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

Sec. i]   THE ROUTE FROM ABDAL TO TUN-HUANG   551

receives practically the whole of the Su-lo Ho drainage (Map No. 74. A. 3, 4). The characteristics of this second section of the route, over 96 miles long, are clearly determined by the fact that, instead of skirting as the first did the utterly desolate shores of the dried-up salt sea, it leads along a great valley. Desert ground as it is, it carries enough subsoil moisture to feed wells of drinkable water at numerous points and close to the surface, and also to maintain more or less continuous belts of reed and scrub growth. The soil is sandy throughout, no longer shör or gravel, and the desert vegetation it supports steadily increases after Yantak-kuduk is passed. Fresh water can be found by digging within a few feet of the surface at most places as far as Besh-toghrak. There reed and scrub grazing is sufficiently abundant, and even a few stunted wild poplars may be seen, which account for the name, the ` Five Poplars ', now given to it by the Lopliks. As we advance from Achchik-kuduk north-eastwards the valley gradually narrows. The long dune-covered ridge on the south approaches closer and closer to the foot of the barren Kuruk-tâgh range ; this rises here to about 1,500 feet above the valley bottom, which at Besh-toghrak contracts to only about five miles in width.

The second section of the desert route as far as Besh-toghrak offers none of the serious difficulties encountered on the first. To the east of Besh-toghrak, however, the character of the ground undergoes a notable change. There, after a distance of about five miles, a belt of dunes rising to 40-50 feet in height is encountered, and after crossing this the route strikes the westernmost of a series of depressions constituting a dried up terminal basin of the Su-lo Ho. In my Personal Narrative a detailed account has been given of the interesting physical features encountered on crossing this basin to a point near its eastern head.° In it I have also explained the special interest attaching to the geographical questions which those features raise.10

The presence of subsoil water within easy reach, which alone makes the valley descending from Besh-toghrak to the eastern extremity of the ancient Lop lake-bed practicable for traffic, is directly due to the fact that a certain portion of the Su-lo Ho drainage, at least during its big summer floods, must still find its way somehow, probably underground, into the depressions of its earlier

Subsoil water of desert valley.

Wells of Béshtoghrak.

Ancient terminal basin of Su-lo Ho.

Subsoil water from Su-lo Ho drainage.

s Cf. Desert Cathay, i. pp. 532 sqq.

10 The geographical importance of this ground induced me to make further surveys there on my next passage in March, 1954. Their results, while confirming my former conclusion as to the general character of this area, necessitate the modification of certain details in its cartographical delineation as presented in Map No. 74. A. 3. They are duly shown in the new series of maps, comprising the surveys made on my third journey as well as on the previous expedition. Among them there is one detail of importance which I am glad to have an opportunity of correcting here.

R. B. Làl Singh, when carrying his plane-table survey in March, 1914, under my instructions to the south of the above described series of depressions, ascertained that the final northward turn given in Map No. 74. A. 3 to the present terminal course of the Su-lo Ho, suggesting that its water could eventually reach the dry basin shown south of Camp 153, is erroneous. Rai Ram Singh, when making a reconnaissance survey of this area in May, 1907, had from Camp 174 sketched the Su-lo Ho bed, down to the approximate point marked by the letters su in the map, with very fair accuracy. He had, however, failed to see that the northward turn of the bed was not final, but only a bend followed by a sharp turn to the south-west. This is succeeded by a westward course of about four miles leading to where the river in

1914 terminated in a narrow lake about seven miles long that stretched from north to south and was bordered by shôr belts. It probably connected at the flood season with the lake-bed shown further south in Map No. 7o. n. 4.

In reality the high and well-marked plateau of conglomerate covered with gravel, which in the map is shown extending north of the actual Su-lo Ho bed as far as the northward bend (wrongly assumed as final), stretches right across westwards and joins the ridge correctly shown on the western edge of Map No. 74. A. 3 and continued in Map No. 70. D. 3. The statement made in Deserl Cathay, i. p. 535 (top), requires to be modified accordingly.

It is worth notice that the continuity and width of this intervening plateau makes it appear still more probable that the drainage, of which I observed such plentiful evidence on my examination of the northern basin in 1914, finds its way there through the northern beds of the delta, now ordinarily dry (Map No. 74. B. 3), and not by percolation from the southern terminal basin, as might be otherwise assumed.

In justice to my surveying assistant on the former journey I ought to add that his mistake must be ascribed partly to the very deceptive nature of the ground and partly to the fact that he had to make his survey under exceptionally trying climatic conditions and at a time when his health was seriously affected.