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

Northernmost hill chain passed.

Descent over Peishan glacis.

Depression of Yen-tun reached.

Dry beds at head of ` Yen-tun trough ',

53o   ACROSS THE PEI-SHAN TO BARKUL   [Chap. XV

direction, while another rather indistinct track continued along the latter north-westwards.' Separating here from Muliammad Yagiab we continued to descend the gradually widening valley. At a point about 8 miles from Ming-shui camp, the bold conical hills of an outer chain receded on the western side and allowed us to sight far away in the distance the glittering snowy crest of the Karlik-tagh. The distance separating us from its nearest point proved to be still fully a hundred miles. ' All the same it was a most encouraging vision and remained a guiding mark on this and the next march.

Farther on the descent continued over a gently sloping stony plateau. After a march of 19 miles we pitched camp at a point where we found a well only two or three feet deep in a shallow bed (Map No. 37. D. 4). Our Chinese called the spot Tung-êrh-shan. A short distance before reaching it we passed through what looked like a gap in an outer chain of hills. Though not rising high above the valley bottom, they looked picturesque with their bold conical forms and much-eroded black slopes (Fig. 284). There seemed to be a good deal of scrub at their foot, and a herd of Kulàns or wild asses was seen grazing on it. One of them was successfully stalked by Afrâz-gul.

On September 21st a long march, impressive in its monotony, took us down, along the same Ming-shui drainage, over the detritus-covered northern glacis of the Pei-shan. The hills, still rising near Tung-êrh-shan to a conspicuous hill-cone about 6,760 feet in height, sank away more and more under that vast covering of detritus and gravel which overlies the northern foot of the Pei-shan mountain system. In the distance beyond the last rocky ridges cropping out from the vast expanse of detritus could be seen the higher line of the first Pei-shan range west of Ming-shui. Saxaul bushes and low thorny scrub were growing plentifully among shallow beds, which, no doubt, at intervals carry occasional floods from the mountains.

Towards the end of the march the valley bottom along which we were moving became more and more trough-like ; it was flanked on either side by well-marked terraces. The track kept on the gravel Sai above them and brought us, after a march of about 25 miles, to where the trough widened into a reed-covered depression about two miles long. In it we found a large spring, flowing open for a distance of about 70 yards and farther on disappearing in a marshy thicket of reeds. The spring, which our Mao-mei informants had mentioned by the name of Chin-êrh-ch`üian, is evidently fed by the subterranean drainage of the Ming-shui valley. Our camp was pitched on a shôr-covered bank above the spring, at an elevation of 4,020 feet above sea-level. The day's march had meant a descent of close on 1,80o feet from Tung-êrh-shan. This, together with the appearance of the ground ahead, left no doubt that we had now reached that extensive depression which the An-hsi—Hâmi road crosses near the station of Yen-tun (Map No. 37. A. 4), and which has been designated in consequence the trough of Yen-tun.2 It marks sharply the dividing line between the Pei-shan and Tien-shan systems and terminates in an as yet unexplored portion of the western Pei-shan.

On the morning of September 22nd we continued our march north-westwards on the detritus covered terrace above the left bank of the wide dry bed. A branch of it coming from the south that had to be crossed close to camp was hollowed out to a depth of 3o feet. The main bed with which it united was close on three-quarters of a mile wide, and after running north-west for about four miles from C. 215 turned to the west. We crossed it about a mile farther down and soon passed beyond it into a maze of small eroded ridges of granite holding little nullahs, filled with scrub, between them. It was the last offshoot of the spur that we had skirted along its western side on the way down. After

1 The point of junction of the two routes was found correctly marked in MM. Grum-Grizhmailo's route map. But, by an erroneous conjecture, the course of the. Ming-shui

bed is made therein to take, farther down, a turn to the west, a wrong course reproduced also in the Futterer map. 2 See Futterer, Wüste Gobi, p. ro.