As we advanced to the west and south-west, the sand continued to increase in quantity. The dunes were at first small and widely scattered, but soon grew higher, running up to 7 and 8 meters. Here again we had confirmation of the law, that the sand increases in quantity from east to west, and from north to south. Here underneath the sand there were always jardangs; but as we were travelling south-west, they occasioned us no inconvenience. They often stuck out of the sides of the dunes, forming platforms and cornices. In the latter part of the day's march, where the sand lay in an almost unbroken sheet, it was evident that it almost entirely neutralised the effects of erosion upon the substratum. But as the erosion gullies in this part of the clay desert are in every case likewise drawn out from north-east to south-west, • it is evident that the wind accomplished its erosive work before the sand was spread out over the desert. This circumstance proves that, after the drying-up of the lake-basin in which the clay deposit was laid down, the bottom of the lake must have been for a long time open and exposed to the wind, before the drift-sand began to accumulate on this side. The depth of the gullies, 2 to 3 m., is a proof that this period of freedom from sand must have been a long one. At the present time, and under the conditions that now prevail, the sand is going on piling itself up in larger and larger quantities in the south-west parts of the Desert of Lop.
Here again we frequently passed pools and sheets of water which had no visible connection with the larger lakes. The water in them came bubbling up from below, and was strongly impregnated with salt.
We pitched Camp No. CLXXII by the side of an offshoot from the lake. The shore of the latter projected towards the west, and there it was quite easy to perceive the flow of the water from the south towards the north-east. Thus there was here a slope in the latter direction, even at such a great distance from the northern bank of the Kara-koschun proper — Prschevalskij's Kara-koschun. Clearly this region corresponds to that part of the surveyed line where we found the surface rising from north to south, the only difference being that, here in the west, there must be a gap or hollow in the threshold (a couple of meters high), which in the east separated the northern depression from the southern depression. The current in the nearest lake, which was most irregular and capricious in outline, showed that the overflow water of the Kara-koschun had not yet reached its most northerly limit, but was still pushing its way on north and north-east. Vast quantities of water are lost in moistening the ground. The little pools I have alluded to are but the skirmishers as it were of the advancing flood, and are soon swallowed up by the expanding lake. The water here had a sp. gr. of i .0069, and consequently was almost undrinkable; even the camels were content with merely sipping at it. As the day advanced, and it grew warm, the water became actually repugnant; but in the morning, after the coolness of the night, the salinity was less noticeable. The neighbourhood of this camp was perfectly barren; there was not even kötäk (dead wood).
On the 28th March we doubled the north-western angle of the new lake; consequently our line of march turned to every quarter of the compass — west, northwest, south, south-east, east, and north-east. At first, after leaving camp, we steered towards the west and north-west, having a number of small lakes on our left, and
IIIF Color HighRes
