still contained water at the time when Lôu-lan was inhabited — i. e., according to the MSS. which I discovered there, in the 3rd century A. D. — and dried up shortly afterwards, the wind has had a period of i 600 years in which to accomplish its sculpturing work in the clay. In other words, the layer of clay which it has had to remove every year has not exceeded 4 mm. This figure varies of course for different parts of the zone in question; for some of the jardangs are only 4 m. high, and even less. The reason why these flat-topped elevations have not been blown away too is no doubt that which we have already discussed when dealing with the region beside the Kuruk-darja, namely the binding properties of the vegetation. At the present time there does not remain, on their summits, the least trace of vegetation, not even withered roots. From this we may infer, that this part of the lake-basin dried up earlier than the portion beside which Lôu-lan stood. It is conceivable that this region never did possess forest like that on the banks of the Kuruk-darja; but there must certainly have been kamisch, although there is not now a trace of it to be seen. The long clay ridges make it however extremely likely that tamarisks at any rate existed there, for nothing but the presence of binding roots can explain the possibility of certain portions having survived. All traces of former vegetation are now entirely obliterated, and since its disappearance the ridges themselves have been exposed to the destructive agency of wind erosion, although at the same time this agency is operative with greater energy in the grooves. The amount of erosion assumed above, namely 6 m. in i600 years, is therefore far from exact; for not only is it probable that this part of the ancient lake was laid dry long before the 3rd century A. D., but the altitude of the ridges themselves would be higher than they are, had their summits coincided with its bottom-level.
After crossing over all these successive belts of the ancient watery region, we traversed some flat, gravelly hills, between which the dry torrents generally made their way south. In two or three places amongst these hills there were self-contained basins in miniature, into which, as the level, cracked clay surface and two or three small living tamarisks proved, water does sometimes gather, being held up there as in a saucer.
All day on the i 8th February there was a storm from the north-north-east, which drove before it along the ground vast quantities of hard material, coarse sand, and small pebbles. This direction of the wind corresponds at all events with the position of the western clay ridges. At first our track led across a peculiar country, consisting of small, narrow, elongated hills, composed chiefly of gravel, though also of hard rock; their sides were often pretty steep, and they were separated from one another by depressions, in which it was the rarest thing possible to find eroded watercourses or any signs of running water. These depressions, in the bottom of which sand had once accumulated, were in places disposed parallel with one another, and like the ridges in general parallel with the clay terraces or jardangs. Yet this is not true of them all, for those in the east extended north and south, whereas those in the west ran north-north-west to south-south-east. The absence of eroded watercourses and the impossibility of determining in which direction these hollows inclined made it tempting to ascribe their origin also to the erosive agency of the wind; which has of course been operative here an inconceivably longer time than in
IIIF Color HighRes
