________________
August. 1917]
A THIRD JOURNEY OF EXPLORATION IN CENTRAL ASIA
169
There could be no doubt that the fort dated, like the Tun-huang Limes itself, from the first military advance of the Chinese into the Tarim Basin, about 104 B.C., and that it represented, as it were, the bridge-head of the desert route by which that advance was made possible. I had become so familiar with that ancient Limes and the technical skill displayed in its construction that I could not help rejoicing at the way in which this work from the hands of the same old Chinese engineers had withstood the attacks of that most formidable enemy in this region, wind-erosion. The walls of reed fascines had nowhere been seriously breached, while inside the circumvallation the force of the wind has worked terrible havoc, scouring out big hollows down to 20 feet and more below the ground-level and reducing a large central structure to a bare clay terrace strewn with scattered débris of timber. Under the shelter of the north wall, however, refuse heaps had survived, and these yielded Chinese records on wood and paper.
Beyond this fortified Chinese station other remains were traced. Of these it must suffice to mention a small ruined fort which occupied a commanding position on the narrow top of a precipitous clay ridge fully 100 feet high. It had evidently served as a stronghold and look-out post for some chief of the indigenous population of Lou-lan. Of the type, habits, and civilization of the Lou-lan people, as the Chinese found them on the first opening of the route through the desert, the Han Annals have preserved some curious notes. The accuracy of these was illustrated in a most striking fashion by the examination of the graves covering the other end of the clay ridge. Here we found the bodies of men and women, probably members of the old chief's family, in a truly wonderful state of preservation, due, no doubt, to the absolute dryness of the climate and the safe elevation of their resting-places. The peaked felt caps of the men decorated with big feathers and other trophies of the chase, the arrow-shafts by their side, the simple but strong woollen garments fastened with pins of hard wood, the neatly woven small baskets holding the food for the dead, etc., all indicated a race of semi-nomadic hunters and herdsmen, just as the Chinese describe them.
It was a strange sensation to look down on figures which but for the parched skin seemed like those of men asleep and to feel brought face to face with people who inhabited, and no doubt liked, this dreary Lop-nor region in the first centuries A.D. The features of the heads closely recalled the homo alpinus type, which, judging from my anthropometric records, worked up by Mr. T. A. Joyce, still supplies the prevalent element in the racial constitution of the indigenous population of Chinese Turkestan and is seen in its purest form in the Iranianspeaking tribes near the Pamirs. The general appearance of these Lou-lan people seemed curiously to accord with the significant juxtaposition in which small bronze objects of Chinese origin were picked up on the slope below the little fort together with stone implements, There were indications elsewhere, too, suggesting that the interval separating the latest Neolithic period in Lou-lan from the advent of the Chinese may not have been a very long one.
Apart from their direct interest, the discoveries here briefly indicated had a special importance by furnishing me with a safe starting-point and some guidance for the difficult taak still before us, that of tracing the line of that famous ancient route through the forbidding desert eastwards. But it was impossible to set out for it at once. Incessant toil in the waterless desert with constant exposure to its icy winds had exhausted our Loplik labourers, hardy plants as they were and pleased with the rewards I gave them. When the last digging at the outlying ruins to the north-east had been done, I had to take them haak