74 FROM KASHGAR TO THE KHOTAN RIVER [Chap. III
bâshi garrison in order to keep watch over a pass in the range beyond ; for in pre-rebellion times Kirghiz raids had been carried out through this pass upon Marâl-bâshi and the high road" This ridge of Shikarwai has the same NW.—SE. bearing that marks off the other isolated desert ranges so sharply from the Tien-shan system, and also shows signs of the same far-advanced erosion.
Arrival at As we moved on October i8th across the flat alluvial plain to the south-east between low
Marâl-bâshi. widely scattered tamarisk-cones, we came, after marching about a mile and a half, upon two ruined huts roughly built of Toghrak trunks. They recalled similar but better-preserved structures that I had seen in 1906 at Bilel-konghan.i8 It was impossible to make out whether they had served for shelter on an old route or as a shepherd station. No grazing exists now anywhere near this point, and when at last, about eight miles farther, the first reed-beds were reached, they proved to be all dead, occupying small terraces around which the ground had been lowered, and was evidently still being lowered, by wind-erosion. Living reeds and scrub, however, made their appearance as the northern edge of Marâl-bâshi cultivation was approached, and this we reached about four miles farther on at the tiny hamlet of Jaren-tola. Beyond it another eight miles' ride, mainly across ground left uncultivated and overrun by low jungle, brought us to my old camping-place in Ayab Mirâb's garden, east of the tumble-down Bazar that marks the town of Marâl-bâshi.
Preparations for desert crossing,
SECTION II.—OLD REMAINS AND ROUTES BEYOND MARAL-BASHI
A short halt was imposed at Marâl-bâshi by the need of careful preparations for my intended attempt to reach the Mazar-tagh on the Khotan river by a short cut through the Taklamakân. From the experience of previous expeditions I could form some estimate of the formidable obstacles and the risks presented by the wide intervening belt of absolutely waterless drift-sand desert. The short desert journey just behind us, though over far easier ground, had served as a useful test of equipment and transport. Guided by it, I decided to lighten our impedimenta still further by sending all baggage beyond an absolutely necessary minimum to Khotan by the caravan route via Yarkand, and to reduce in the same way my camp to a few indispensable followers. By thus keeping most of our fine camels available for the transport of water in my six galvanized iron tanks and the forty odd goatskins I had brought from India, I could hope to overcome the difficulty about water. The approach of autumn weather, cooler than that experienced on our marches from Kashgar, would help our camels to face their long fast and the other trials before them. There was thus compensation for the few days' delay which the wait for our runaway camels was bound to involve in any case.
While the camels were given a good rest with such abundant grazing as the vicinity of Marâlbâshi affords and our apparatus for the carriage of water was being thoroughly overhauled and tested by Naik Shamsuddin, I was able to spare two and a half days for a visit to the hill range of Lal-tagh and its reported ancient site. Starting on October 21st with hired pony transport and a small number of labourers, I moved first to the north-east across the village tract known as Nôr, which extends beyond the left bank of the Kashgar river-bed. Its name is derived from the peculiar system of wooden conduits (nor) by which the irrigation water supplied, here as elsewhere in the Marâl-bâshi area, by the Yarkand-daryâ is carried across the deep-lying Kâshgar-daryâ. I have already before had occasion to refer to the difficulties that impede intensive cultivation around Maral-bàshi. They consist partly in the inherent drawbacks of an irrigation system dependent on a canal head far below the spot where the river debouches and ever threatened by a shift of its bed ; and, probably to an even greater degree, in the deficient agricultural aptitude
17 Cf. Desert Cathay, ii. p. 426. 18 See Serindia, i. pp. 272 sq.