SECTION I.—THE RUINS OF BILÉL-KONGHAN
AFTER a day's halt at Imam Ja`far Sadiq, made necessary by a variety of urgent practical tasks, I started on the morning of November 1, 1906, across the high sands eastwards in order to revisit the tract near the end of the Endere River before moving on to Charchan. In 1901 I had explored there the ruins of an old fort, half buried under drift sand, and a Stûpa. Want of time had then prevented a thorough survey of the whole site. The existence of other remains was indicated by various evidence, and my archaeological conscience would not have allowed me to forgo a fresh visit, even if the shortest route to Charchan and Lop-nor had not led quite near. But a particular inducement was supplied by a curious acquisition which I made during my stay at the Niya Site.
On hearing of my intended move to Endere, Sadak, the young cultivator from the Mazar, whose guidance had proved useful already on my first return to that site, told me of an inscribed Takhta' he had come upon a year or two before, while ` searching for treasure ' close to the old fort of Endere. When on my arrival at the shrine he brought it for my inspection, I was greatly surprised to find that it was an irregular oblong tablet fairly well preserved and showing clear Kharosthi writing on both sides.' The writing proved that it belonged approximately to the same period as the wooden documents of the Niya Site. Yet my own finds made in the old fort of Endere in 1901 had established the fact that this ruin had been occupied at the beginning of the eighth century and had probably been abandoned to the desert soon after. Thus an interval of not less than four centuries seemed to separate the remains explored by me in 1901 from the document now put before me by Sadak. I was at first inclined to suspect in the latter a find clandestinely brought away from the Niya ruins. But Sadak, as well as his father Samsak, a withered old shepherd of the Mazar flocks, clung stoutly to the story first given, in spite of all my critical questionings. Assuming their statement to be true, there was here an interesting archaeological puzzle which could only be solved on the spot. Its solution, as I shall soon be able to show, had a wide historical bearing:
Owing to the necessity of saving our hard-worked camels and men any additional hardships, I could not strike across the desert to Endere by any other route than the one safely followed in 1901. Otherwise I should have preferred to use the opportunity for crossing the unexplored belt of desert between the Niya Site and the end of the Yar-tungaz River. As our marches, therefore, lay by the route already surveyed, there is no need to describe them .2 The small settlement of Yar-tungaz Tarim of which I have given some account in my former Report, and which I reached once more on November 3, presents all the characteristic features of a terminal oasis on a small scales So it would be naturally tempting to look for the remains of earlier occupation in the desert beyond the present river end. But the existence in that direction of ancient remains of any sort was uniformly denied by all local informants, and there was the additional fact that Prof. Huntington,