Sec. i] EXCAVATION OF RUINED DWELLING, L.A. I. 375
called to two fragments of pottery, L.A. I. iv. ool, oo2, which have retained their glazed surface, in two different tints of green. Its preservation is manifestly due to these pieces having been embedded in refuse. The rarity of such glazed potsherds elsewhere at the site is accounted for by the corrosion to which all pottery remains, left unprotected on the ground, are exposed here from wind-driven sand. L.A. I. iv. 002, with its finely crackled leaf-green glaze, suggests Chinese pottery of the Han type, like the similarly crackled glaze in red of L.A. oo5. In the other fragment, L.A. I. iv. ooi (Plate Iv), which is decorated with an incised pattern, the dark green glaze recalls Western influence.
Finally I ought to record here that the careful search of the débris-covered slopes of L.A. I. resulted also in the recovery of seven Chinese coins, mostly fragmentary. They all belong to the type marked by the legend Wu-chu and assumed to have been first introduced by the Emperor Kuang Wu-ti, A.D. 26-57.
SECTION II.—EXPLORATION OF RUINED DWELLINGS, L.A. lI—vt.
After completing the search of the isolated dwelling south of the Stûpa, I moved my band of Brick-built
labourers to the large group of ruined structures on the south-west. It occupies a terrace-like piece structure,
of ground, about two hundred yards broad at its widest, near what my subsequent surveys have proved to have been the centre of the walled square enclosure marking the ancient Chinese station (see Plate 23). Surrounded on all sides by wind-eroded depressions, twelve feet and more deep, this ground had escaped being broken up into Ylydangs through the protection which the various ruined structures occupying it, L.A. II—vI, had afforded. In the main building, L.A. II (Plate 25), which appears to have originally had a rectangular enclosure solidly built of large sun-dried bricks, the thick walls of three narrow apartments still rose to a good height, as seen in Fig. IoI. Here I easily recognized the spot where Dr. Hedin had come upon his ` find ' of Chinese records on wood and paper. From the rubbish layers which covered the floor of the easternmost and narrowest of these apartments, L.A. II. ii, he had recovered forty-two narrow tablets or ` slips ', no doubt of the type already described, and about two hundred pieces of inscribed paper, most of them torn fragments.'
A first rapid inspection sufficed to show me that the contents of this refuse-heap, originally so Documents
well sheltered, had not been completely exhausted. Notwithstanding the statement that ` the from refuse
contents of the stall were sifted to the very last grain we came hard bare rain until in fact to the hd b ground hep, L.A.
ground II.1i. underneath, and had literally cleared out the whole of the interior ', the thorough search I made here revealed a great quantity of fragments of inscribed paper and wood. A number of these, including a wedge under-tablet with Kharosthi writing, L.A. it. ii. oo3, were found in the rubbish still about two feet deep, which Dr. Hedin's men had dug up but not removed from the small apartment. But far more turned up in the refuse which had been thrown out and left to litter the eroded slope immediately south (L.A. II. i). From this two packets full of Chinese fragments were ultimately recovered. The thin, curled pieces of wood which formed the majority were obviously shavings from Chinese slips of the regular size which had been scraped down in order to be used afresh for writing.2 The economy practised by the fresh use of this old stationery finds its obvious explanation in the fact that the material of these slips seems to have been ordinarily a pliable smooth wood, often of some conifers, which could not have been obtained locally.
It is worth noticing that among the Chinese documents here recovered there is a complete
1 Cf. Hedin, Central Asia and Tibet, ii. p. 132 ; Central ments, Nos. 739-41, Pl. XXIII; Nos. 878, 885 (Pl. xxvII).
Asia, ii. p. 632. For other records from L.A. u. i, ii, cf. Documents, Nos. 879-
2 For specimens of such shavings, see Chavannes, Docu- 884, 893, 896-902, 905-909.