CH. IX CHANGE AFFECTING `DRY RIVER' 151
in 1915 to the Kuruk-darya in the desert, farther west where its bed could be clearly followed along the foot of the utterly barren Kuruk-tagh, the `Dry Hills', I found that within it brackish water could be reached by digging shallow wells in some hollows. Hence I could not feel altogether surprised when I learned on my fourth journey in the Tarim basin (I 930-31) that a recent great hydrographic change affecting the course of the Tarim had caused the greater portion of its summer floods to meet the Konche-darya much farther north than before and thus to deflect the united waters of both once more into the `Dry River' and towards the area of ancient Lou-lan. The hoped-for chance of studying this latest change affecting the Lop basin has, I regret, been denied to me through Chinese obstruction.
The time for more exciting work came when by the middle of February I could turn to the main task which had drawn me back to this desolate ground of Lou-lan. It had been duly prepared for by reconnaissances made mainly with the help of Afrazgul Khan, the young but zealous and intelligent Pathan draughtsman who had joined me as a Sepoy from the Khyber Rifles and who has since, by his merits, risen high in the service of the Survey of India. Those reconnaissances, on ground untouched by human feet for many centuries, led to the discovery of a succession of remains to the north-east, which clearly indicated that the ancient Chinese trade and military route, which I was anxious to trace towards Tun-huang through the desert, had followed that direction, at least in its initial portion.
The nearest among those remains was an ancient burial-ground situated some four miles from the Lou-lan station on the top of an isolated clay terrace or Mesa rising some
IIIF Color HighRes
