, the ground was much too dry. Thus the farther we advanced southwards, away from the northern depression which once contained the lake of Lop-nor, the fewer grew the remains of the ancient poplar forests, and of the tamarisk and kamisch steppe. The day's march brought to our notice the trunk of only one solitary poplar.
Limncea shells were still abundant everywhere, though mostly in the shape of thin fragments scattered amongst the sand; they seemed to have been transported thither by the . wind from farther north. The sand was now rather heavy, although the dunes seldom exceeded 6 m. in height. Yet it is not evenly distributed, but appears to be arranged in huge waves, so that in some places it is heaped up higher
IIIFカラー高解像度
