282 THE SEARCH FOR THE ANCIENT CHINESE HIGH ROAD [Chap. VIII
and a carefully arranged store of fuel was prepared. Out of the twenty camels available for my own party, which included Afraz-gul and Shamsuddin as well as Tokhta Akhûn and a younger Lop hunter, eight were to carry ice, four fuel, and the rest what indispensable baggage we had brought and the provisions which were to last us to Tun-huang. For the corresponding needs of Lal Singh and his three men, Abdurrahim's five sturdy camels provided reliable transport.
The task allotted to Lal Singh was to survey the north and north-east shores of the great salt-encrusted basin which represents the fullest extension of the ancient Lop sea, together with the barren hill ranges of the Kuruk-tagh overlooking its shores on the east. His task was somewhat lightened by the fact that for the first two marches he could count upon finding ice and a little grazing at the salt springs of Yetim-bulak and Kaurük-bulak (Map No. 32. A. 2, 3), the last springs to the east known to Abdurrahim. After carrying out his survey he was to meet me at Kum-kuduk, one of the wells on the caravan route leading through the desert from Tun-huang to Miran (Map No. 32. D. 4)
The plan I had long kept before me and which the time had now come to carry out was to search for the ancient Chinese trade route towards Tun-huang, the Wei lio's ` route of the centre ',3 from the point where it left the edge of the once inhabited Lou-lan area. Thence we should endeavour to trace it over whatever ground it might have crossed, right through to where it probably struck the line still followed by the caravan track leading from Tun-huang and the western end of its ancient Limes to the southern shore of the Lop sea and thus on to Miran and Charkhlik. The combined geographical and historical interest of this task appealed to me greatly, and the knowledge of the serious difficulties and even risks attending its execution could in no way detract from its fascination.
The discoveries attending the successful search made to the north-east of the Lou-lan station had supplied me with what promised to be a safe starting-point for our quest of the desert route, abandoned though it had been for close on sixteen centuries. But the foretaste then gained of the ground ahead made it also certain that we could not hope for water, nor, over most of it, for fuel to melt our ice with, before striking the Tun-huang caravan track in the vicinity of Kum-kuduk. It was a matter of some ten days' hard marching, so far as our former surveys enabled us to calculate the distance. There was a limit to the endurance of our brave camels, and with the heavy loads of ice, fuel, and provisions that had to be carried for the sake of safety, I could not expect the animals, already severely tried by the preceding weeks' work, to hold out for more than ten or twelve days without grazing or water. It was impossible to foresee what physical obstacles might be encountered and might delay us in this inhospitable wilderness, now more barren, perhaps, by reason of the total absence of water, than any similarly large area of our globe.
There were other disquieting aspects presented by the problem of hitting on the exact line of the ancient route and of tracing it through on ground which long before the dawn of historical times had ceased to offer any possibility of human occupation. Few, if any, vestiges of human passage and activity could have survived the destructive force of wind-erosion. There would not be time to make careful search over extensive stretches for any relics left behind by the traffic which had once passed through this region. The indications preserved by our Chinese sources as to the line followed by the ancient route were far too few and vague to offer definite guidance ; though once the route line was traced, there might be hope of locating the few points they mention along it. Accordingly, in spite of all the care I had devoted to the study of the problem, I could not help realizing that our success depended largely, if not mainly, on good fortune.
3 See Serindia, i. pp. 418 sq. ; ii. pp. 555 sqq., for an results previously recorded of the search.
analysis of the Chinese notice of this route and the main