414 THE LOU-LAN SITE [Chap. XI
sufficiency of this explanation for the same use of a foreign administrative language might well be open to doubt.
In the present inadequacy of our historical data it is impossible to assert whether a better solution of the problem is to be looked for in the spread of Buddhism, which may have carried the language and script prevalent in the extreme north-west of India with it into common use throughout the Tarim Basin ; or whether in this adoption of a Prakrit language, closely allied to that current on the Indus in Kusana times, we ought perhaps to recognize a lasting impression left by that temporary extension of Indo-Scythian power from across the Pamirs of which we catch dim glimpses from Buddhist tradition in China.2 But seeing the importance of the question raised, I feel all the more gratified by the fact that the conclusion first drawn on the spot as to the language and character of the Kharosthi documents from the Lou-lan Site is now fully confirmed by the careful examination which Professor Rapson has been able to make of them since my return in 19(39.
The abstract of the contents of these documents with which Professor Rapson has very kindly supplied me embodies the main results of his decipherment down to December, 1916, and makes it perfectly clear that in character, language, phraseology, and other respects they show the closest agreement with the Kharosthi documents brought to light in so much greater numbers from the ruins of the Niya Site. Just as there, we have a variety of deeds, letters, official orders, lists, and the like worded in the same early Prakrit dialect with an admixture of queer Sanskrit phrases in complimentary introductions, etc. Peculiarities of style, phonetics, and spelling leave no doubt as to identical standards having been followed by the chancelleries from Khotan to Lop at the period to which the records of both sites belong. Many, if not most, of the personal names which we meet in the Lou-lan Site documents occur also in the Niya series, though this does not, of course, imply identity of the individuals. Just as in the Niya series, we find numerous names of unmistakably Buddhist or Indian derivation, such as Anarndasena, Bhati§ama, Bhimaya, Budhamitra, Dhalnnapala, Kumudvati, Purpnadeva, Caraka, Rutra, Sujada, Vasudeva, side by side with others which seem of local origin, e. g. Cauleya, Cuvalayina, Kapgeya, Kalpisa, Kipsa, Kitsaitsa, Lampurta, Maldraya, Porbhaya, Pulkaya, Signaya, Tasuca, Tameca, Varpeya. The official titles of Cojhbo, Gugura, Kori, Vasu are common to both Lou-lan and Niya records.
The rectangular double tablets L.A. iv. ii. 1, 2, 3 contain deeds, and in accordance with the practice uniformly observed in such formal records are exactly dated in regnal years. But only in the case of L.A. Iv. ii. 2 (Plate xxxviii), which relates to a transfer of land by one Sigayita to a woman Kosena, can the name of the reigning king be made out with certainty. He is designated as Malzraya Amgoka devapzzlra. His name and style curiously recall the JWaharaya 7ilu[nz]gha Anzkoizga [or Amvaga] devatzclra mentioned in the dates of two rectangular tablets from Niya, N. xxi. 6. a, 7+4, full transcripts of which Professor Rapson's kindness has made available to me in the proofs of his and Messrs. Senart and Boyer's text publication now passing through the press. In view of what has been shown above as to the dependence of the territory of Ching-chtieh, of which the Niya Site represents the chief place, upon Shan-shan or the Lop territory,3 there is a temptation to assume that the same ruler is meant in the records of both sites.
Though in L.A. iv. ii. 3 the year and name of the reigning king can no longer clearly be read, this document is of considerable interest. It contains a deed recording the sale of a piece of land by Camaka, a man of Kroraina settled at Calmadana,4 and conveying full rights of possession to