Sec. ii] ANCIENT REMAINS AROUND CHARCHAN 303
Saif-ullah called it by the name of Nan yaigapt "Tali" (` the Tati where the meal was eaten '), explaining it by a whimsical story.
The ground showed a uniform surface of fine gravel. Its flatness was broken only by occasional low swelling ridges which, however, bore signs of being swept and scoured with equal force by the winds. Their erosive power was attested by the general smallness and wind-worn appearance of the potsherds which covered the ground in abundance. The material was mostly dark-brown, black, or a deep red on the outside faces. As the specimens taken (Char. 001-007) show, both hand-made and wheel-made pottery are represented. A very conspicuous feature in the midst of this ` Tati ' was a big ` witness' of pure riverine loess, about eighteen yards long and eight across, rising with steep slopes to a height of not less than twenty-three feet above the ` Sai ' surface. Potsherds lay plentifully at its foot and also on all the small terraces breaking the slopes. Thence it may be safely assumed, I think, that the height of this ` witness ' marks the extent to which the ground-level has here been lowered by wind erosion since such relics of ancient occupation came to be deposited. In the sides of this ` witness ' I was unable to trace either clear stratification or embedded pottery débris. This suggests that the loess deposit is due mainly to aeolian growth and took place during a period preceding settled occupation.
There is every reason to assume that this area received irrigation mainly by a canal taking off from the Charchan River. But the flood-water of the Ayak-tar stream, which after heavy rain in the mountains is said still to carry water on occasion as far as the ` Tati ', may also have been utilized as an additional source of supply. To judge from the configuration of the gently sloping alluvial fan, it seemed clear that the water of the Charchan River could still be brought here without any difficulty. It is equally certain from the instances of similar reclamation which I observed near Khotan (e.g. on the ` Sai ' south of Sampula) and Kucha, that the soil of all these ` Tatis ', which far-advanced deflation has reduced to a surface of fine gravel, could be rapidly improved and rendered capable of cultivation by adequate irrigation ; for the silt-depositing process which results from irrigation is everywhere in this region greatly aided by aeolian action, i.e. the accumulation, on moist soil and on all ground protected by vegetation, of the fine dust which fills the air of the Tarim Basin so abundantly during the greater part of the year.? But, with many thousands of acres of fertile ground available lower down and near the river, renewed cultivation is not likely to take this direction.
The Tati' of Lalulik, to which I was subsequently taken at a distance of about two and a half miles northward, proved to be of the same character, except that small witnesses' of loess, about four to six feet in height, were here frequent. Pottery débris covered the ground for over a mile towards the existing oasis. A third ` Tati ' area in this neighbourhood, called Koyagii-keime, I was obliged from want of time to leave unvisited. I had seen enough to make sure that none of these ` Tatis ' offered scope for systematic archaeological labour. Nor did I succeed in discovering among their small débris anything affording a definite clue as to their age. Some gauge, however, as to the relative chronology of these débris areas is afforded by the fact that, while to the south the ground once cultivated has been uniformly denuded right down to the underlying Piedmont gravel, the ` Kône-shahr' ' Tati ' close to the present oasis exhibits remains resting on loess soil. Here the process of erosion is still in progress, and will be continued until the new fields being pushed southward have invaded the whole area. The ` Lalulik ' area, with its numerous surviving small loess banks, may represent an intermediate stage. However this may be, and whatever intervals of time may divide the periods of abandonment of these different ` Tatis ', their extent is quite sufficient to prove that the ancient Chti-mo must be located in the position of the present Charchan oasis and its immediate vicinity