| || || || || || || || || || || |
182 REMAINS OF AN ANCIENT DELTA [Chap. VI
was now completely dry. But a deep hollow near its northern end, where we encamped and where before we had found a pool with water drinkable for animals, still held ice tasting slightly brackish. When in March, 1915, Afraz-gul returned to the Chainut-köl from the north, he saw its basin being filled again' by flood water running in numerous channels from the side of the Kakmak-chash lagoons.
On the exceptionally clear but bitterly cold morning of February 4th, standing on the summit of a conspicuous tamarisk-cone near our camp, I could discern far away to the south the outermost range of the Altin-tagh and thus fix our position on the plane-table by some peaks to the west of B5.sh-kurghan that Lai Singh had triangulated in the autumn. The day's march, lying over ground easy for the camels, was a long one and led straight to the north-east along a line which was practically the same as that followed in 1906 between Camps 119 and 121. The physical features observed on this ground have been fully described by me in Serindia,4 partly from the notes taken on my second passage. I may therefore restrict my remarks here to such changes as struck me after the seven years' interval. They were mainly the result of the complete drying up of the Yangi-köl depression which Dr. Hedin in 1900 and 1901 had seen occupied by an extensive system of lagoons. In 1906 some low-lying portions of this area still retained numerous small lakes and pools filled with salt water ; but now the beds of all these were found quite dry, except for boggy patches here and there and a few small pools between the basins of Tokhta-Akhûn-ku-atkan-köl (a name given in honour of our hunter-guide) and the Juduk-köl (Map No. 30. c. 1). Their water was so briny that no ice would form on them.
Round the largest of all these dry basins farther north, the Kurba.n-kullu-köl, I noticed a regular fringe of tamarisk-cones almost all dead, marking an old shore-line that had probably been maintained for a long period. In 1906 the interior of the basin was found quite bare of vegetation. But now prolonged dryness had allowed young tamarisks to take root round patches still retaining subsoil moisture, and if these tamarisks survived they would in time grow to form a fresh but much smaller ring of cones. Beyond this basin the route passed first among a belt of small tamarisk-cones, all living ; then among rudimentary dunes and then on to ground showing, as already described, the effects of wind-erosion, but manifestly of comparatively recent date. Large patches of the surface were found here still protected by closely matted stalks of dead reeds, all laid flat in the prevailing wind direction from ENE. to WSW. Wherever this covering had been blown off, the bare clay below was cut up by shallow Yardang trenches all running with precisely the above bearing, as they do over the whole of this desert area. That none of these trenches were cut deeper than 3 to 4 feet sufficiently proves that wind-erosion on this ground, close to the northern edge of the Tarim terminal basin as we know it, is not likely to have been at work for more than a few centuries.
Fully in keeping with this was the curious fact that in the midst of this desiccated ground now exposed to wind-erosion there were patches of living reeds occupying small depressions. The most probable explanation is that on the occurrence of exceptionally high floods in the Tarim, water penetrates into wind-eroded depressions of this transitional belt between the true marginal area of the Tarim delta and the utterly barren desert to the north, once watered by the delta of the Kurukdarya. A similar explanation may account for the young tamarisks that we found here growing abundantly on small cones, only 2 to 4 feet high, on level ground within what looked like small wind-eroded basins without any trace of recent moisture. The fact that the direction of these small depressions containing living vegetation coincided roughly with the SW.—NE. bearing of our route supports the above explanation of their origin and incidentally helped to facilitate our progress.
a Cf. Serindia, i. pp. 352 sq. ° See ibid., i. p. 353.