________________
AUGUST, 1917)
A THIRD JOURNEY OF EXPLORATION IN CENTRAL ASIA
167
The exact antiquarian evidence here obtained has its special value, because it enables us to date a variety of physical features which I could observe in the immediate vicinity of the ruined settlement. They throw fresh light on the hydrography and early occupation of this part of the Lop-nor region during historical times and those immediately preceding them. For the latter the abundant finds of stone implements, such as Neolithio arrowheads and jade oelte, which were picked up from the eroded surface of the ground near these ruins afforded a very useful guide. The fact that these finds of stone implements continued over most of the wind eroded ground up to the Lou-lan site had a significant bearing on the so-called "Lop-nor problem," the discussion of which has long been carried on without an adequate basis of surveys.
It was similarly important that on the two long marches which brought us there we met & succession of ancient river-beds all lined by rows of dead Toghrak (wild-poplar) trees, and clearly recognizable by their direction as having branched off from the “Dry River" skirting the foot of the Kuruk-tagh. It was plainly a considerable delta, not a large terminal lake, which had existed here during the historical times accessible to antiquarian evidence, and our new surveys have shown how far it extended south and south-west. Finds of Chinese Han coins and of small metal and pottery fragments of undoubtedly the same historical period mingled freely with those of the Stone Age, just on the ground where (according to a recent theory) we ought to have been crossing the position assumed for the Lop-nor of the epooh when Lou-lan was oooupied.
It was long after nightfall on February 10 that we struggled through to the old Chinese station marked by the chief ruins of the Lou-lan site. It was very trying ground we had to cross all day, out up by wind erosion into an unending succession of narrow and steep olay terraces all running east-north-east to west-south-west, the direction of the prevailing wind, and very difficult for the camels to pass. From our base camp at the foot of the familiar Stupa ruin I pushed out reconnaissances into the unknown desert to the east and north-east, while keeping my diggers at work on deeper deposits of refuse, etc., which had escaped attention during the stress of our previous visit. Among the numerous finds of ancient documents on wood and paper which rewarded this clearing, I may specially mention one, unfortunately fragmentary, which shows a script as yet unrepresented among all our former collections. The rest were in Chinese, Kharoshthi, and the Iranian language known since my finds of 1906-07 as Early Sogdian.
Quite as interesting to me were the series of close observations I was able to make on ground immediately adjoining the ruins, as to the levels at which the process of denudation and wind-erosion had been arrested from time to time by a temporary return of moisture and desert vegetation affording protection to the soil. These clearly showed that the process, striking as its effects everywhere are, had been neither constant nor uniform during the sixteen hundred years which have passed since the abandonment of the station. Hence & mere line of levelling carried across areas which wind-erosion has affected in such different ways, could not, in the absence of dateable marks in the shape of structural or other remains, be expected to yield reliable outlines of the hydrographic configuration of the ground at earlier periode.
But the chance for more exciting work came when I could follow up what the reconnaissance surveys, carried out particularly by Afrazgul Khan, my young Pathan surveyor, with great zeal and intelligence, had revealed towards the north-east. There on ground wholly untouched by human feet for so many centuries, I had hoped to find ruins near what