| || || || |
Location of Chu-pin at Ying-p`an,
766 ON THE ANCIENT ROUTE ALONG THE KONCHE-DARYA [Chap. XXI
descending from the Kuruk-tagh to the south, is sufficiently accounted for by what we have learned above of the extent of the mountain area drained by that river, and of the height of the Hsi-ta-shan range in which its feeders rise. A reference to the map shows that similarly favourable conditions for the collection of drainage do not exist in any other portion of the western Kuruk-tâgh.
In view of the archaeological evidence obtained by me at a series of abandoned sites all along the southern edge of the Taklamakân and in the northern portion of the Tarim basin, I think we may safely recognize that the amount of water available for irrigation has diminished during historical times throughout this great region of innermost Asia, whatever may be the direct causes, rate of progress, and other factors connected with this process.17 Nor does the probability that cultivation on any appreciable scale is at present impracticable at Ying-p`an affect the question of the former importance of the site ; for the number of the ruined shrines and the size of the circumvallation traced there conclusively attest the former existence of a fairly large settlement at this point. The Kharosthi documents, mere fragments as they are, found at shrine i suffice to prove that the period of occupation of the site goes back as far as that which dated records enable us to assign to the Lou-lan station L.A. This chronological evidence in respect of the Ying-p`an site indirectly supports the location of ` the town of Chu-pin ' at that place ; for the account of Li Tao-yuan, or rather the record on which it is based, shows that at the time when the information was obtained both the ` towns ' of Chu-pin and Lou-lan were still actually occupied.
We know that the settlement at the latter place must have been abandoned about the middle of the fourth century. Occupation at Ying-p`an on the other hand probably continued into Tang times, as shown by the Tang coin picked up there and by collateral evidence found at the watch-stations of Kurghân and farther towards Korla. But this is exactly what was to be expected ; for long after the road through Lou-lan had ceased to be used, traffic from the side of Shan-shan, i. e. the Lop region, towards Korla is likely to have clung to the old and convenient route which led to it past Ying-p`an. The presence of surface water and the consequent possibility of some cultivation must have alone assured the continued occupation of the site, just as the springs and the grazing in the neighbouring portion of the Kuruk-daryâ bed have caused Ying-p`an to remain to this day a regular and necessary halting-place for travellers following the direct route from Lop to Turfân.
The above explorations will sufficiently explain my special interest in the region which separates the bed of the Kuruk-daryâ, at Ying-p`an from the present course of the Konche-daryâ. A reconnaissance made on March loth, the last day of our stay at Ying-p`an, had shown that the bed of the ` Dry River ' bending northward from our camping place approaches quite close to the end of the several flood-beds from the Shindi river below the ruined site ; thence it could be followed to the south-west for close on three miles, before it became indistinct. Continuing farther in the same direction, we passed rows of dead Toghraks striking to the south-east and suggesting successive shifts of the Konche-daryâ from its ancient to its present bed. Lal Singh had made the same observation when he followed the track leading from Konche-örtang (also known as Turfânkaraul) to Ying-p`an. But obviously such dune-smothered channels as these rows of dead trees probably mark lie too far south to be considered as possible feeders of the Kuruk-daryâ.
On the evening of the same day the additional supplies arrived from Tikenlik, but to my disappointment the Loplik guide asked for from Singer and again from Ying-p`an failed to appear. His absence, probably due to apprehensions of official displeasure, would necessarily hamper us in our work towards Korla. All the same I kept to my previously arranged plan. We should
17 See Serindia, i. pp. 243 sq., 286, &c. (see Index) ; - 459 sq. ; ii. p. 569, [and my paper, Geogr. Journal, lxv. pp.
Ancient Khotan, i. pp. 384 sq. ; above, i. pp. 71 sq., 435, 487 sqq•]•