Sec. vii] GENERAL OBSERVATIONS ON DANDAN-UILIQ SITE 285
the colonies, which in our own days we have seen springing up with such astonishing rapidity within the areas reached by great irrigation works like the Chinâb and Jhelum canals, would lapse within an equally short time into the arid uninhabited waste from which they had emerged.
But apart from the effect which political conditions must thus have upon cultivation in the Khotan region, there are, no doubt, physical changes, too, which gravely threaten it and represent a far more permanent danger. The question of the character and extent of these changes concerns the geographer rather than the archaeologist. Yet it is plain that we must touch on it here, if only in order to explain why the site, which was once capable of supporting a flourishing settlement, was never occupied again after its first abandonment. That the main danger to the maintenance of cultivation does not lie in the advance of the drift-sand itself, but in the reduction or failure of the water-supply needed to cope with it, must be clear from what has already been said. Therefore before we attempt to form any opinion about the physical changes which caused this ancient settlement to be definitely absorbed in the desert, it is essential that we should realize what the source and conditions of its water-supply were.
The preliminary question thus raised presents at the first glance far greater difficulties at Dandân-Uiliq than at any other of the ancient sites I examined. Whether they were ` Tatis' near existing oases, like those between Gttma and Piâlma and on the outskirts of the Khotan oasis, or marked ancient terminal oases since completely abandoned, like the Niya River Site and probably Endere, no doubt ever arises as to the source from which they received irrigation. In each case we find the river or stream, which alone could have supplied it, unmistakably indicated by existing hydrographic conditions, In the case of Dandân-Uiliq we are confronted by a series of possibilities, and only a far more exact survey of the whole desert region around than could be attempted by me, including a series of levellings, will enable us to decide the question conclusively. If we merely look at the map the position of Dandan-Uiliq about halfway between the Yurung-kâsh and Keriya rivers suggests the possibility of water having been brought to the site by canals taking off from either. Again, the map shows us that DandânUiliq lies exactly in the natural line of drainage for the rivers of Chira, Gulakhma, and Domoko, which now lose themselves in the desert due north of these closely adjoining oases.
Archaeological evidence, so far as I could gather it at the site, does not help us here. Though there were little inequalities of the ground at several places left bare of sand, which looked like possible remains of small embankments for irrigation cuts, it was only in one place, to the north of D. VII (see Plate XXIV), that the trace of such a channel, about 2 feet broad, could be definitely followed for a short distance between the dunes. Its apparent direction was from SSE. to NNW., but seeing the smallness of the channel and the shortness of the distance (less than 200 yards) over which its traces could be picked up, it is manifestly impossible to base upon this any safe conclusion as to the direction and origin of the main canal from which it was fed.
As already indicated in my Preliminary Report, a series of subsequent observations about the gradual receding of cultivation near Chira, Gulakhma, and Domoko incline me to the opinion that Dandan-Uiliq had received its water by an extension of the canals which down to a much later date irrigated the area now abandoned to the desert north of the above-named oases. The débris-covered site of Uzun-Tati which I surveyed there, and which I think can safely be identified with the Pi-mo of Hsüan-tsang and Marco Polo's Pein, is proved by unquestionable evidence to have been occupied for at least five centuries longer than DandanUiliq. It lies fully twelve miles, if not more, beyond the northern edge of the Chira oasis;
| || || || || |