to that of the Chinese records on wood brought to light in 1901 from the rubbish heap N. xv at this site.23 Speaking without any knowledge of Chinese palaeography I must own to the impression that the general style of the characters in these beautifully penned labels of N. xiv. iii approaches more closely to the ductus observed in the documents of Former and Later Han times which I discovered along the Tun-huang Limes, than to that shown by the records of the Chin dynasties' epoch among my finds at the Niya and ' Lou-lan ' Sites.24 But it must remain for competent Sinologists to decide whether this impression is right, and if so, whether this resemblance is not due, perhaps, more to archaic tendencies of calligraphy than to proximity in time.
As to the other small relics which have survived in this dustbin, several among them present points of special interest. It is true, we can scarcely hope ever to obtain a sufficiently close knowledge of the development of ancient weaving industries in Central Asia and China to derive chronological indications from the manifold fragments of fabrics in silk, wool, and felt which the Descriptive List below shows (see N. xiv. iii. ooi, oo3, 007). But the total absence of cotton among these materials, as far as Dr. Hanausek's analysis of characteristic specimens shows, is noteworthy. The conical head-gear, N. xiv. 004, in carefully gored yellow felt is curious on account of its Phrygian cap-like shape. The two arrowheads in bronze and iron, N. xiv. oo8 ; iii. 0032 (Plate XXIX), do not differ materially in shape from those found elsewhere at this or the ' Lou-lan ' Site. The small bronze plates, N. xiv. 009-0011 (Plate XXIx), which look as if they had belonged to scale armour, are certainly peculiar, but were found on eroded ground near N. xiv. iii, not in the dustbin itself. The chopsticks, N. xiv. iii. 0020, 0021, imply Chinese customs of eating, and the lacquered ones are, like the piece of a fine lacquered frame, N. xiv. iii. 0025, probably of Chinese origin. The carefully worked ornamental button-edging in leather, N. xiv. iii. 0033 (Plate xxix), and the pendant or button in paste inlay with elaborate design, N. xiv. iii. 0035 (Plate xxIX), attest the use of elegant apparel by those who resided close by. Furniture is represented by the two well-designed chair-legs in turned wood, N. xiv. iii. 0036, 0037. Small but of distinct interest are the two morticed wooden pieces, N. xiv. iii. ooio, 0017 (Plate xxVIII), which after the analogy of numerous similar finds from the Tun-huang Limes have been described as seal-cases. Their real use is not quite certain. But in any case it deserves attention that, abundant as finds of small objects in wood serving stationery purposes have been both at the sites of Niya and ` Lou-lan ', yet the only analogous pieces have come to light from the remains of those old watch-stations of Han times in the Tun-huang desert.86
Even before the clearing of the refuse layers of N. xiv. iii was completed, the number of available men had enabled me to commence the excavation of the chain of smaller ruins stretching south, Naik Ram Singh supervising. When I was able to bring the whole of the men to this task the progress made was rapid. Some of the dwellings had suffered much from erosion, and within their broken walls but little sand had accumulated. Others had been better protected, and it cost great efforts to clear the high sand which filled their rooms, in one or two instances to the very ceiling. But the men wielded their Ketmans with remarkable perseverance, in spite of the trouble caused by the necessarily limited water rations, and the encouragement of small rewards paid for the first finds of interest in each structure sufficed to keep them hard at work for ten or eleven hours daily.
At a group of much-eroded dwellings of small size, about half a mile to the south of N. xiv, the ruin first cleared, N. xv (see plan in Plate io), yielded some well-preserved Kharosthi tablets of varying shapes as well as two stick-like implements, probably used in weaving, and a much-fissured double-bracket in wood, as shown by the photograph in Fig. 52, over five feet long. The rather