Sec. vi] DECIPHERMENT OF KHAROSTHÏ AND CHINESE DOCUMENTS 369
within India peculiar to the region that had in Taxila its oldest historical centre. It does not appear possible to account satisfactorily for either the language or the script of our documents by the spread of Buddhism alone, seeing that Buddhism, so far as our available evidence goes, brought to Central Asia only the use of Sanskrit as the ecclesiastical language, and the writing in Brahmi characters. Nor would the assumption of a temporary extension to Eastern Turkestan of aka or Kusana power from the north-western borderlands of India be sufficient to explain the transplanting of an Indian language and its adoption for ordinary use among the people; for on the one hand, such a political connexion, if it ever really existed, must, in view of the Chinese historical records, have been very transitory, while on the other hand the forces that might have effected it were themselves Central-Asian rather than Indian.
It seems strange that ruins far away in the barbarian North, overrun by what Hindu legend vaguely knew as the mythical ` Ocean of sand' 27, should have preserved for us records of everyday life older than any written documents (as distinguished from inscriptions) that have as yet come to light in India itself. But from the first there was ample evidence pointing to this chronological conclusion. The close agreement in palaeographical features between the documents discovered and the Kharosthi inscriptions of the Kusana kings led me at once to the conclusion that those records on wood and leather must belong approximately to the period covered by Kusana rule in the extreme North-West of India. We know, in spite of all uncertainties as to the era or eras used in the inscriptions, that the domination of the Kusana dynasty in the Indus Valley and Punjab must mainly fall within the first three centuries of our era. It is certain that the Kharosthi script ceased to be used soon afterwards, in those territories which had formed its Indian home ; and it seems very improbable that it could have remained in current use in Khotan for a long period without undergoing perceptible changes. Valuable collateral evidence was provided by an interesting find made in the ruined building, N. vin, which will be described below. A narrow, Takhti-shaped tablet (N. xx. i) excavated there shows on one side a single line of Kharosthi, while on the other side I discovered to my surprise three lines of Brahmi characters, the only specimen of this writing from the site. Unfortunately the ink on both sides has become very faint, and of the Brahmi text in particular it has become impossible to make out more than a few detached characters here and there. In these, however, I thought that I could recognize characteristic features of the Brahmi writing of the Kusana period. There is nothing in the appearance of the tablet to suggest that the two texts were written at appreciably different times. Hence the close agreement of the palaeographical indications furnished by the Brahmi and Kharosthi writings supplied further evidence in favour of the conclusion already indicated.
I could not fail to recognize another proof of considerable antiquity in the use of wood as the only writing material apart from leather ; for the total absence of paper among the remains of the site clearly marks a date earlier than the fourth century of our era, from which onwards the use of paper in Eastern Turkestan is palaeographically attested by MS. remains from Kucha28. Finally, numismatic evidence helps to confirm the conclusion, since the coins which were picked up at different points of the site during my stay, seventeen altogether, were, with the exception of one of doubtful origin, all Chinese copper pieces current under the Later Han dynasty 29.
27 Comp. Râjat. iii. 279 sqq.
28 See Hoernle, Report on Central-Asian antiquities, ii. pp. 13 sq.
29 As will be seen from App. D, 8 coins from the Niya Site belong to issues showing the symbols Wu-chu (see for specimens Nos. i 2-r 4 in Plate LXXXIX); seven are small
pieces without any legend, such as are usually attributed to the reign of Hsien ti (190-24o A. n.) ; specimens of these are seen in Nos. 7, 8, ro, Plate LXXXIX. No. 7 with another piece was picked up on eroded ground close to N. xv. One piece is in too poor a state to be determined. Another coin, the characters on which could not be made out by me at the