Sec. iii] REMAINS SOUTH-EAST OF KUCHA AND LIST OF ANTIQUES 821
another enclosure, of an irregular oblong shape, and showing low walls of inferior strength and apparently of later date.
The only relics that Afraz-gul had picked up when searching the surface of this site were the fragments of red and grey pottery described in the List below. Inconspicuous as they are, they furnish an interesting archaeological indication ; for, as Mr. Hobson has pointed out in his Notes on the ceramic specimens of my collection,2 the ornamentation and fabric of these fragments from Khitai-shahri prove that they belong to types which are plentifully represented among the pottery remains found at the Lou-lan sites and also along the Tun-huang Limes. The result of Mr. Hobson's examination points to an early occupation of the site, and on this account I particularly regret that I was unable to inspect its remains myself. I can only hope that the present record may draw the attention of some competent future visitor to the place. It derives additional interest from a topographical observation noted in the Surveyor's route report. About a mile and a half before reaching Khitai-shahri from the north-east, he crossed a well-marked road leading from NW. to SE., which his guide declared to be the regular route for such traffic as passes between Kuchâ. and the Lop region. Considering the importance of this route during the early period when the great high road from China passed through Lou-lan, significance must attach to the fact that we find it guarded by a fortified post just at the point where it reaches the south-eastern extremity of the once cultivable area of Kuchâ.
The return journey to Kucha from Khitai-bazar, across grazing grounds containing an abundance of reeds and scrub, brought us late on the same day to Tim. This outlying hamlet, at the extremity of a narrow strip of cultivated ground irrigated from one of the main canals of the Muz-art river, takes its name from an ancient ruined mound, which was too much destroyed to furnish any indication of its original character. Our next march, on May 4th, led us for some twelve miles along that narrow cultivated belt, and then for another nine over bare steppe ; this forms part of the alluvial fan of the Kucha river, but no longer receives water from it. It was only about six miles from Kuchâ, town that we passed into the area that had manifestly been long and continuously cultivated. The route followed by Afraz-gul from Khitai-shahri to Kuchâ. town lay, as Map No. 17. c. I, 2 shows, some distance farther east. It enabled him to determine the extreme limits of the irrigated ground on that side ; most of it had been but recently reclaimed from scrubby waste. But the ruined mounds that he found in two places, and at a third the manifestly ancient walled post of Sang-khan-atam which he examined,3 afford proof that in this direction, too, the permanently occupied area was once greater than it is now.
Owing to lack of time, and the wide extent of the ground over which the old remains of Kuchâ. are scattered beyond the present limits of the oasis, my visits to the sites described above were necessarily very rapid. Yet the observations I made sufficed to familiarize me to some degree with the conditions under which the miscellaneous antiques acquired during my stay in the oasis, and described in the List below, were probably found. All these small objects correspond closely in type to the Tati' finds familiar to us from the sites of ancient occupation around the Khotan oasis, which have been abandoned to the desert since Buddhist times. Since wind-erosion is at work outside the irrigated area of Kuchâ., though to a much smaller degree than south of the Taklamakan, we can believe the statements of Mir Sharif and Aziz Palwan', who supplied most of these `finds', that they were picked up at ` Tatis ' of Dawan-kum and similar localities to the west and southwest of the oasis. In any case it should be noted that remains of stucco relievos, which the exploitation of ruined shrines might have yielded, and decorated terra-cotta fragments, such as are so abundantly recovered at Khotan by washing the soil in ancient ` culture strata ', are almost
2 See below, Appendix D. 3 For a sketch-plan, see PI. 44.
II 5 N