The peculiar features of the ground just described appeared still more pronounced during the next march. All day we followed a narrow strip of marshy ground plentifully covered with tamarisk, reeds, and scrub, and unmistakably marking the edge of an old lake basin. The bare salt-encrusted plain stretched unbroken to the north of the track, but the terrace-like fall towards it grew less distinct after the first four miles, and in parts disappeared altogether. Curiously enough the ground adjoining the marshy strip from the south showed now a similar crust of hard salt cakes, and was so difficult for the animals to move over that we had to keep to the track, water-logged though it was, almost throughout. A slight rise in the ground level was generally observable between the marsh belt along which we were moving and the barren salt steppe southward ; and as the hardness of the salt crust increased the farther off I reconnoitred from the line of springs and of the route, I took this for an indication that the level of moisture here had been gradually shrinking.
In any case I felt justified in assuming that this narrow belt of marsh and vegetation, which we skirted for about thirty miles altogether, marked a ` shore line ' of far more recent date than the one which we had followed along the littoral clay terraces up to Koshe-langza. In this connection I may mention that the elevation indicated for Panja by our aneroid observations, 2400 feet above sea-level, was the lowest noted in the Lop-nor basin. The water in the marsh springs was brackish, but not altogether undrinkable for animals. Plentiful tracks and droppings amidst the scrub along the route showed that wild camels, deer, and other animals used to come here from the foot of the mountains for water and grazing. The presence of the springs themselves is fully accounted for by the direction of the Altin-tagh drainage, which, as the map shows, sends a main bed, no doubt mostly dry, into the sandy desert just south of this portion of the route.
The level bareness of salt steppe and marsh had been relieved only by a single isolated clay terrace which we passed about eight miles from Panja. But on approaching the end of the day's march we found the
IIIF Color HighRes
