they gave me make me inclined to accept its substantial accuracy. From their fathers they had heard that the basins had before held water for certain periods, and Tokhta Akhûn himself remembered an exceptional flood about 1892, which had carried water into the lagoons, within
a day's march north of Chainut-köl. If such intermittent inundations had occurred for a generation or two past, it seemed easy to account for the thin beds of reeds, dead or living, which we found
over extensive patches of ground here. On the other hand, the total absence of vegetation in certain intervening depressions was attributed by my hunters to the depth of the water which they periodically held.
I may conveniently record here that the observations made on my renewed visit in 1914 proved the drying-up process in the area affected by such inundations from the Yangi-su or Kakmak-chash
bed, as it is also known, to be still in progress. Throughout our march from Alam-khöja-köl to
Chainut-köl, on February 3, 1914, all the lagoons that we saw were dry, though our new route lay considerably more eastward and thus nearer to the feeding bed. The Chainut-köl itself held no
water, except in the pool previously mentioned, and even the ice there tasted very brackish. Beyond Chainut-köl we kept close to the route of 1906 as far as Kurbân-kullu-köl, but saw no more sheets of open water, only boggy patches and a couple of very small salt-pools in the deepest portions of the basins. It was obvious that since about 1903 no exceptional flood had penetrated so far.
Our march on December 14 took us over ground where the increased difficulty of progress for our heavily laden animals was compensated to me by the interest which the observation of some
novel features afforded. The sixteen miles covered that day carried us across what I recognized at
the time as a transitional belt between the true marginal area of the present Tarim delta and the absolutely barren desert northward once watered by an earlier delta. This impression, I may add
at once, has since received full confirmation from the observations which our surveys of 1914 and
1915, carried across the same area and on three different routes all radiating from Kurbân-kullu-köl, have yielded. All day we passed a succession of dry, salt-covered lake-beds, large and small, with
occasional salt-pools in their deepest portions, showing plainly that we were still within the Yangi-köl
depression liable to inundation by exceptional floods. But after scarcely more than two miles from camp we left behind the area where subsoil moisture sufficed to maintain a continuous if scanty
desert vegetation, and two miles further on we were met by the first indication of close approach to that zone of strongly marked wind-erosion which, as I knew from Dr. Hedin's description, constitutes the most striking feature of the northern portion of the Lop desert.
It was a belt of narrow ridges or terraces of hard clay, separated by small nullahs or trenches, as yet only three or four feet deep, but showing sharply-cut banks such as only the erosive action of
wind and driven sand could produce in this region. The top of the terraces or Yârdangs, to use the convenient Turki term adopted by Dr. Hedin, was invariably covered with a network of shallow
parallel furrows, all running like the trenches in the direction of the prevailing winds which had
carved them, from east-north-east to west-south-west. The soil exposed on the sides of the Yârdangs was a hard stratified clay, clearly the sediment of an ancient lake-bed, but showing to the
eye no trace of saline impregnation. Erosion could not have been very long at work here, for in places the top of the terraces still retained a protective cover of matted dead reed stalks, all laid flat in the same wind direction. Elsewhere returning moisture was likely to have temporarily stopped denudation, and this would account for the dead reeds which, as I found in places, had grown between the Yardangs.
Beyond this first outpost line of the wind-eroded clay desert we passed again flat saline ground with dry lagoons large and small. In their deepest portions they held rare pools with water so
1374 z Z