Sec. vi] DECIPHERMENT OF KHAROSTHÏ AND CHINESE DOCUMENTS 373
administrative agents were left a preponderating share in the management of affairs directly concerning the local population. The contents of our ancient rubbish-heap seem strikingly to illustrate this view. Since Kharosthi and Chinese documents intermingled throughout its layers, there can be no doubt as to the simultaneous existence of the two administrations working side by side. Yet at the same time the overwhelming proportion of Kharosthi records proves that the range of affairs left to native management must have been far the wider.
The occurrence of the Chinese slips of wood throughout the rubbish accumulations and at greatly varying depths is of special chronological interest. We know that after the reign of the Emperor Wu ti (265-289 A. D.) Chinese political relations with the ` Western Countries' completely ceased 41. Taking then into account that the document N. xv. 326, accurately dated in 269 A. D., was found only about one foot above the floor, and that a considerable number of Chinese records with contents of similarly official character were found fully 3 to 32 ft. from the floor, and thus close to the surface of the extant mass of refuse 42, it is evident that we are justified in recognizing the period covered by Wu ti's reign as the time during which the whole of the records, &c., found in N. xv. must have accumulated. The chronological limits thus ascertained make it less of a loss to us that the other Chinese documents are either undated or only by month and day 43, and that we cannot at present fix the chronology of the reigns to which the fully dated Kharosthi tablets refer.
Considering the unsubstantial construction of the building, it appears very improbable that N. v. could have continued to be inhabited for many years after Wu ti's time ; and as its fate may safely be assumed to have been typical of that which the other structures within this portion at least of the site underwent, we are led to conclude that the end of the third century of our era must have seen the site abandoned to the drifting sand of the desert. Whether this abandonment was indirectly connected with the political and economic changes which undoubtedly accompanied the withdrawal of Chinese authority, or was due solely to natural causes affecting irrigation and therefore cultivation, is a question regarding which the available evidence would scarcely justify the expression of a definite opinion at present.
It was interesting to trace the close connexion with China also among the relics other than documents which this ancient rubbish-heap had preserved, and which may now be briefly reviewed in conclusion. The fragments of a well-finished circular lacquered bowl (N. xv. ooi), decorated outside with black and red bands (see Plate LXV), and showing a highly-polished red surface inside (see Plate LX X), unmistakably indicate Chinese workmanship ; and the cane material too, clearly points to that origin. The pieces of delicate plain silk (N. xv. of I), of which Plate LXXVI reproduces portions, may also have come from the Far-eastern portions of the empire. The fragments of plain and cut glass, N. xv. ooi. b—h (see Plate LXXIV), mostly greenish in colour, showing great transparency, and very different from the coarse material found at other sites, may be the remains of Western imports. Glass was until the middle of the fifth century of our era known in China only as an import from Ta-chtin, the Far West 44 ; and it appears very improbable that the introduction of glass manufacture could have been delayed so long, if the making of glass had been an art practised in Eastern Turkestan when the latter was under Chinese control during earlier centuries. It is significant that, according to the Pei shih, the first makers of glass in China were traders ' from the country of the Great
41 See M. Chavannes' remarks on N. xv. 93 in App. A, part ii. 43 Dates by month and day only are found on N. xv. 348,
44 See, e. g., N. xv. 337, 339, 349, 351, 353, 02, o8, 09, xoo, 207, 351.
010. " See Hirth, China and the Roman Orient, pp. 228 sqq.