8io KUCHA AND SOME OF ITS ANCIENT SITES [Chap. XXIII
abandoned patch of cultivation ; it may have dated from the time when, under Yagûb Beg's régime, digging for saltpetre was carried on within the ruined circumvallation.
Ruined This, as the plan (P1. 40) shows, forms an approximately orientated square, with walls 168
circumvalla- yards long on each side, protected by small bastions of varying size. Both walls and bastions are tion.
built of sun-dried bricks, mostly measuring 15 by 8 by 3i inches ; among these are interspersed flat pieces of hard clay or kisek. The walls, about i8 feet thick on the average, still rise in places to a height of i8 to 20 feet. The gates leading through the northern and southern faces are protected by short ` curtains ', behind which the opening in the main wall is approached through outer courts. In the court at the northern gate the foundations of the walls of several rooms built against the protecting ` curtain ' could be traced. Mir Sharif stated that, on a previous search, he had come upon several wooden documents in the court to the north-east ; but only a fragment of a pottery vessel rewarded our clearing of these rooms. This fragment, Tong. of i, shows, like other potsherds picked up at this site, a fine terra-cotta-like body.
Date of No structural remains whatever could be seen within the enclosed area. Considerable refuse
Tonguz- heaps extend along the foot of the walls to the north and west. Our search could not be carried
deep, but it brought to light a child's well-made string sandal, Tong. 02, of a type rendered familiar
by finds along the Tun-huang Limes ; fragments of ribbed silk, Tong. 04 ; a mass of raw cotton, Tong. oi, &c. Among the pottery debris, all of fine terra-cotta-like ware, Tong. o8 deserves notice on account of its dark-green glaze, slightly iridescent on both sides, which points to Tang times or the period immediately preceding.la A paper fragment with Chinese writing, also found here, was too small to afford any definite chronological indication. But on general grounds there is every reason to believe that the circumvallation dates from pre-Muhammadan times, like the ruined shrines in the vicinity to be mentioned later. Tonguz-bash is the junction of the routes followed by those who wish to gain the caravan road along the Khotan river from Kucha town and from the Yulduz-bagh portion of the oasis. In view of this fact and of the analogous position occupied by the ruined circumvallation of Khitai-shahr (Map No. 17. D. 2) situated on the direct line leading towards Lop,2 the conjecture may be hazarded that the site is that of a fortified station intended to guard the shortest line of approach to Kucha from Khotan territory lying southward.
Remains of At a distance of a little over a mile to the ESE. of the town' of Tonguz-bash my guides
Buddhist showed me the remains of what they called its ` But-khana '. On approaching it I noticed the faintly shrines.
marked line of an old canal, about 8 feet wide, raised about io inches above the level ground, which here showed signs of having suffered some wind-erosion. Beyond it, badly broken walls scattered, as the rough sketch in Pl. 4o shows, over an area about 13o yards long from north to south mark the position of what were evidently once small Buddhist shrines and monastic dwellings. The structures were all said to have been searched by M. Berezowsky's men, and had probably also since been dug over by antiquity hunters. They were mostly built of sun-dried bricks of the same size as those in the walls of Tonguz-bash ` town ' ; here and there also of stamped clay, into which wooden beams and posts, since decayed through exposure, had been inserted to strengthen them. On the eastern side of the area, there remained the foundations of a small structure of timber and wattle walls, marked a in plan, which apparently enclosed two small shrines built back to back. Among completely broken and weathered fragments of decorative wood carving, we here recovered the piece of stucco relief, Tong. oio, showing half a warrior's helmeted head.3 It had obviously belonged to some such scheme of sculptural wall decoration as is illustrated by the numerous small reliefs that I recovered from the Buddhist shrines of the ` Ming-oi ' of Shikchin, near Kara-shahr.4
la See below, Mr. Hobson's Appendix D. 3 For a description see the List below, ii. p. 822.
2 See below, ii. pp. 82o sq. 4 Cf. Serindia, iii. pp. 1191 sqq.