Sec. iii] ACROSS THE SALT-ENCRUSTED LOP SEA-BED 297
the rounded shôr-covered slopes bear out the assumption that wind-erosion was still here carrying on effectively its work of paring and scouring. It has since occurred to me that there might once have been shallow recesses or terraces cut into the sides of the ridge, to provide some kind of modest shelter, such as I had found on the slopes of the similarly isolated clay hillock, T. xiv, once occupied by quarters at the ancient Jade Gate of the Limes.3 The small relics might have been left there by wayfaring occupants. The gradual disintegration of the clay would obliterate such semi-troglodyte hovels, while at the same time erosion, proceeding however slowly, would expose any hard fragments once hidden away on their floor. But all this must remain purely conjectural.
Another problem, and one which at the time was bound to engross me far more, was presented by the outlook from the top of this Mesa. To the east, to which I turned with natural eagerness for a sight of the ground over which the continuation of our supposed route line would take us, the view was obstructed, beyond half a mile or so, by what seemed an unbroken maze of high parallel ridges, Yardang-like in shape, but all glittering in a white coating of salt. The open bed of the dried-up sea I was looking out for was entirely masked by them. The view of this belt of forbidding salt-coated terraces at once brought into my mind the dreaded ` White Dragon Mounds ' mentioned in all the early Chinese accounts of the direct Lou-lan route. How we should make our way through it was uncertain. So much, however, was clear, that the choice of the little Mesa on which I stood for a halting-place had probably been determined by the very nature of the ground which now faced us. There could be little doubt as to the reason for which it had been selected. At its foot was the first piece of ground, level and tolerably clear of salt, which ancient travellers moving westwards would strike after the trying march through the bed of the dried-up sea, with its blocks of hard salt, as described in Li Tao-ytian's notice quoted hereafter. The equally forbidding maze of white salt-encrusted ridges which now faced us and manifestly marked the proximity of its shore had also to be crossed before they could camp in comfort.
The encouragement derived from our fortunate discovery on the Mesa induced me to continue the eastward direction of our march. The soil turned rapidly into shôr, and as soon as we had passed into a depression masked by the outermost line of salt-coated ridges, this assumed the form of crinkled cakes of hard salt suggesting the petrified ripples of an inlet from the great dried-up sea. This surface was so trying for the camels, that after covering about a mile and a half from the Mesa I had to change the bearing to N.8o° E. in order to reach ground where easier going was offered by patches of soft brown shôr, i. e. salt-impregnated clay, between the salt-coated ridges. Among these patches I noticed small lakelet-like depressions in which large flat surfaces of pure salt were broken up by cracks into more or less regular pentagons. The position and flatness of these salt surfaces suggested that their formation was originally the effect of moisture. Their disruption, while drying, into cakes with crumpled edges and the subsequent corrugation of these cakes through contraction after renewed access of moisture seemed to offer a possible explanation of the origin and character of the process ; operating on a large scale both as regards time and ground, it has covered vast stretches of ancient sea bottom, such as we subsequently encountered, with solid blocks of hard salt, heaved up and contracted into endless blisters and confused pressure ridges.4
Thus we toiled on painfully for two more miles crossing line after line of salt-encrusted hillocks, all drawn out parallel like Yardangs, but stretching from NNE. to SSW., and thus almost at right
3 See Serindia, ii. pp. 684 sq. ; also p. 72r, concerning another such shelter cut into clay at tower T. xxiii. a.
4 The above-recorded observation fully bears out Professor Huntington's explanation of the roughness of the