National Institute of Informatics - Digital Silk Road Project
Digital Archive of Toyo Bunko Rare Books

Wind-erosion watched at work.

36o   THROUGH THE LOP DESERT   [Chap. X

patches of dead reed-beds covered the top of Yârdangs. On my route of 1906 they were met between four and six miles to the north of Camp 122 (about 40° 21-22' lat.), and on my journey of 1914 I passed a similar belt of dead reed-beds, some three miles further west, in about the same latitude. In both places the appearance of the low but thick stubble of dead reeds struck me as not very ancient. The very different levels on which they were found, varying as much as eight feet according to the greater or lesser height of the Yârdangs above the same wind-eroded trough, suggested that these reed-beds had grown up during a temporary and somewhat recent submersion of the ground after it had already undergone the effects of prolonged wind-erosion. It has since occurred to me that the fact of Dr. Hedin's line of levels having shown the bottom of the above depression to descend, over a portion of its width which he roughly estimates at about 4 km. or two and a half miles, to an average of one metre (3' 3.41 below the level of the Kara-koshun during the spring-flood of 1901,13 may possibly furnish an explanation. During a period of such excel:). tional floods water might have found its way from one direction or another into this belt, more deeply eroded than the rest of the Yârdang area south of Lou-lan, and remained long enough to cause a temporary growth of reeds, which, however, were bound to die again when those floods finally ceased.

The other point to be noted is that nowhere on my two crossings of the belt corresponding to Dr. H edin's depression did I come upon ground showing the hard salt-encrusted surface which invariably marks the bottom of dried-up ancient marsh beds in the Lop region. Nor was any other form of salt-impregnation met with. Further to the east, beyond Dr. Hedin's line of levels, a vast continuous area of the Lop desert was proved by our surveys of 1914 and 1915 to be covered by such a hard crumpled salt crust, and I have strong reasons to believe that the beds of the Kuruk-darya delta, during the historical period of Lou-lan occupation, carried their water to its edge.

During the night from the 15th to the 16th of December we experienced for the first time the blasts of the icy north-east wind of the Lop desert, which, unlike the winds affecting the Taklamakân, does not relent for long even during the winter, and which continued to hold us in its clutches during most of our stay in this region. Its erosive effect on the Yârdangs could clearly be watched, during the day's march, in the steady drift of sand which was undercutting the clay banks. The sand seemed to become, from here onwards, of a slightly coarser and heavier grain, and as the velocity of the wind was not great enough to raise its particles high, I could, before leaving camp in the morning, for the first time sight the reddish-brown line of the Kuruk-tâgh foot-hills far away to the north. Our progress towards it lay that day over ground bearing the same general character as that crossed on the preceding march. The closely packed Yârdangs showed the same uniform direction from east-north-east to west-south-west, and their tops were scored with furrows reproducing the same surface configuration on a small scale. The interrelation between this configuration of the ground and the wind which is its creator was brought home to us with painful directness by the fact that nowhere, even in trenches cut down to a depth of twelve feet, could the slightest shelter be found from that freezing blast. I particularly noticed that at the south-western end of the terraces, where one was naturally tempted to seek for some protecting bank, the cutting force of the wind was even increased. There the Yardangs invariably ran out into a gradually sloping and narrowing tail-like end, where piercing currents of air met from both flanks of the Yârdangs, as if in an eddy, and carried on the work of abrasion with additional force.14

Absence of salt-crust.

Cf. Central Asia, ii. p. 326.

14 This very phenomenon, carried back into an earlier geological period, accounts for the peculiar appearance of the

salt-coated terraces or Mesas to which the early Chinese records of the desert route to Lou-lan apply the graphic designation of ' White Dragon Mounds ', and the position of