Sec. v] DISCOVERY OF DATED DOCUMENTS 267
they had left behind there. The second document (B), dated only by month and day, contains a military requisition from the Li-hsieh garrison, addressed by Yang Chin-chting, commandant of the place, to the civil magistrate, asking for skins to re-cover drums and for quail's feathers to refit arrows. The third, dated in the seventh month of the seventh year Chien-chung (786 A. D.), records the issue of a loan of 15,000 pieces of money on certain conditions to a person who appears to have been of Li-hsieh ; for a note subsequently added and dated in the tenth month of the same year records that i o,000 pieces of money have been repaid on account of that sum by an inhabitant of Li-hsieh.
With the records just described the Chinese documents excavated by me in D. v and in three other ruins of Dandan-Uiliq (D. vII, vIII, ix), show the closest agreement, not only in general appearance, paper, and style of writing, but also in character of contents and in dates. A comparison of these documents, as translated by M. Chavannes in Appendix A, conclusively shows that they are all formal records of public or private transactions. Records of loans, exactly resembling in their terms and provisions the third of the above documents (C), are the most frequent. On the other hand, we see that their dates, ranging from A. D. 78 t to 790, fall within, or quite close to, the period covered by the two dated ones among Dr. Hoernle's previously discussed documents.
In view of this complete agreement of contents, dates, and palaeographic evidence, it seems practically certain that the latter represent some of the finds of ' Khitai khats ', i.e. papers with Chinese writing which Turdi well remembered to have made on a visit to the site some years previously, and which with other ' old things ' he had sold to Badruddin Khan, his usual employer at Khotan. It is impossible now to ascertain whether these particular documents, which had all three been transmitted to Mr. Macartney in one batch from Khotan 7, came from one of those rooms in the ruined house (D. v) which I found to have been roughly searched, or from some of the other ruins that had similarly been ' explored' before by Turdi's parties. But even without such specific information the comparison of D. v. 6, the first Chinese document I unearthed at the site, makes it clear beyond all reasonable doubt that Li-hsieh, the locality all these documents name, must be identified with the settlement or small tract to which the ruined shrines and dwellings of Dandan-Uiliq belonged.
M. Chavannes, in his first note on the document A, points out that the reading of the Li-hsieh, name 01 sit Li-hsieh is rendered uncertain by the first character not being found in Chinese anc entf dictionaries 8. The doubt as to its true sound could be removed only by the discovery of the settlement.
local name in a printed Chinese text, or else, perhaps, by its identification in one of the Brahmi documents from Dandan-Uiliq. Fortunately no obscurity can attach to the interpretation of
the term 71-; Liu-ch`ênK ` Six Cities ', given in D. v. 6 as well as in the document A, as the
designation of the administrative division to which Li-hsieh belonged. Already, on my return journey through Kashgar, when Mr. Macartney and his Chinese literatus Sun Ssit-yeh kindly
furnished me with transcripts and preliminary translations of the Chinese documents brought back by me, I had ascertained that the term ' Six Cities ' is still well known by Chinese officials in the ' New Dominions' as an old designation of the Khotan territory. In a note
° See Dr. Hoernle's note, Report on C.A. ant., i. p. vi., on batch M. 9, received by him in Oct., 1898. It is significant that the eight sheets of paper with cursive Brahmi writing comprised in the same batch are all documents in the ` unknown' Eastern Iranian language, corresponding exactly in writing and character to the Brahmi documents
discovered by me at Dandan-Uiliq along with Chinese records ; see Report on C.-A. ant., ii. pp. 36 sqq., and below, pp. 269 sqq.
8 The difficulty had already been noted in Dr. Hoernle's Report, ii. p. 22.