Kum-tscheke (r.); Darghalik (1.); Tschapal (1.); Kasan-sindi (1.); Schirdak-uji (1.); Arsu Beki-dschajiri, Turfan-karaul, Basch-toghrak, and Toghri-köl.
From Dilpar three routes lead to Jing-pen: (I) Alongside the river, through forest, round the north-going windings as far as Turfan-karaul, and thence along the usual Turfan road. (2) Across the barren desert, first north as far as Kalta, and thence straight to Jing-pen. This route appears to follow the old bed of the Kontschedarja, though there is of course no water in it. Beside it there are reported to be two or three old toms, or ftotajs, i. e. pyramidal »mile»-posts, which evidently mark, the ancient road from Lop-nor to Korla. This road, the northern prolongation of which I discovered in 1896, ran in its time along the left or northen bank of the united river (Tarim + Kontsche), and then, after the confluence, on the left bank of the Kontsche-darja. (3) The great winding road along the southern foot of the Kuruktagh, which I shall briefly describe in the following pages. None of these routes is of any importance: they are seldom used, and then only by hunters, or very exceptionally by a chance merchant going to Turfan.
The distance between Jangi-köl and Dilpar is 31.6 km. Thus the tract between the two rivers is here pretty broad, and as water is extremely scarce, the Jaman-ilek being the only stream, the vegetation also is excessively poor. Unless the river shifts its bed again, or unless a fresh bifurcation takes place, this strip of land will gradually assume the character of desert.
Almost the same distance, that is to say, 32.7 km., separated us on 7th May from Suget-bulak, which lies N. 28.3° E. of Dilpar. Our route therefore ran at right angles to the river, leading through the toghrak forest of Dilpar, in which, strange to say, several of these fine trees are dead. The biggest stand just on the outskirts of the forest, where it yields to the steppe of scrubby jantak and tamarisks. There is another narrow belt of withered poplars crowning a terraced elevation, parallel to the river, and bordered on the north by an old winding river-bed, the course of which is plainly indicated. On the left bank of the old river-bed there still remain three huts, consisting of poles driven vertically into the ground, and surrounded by kotäk, or »dead trees». A few hundred meters farther on we crossed yet another similar river-bed, it too running parallel to the Kontsche-darja. This again is distinctly outlined, being deeply sunk in the ground; it also has kötäk on its northern bank. The natives have a tradition that several decennia ago the Kontschedarja for a long time used to oscillate between these two river-beds, and showed considerable irresolution before it finally settled down into its existing channel. It is probable that the old desiccated course I discovered at Kalta in 1896 is a continuation of the two river-beds in question; at any rate these last undoubtedly mark the last stage but one in the successive south-westward migrations of the Kum-darja.
Next comes thin tamarisk steppe, sparsely dotted with mounds or else without them. The soil consists of white crystalline, tolerably hard schor, which crackled under our feet, although it was overlain by a thin layer of sand. After that we crossed a narrow zone of jardangs, that is hard, dry clay terraces, sculptured at the bottom of some natural watercourse, in which wind erosion played an active part. The region was almost entirely barren; occasionally we caught sight of a withered tamarisk, seldom of a living one. A third watercourse, not quite so dis-
IIIFカラー高解像度
