Sec. v] DISCOVERY OF DATED DOCUMENTS 271
also in my Brahmi documents can be definitely asserted on the basis of Dr. Hoernle's preliminary notes 19. Without awaiting the results of his detailed analysis it is impossible to say whether the contents, too, furnish definite proof of identical origin. But in this respect we can fortunately rely on the parallel evidence of the Chinese documents. We have seen already that two of these contained in Dr. Hoernle's collection must have been written at Li-hsieh, which the testimony of my own find (D. v. 6) enables us to fix at Dandan-Uiliq, and from the same batch (M. 9) which furnished those two Li-hsieh records, came eight of Dr. Hoernle's best preserved specimens in Brâhmi script with an Eastern Iranian language.
The minute analysis of the Brahmi documents previously at his disposal enabled Dr. Hoernle to establish several philological facts which are of very great interest. By determining a number of words, mainly numerals and terms used in the dating, he succeeded in proving the Eastern Iranian type of this ` unknown language ' and its special connexion with the Galcha dialects of the Pamir region 20. He clearly ascertained the important fact that the majority of the complete documents are fully dated, though the key to the reckoning of years has yet to be discovered 21. A number of ingenious observations, such as the discovery of lists of names at the end of documents, accompanied by what are manifestly the marks of witnesses, the frequent occurrence of Chinese signatures or office-stamps in the same place, &c., permitted Dr. Hoernle to arrive at the undoubtedly correct conclusion that we have in them records of official or private transactions similar in character to the deeds of loan, requisition orders, &c., represented by the Chinese documents already discussed. Yet in order to invest these observations with their full historical value, it was essential that the place of origin and the period of these records should be fixed beyond all doubt. The special importance of the Brahmi documents brought to light by me at Dandan-Uiliq lies in the fact that, few and fragmentary as they are, the certainty which exists as to all circumstances of their discovery supplies just what the student of these records needed.
The close association of Brahmi documents with Chinese ones, not only in D. v, but also in other ruins (D. VII, vIII) to be discussed hereafter, permits us to fix the time of the former with approximate certainty. We shall see that the dates of the Chinese records discovered by me all range between the years 781 and 79o, and that very distinct archaeological evidence points to the buildings containing them having been finally abandoned soon after. Insignificant in size and material as the Brahmi documents are, it appears improbable that they should date back to a period appreciably earlier than that of the Chinese papers with which they were found. Hence the conclusion seems justified that the Brahmi documents, too, must belong to the last quarter of the eighth century. If, in conjunction with this chronological fixing, we consider the character of those Brahmi documents which Dr. Hoernle has fully analysed, it becomes evident that their language must have been that actually spoken by the inhabitants of the ruined settlement during the period immediately preceding its abandonment. This then was the ` barbarian language ', to the use of which, by the people of Li-hsieh, the Chinese documents D. v. 6 and A both distinctly allude 22,
I have had occasion, in a previous chapter, to point out that the fact of this language having proved to be of Eastern Iranian origin is in full accord with what indications we other-
is See Dr. Hoernle's notes v., x., xiv., xvi., xvii., xviii., in inventory list.
20 See Report on C.A. ant., ii. p. 32 sq.
21 See loc. cit., p. 35.
22 In D. v. 6 we read of the petitioner's complaint being