Sec. i] THE RUINED FORT OF L.K. 183
Beyond a lake basin of some length (Map No. 29. c. 4), which still retained evidence of recent inundation in the form of small moist patches, we reached a strip of ground where beds of living reeds and thorny scrub stretched for some distance below a line of high tamarisk-cones. Tokhta Akhûn had previously marked this spot as the last where our camels could find some scanty grazing, and there we halted for the night, having covered some 18 miles on the march.
On the morning of February 5th I ascended a tamarisk-cone, fully 40 feet high and still showing live growth on the top;- close to where our Camp xc stood, and from it with the binocular I made out, far away to the NNE., the ruined fort, L.K., which Tokhta Akhûn pointed out as the place he had visited. To the west there extended unbroken an area of high dunes, the same, no doubt, that I had crossed in January, 1907, on my march from the Lou-lan Site to the Tarim. To the north there lay before me an utterly desolate landscape of the type I remembered well around the ruins of Lou-lan. The flat expanse of bare clay, cut up by wind-erosion, was only broken here and there by a few scattered cones with dead tamarisks, and at rare intervals by strips of light drift-sand where lines of dead Toghraks lay fallen in rows.
We passed the first of these rows of dead trees, clearly suggesting the vicinity of an ancient bed with running water, within half a mile of our camp. Here a salt-encrusted depression marked the last living vegetation. Beyond it the ground became greatly eroded with Yardangs big and small, in the midst of which was a long winding depression holding near its centre a small salt pool.e I t was manifestly a part of the last dry lake-bed we had crossed in 1906 south of Camp 121, and the little pool was the last shrunk remnant of those which we had then found there.' From a dead tamarisk-cone near the northern end of this bed the ruined fort L.K. to the north now became clearly visible at a direct distance of only three miles, and to the WNW. of it another smaller ruin the existence of which Tokhta Akhûn had discovered on his recent reconnaissance.
More than a mile from the same point our route passed a curious wind-eroded hollow fully 25 feet deep, with its bottom moist and showing salt efflorescence. It illustrated in a striking manner how subsoil water from depressions farther south, to which an exceptional flood from the Tarim had extended, might percolate to ground that, owing to its total want of moisture, had for many centuries been subject to wind-erosion. The observation was of interest as showing that in this area a belt of depressed ground, such as Dr. Hedin's levelling had shown south of the Lou-lan Site, does not necessarily mark an old lake-bed, but may be the result of long-continued wind-erosion.
We were still about two miles and a half from the ruin L. K., for which we were steering, when relics of the stone age were first met with on the wind-eroded soil. They were miscellaneous small pieces of worked stone (L.K. 073-8, 0130), and were very soon followed by abundant further fragments of stone, together with remains of coarse pottery, all of the same type as those found seven years earlier on our first march over eroded ground beyond Camp 121.8 As the position of the latter lies about three miles to the east of L.K., these finds afforded conclusive proof that the whole belt of ground here was occupied by man during some period or periods of the stone age. The various stone remains picked up on our approach to the site of L.K. will be found briefly described in the List below (L.K. 085, 0111-12, 0117-20, 0127-30, 0135-J4, 0155-62), but have not yet undergone expert examination as to material and make, such as Mr. Reginald Smith has bestowed