EXPLORATIOT^ IN CHINESE TURKESTAN. 769 



nients on leather — a material one could hardly expect to find among' a 

 Buddhist population with an Indian civilization. 



Many of the Kharoshthi tablets unearthed are in excellent preserva- 

 tion, and still retain the original clay seals and strings with which 

 the}' were fastened (pi. v). We are thus al>le to study exactly the t(>ch- 

 nicalities connected with the use of wood as a writing material. This 

 is not the place to discuss such details, but I may mention at least that 

 each document intended as a letter or record of some importance, 

 whether wedge-shaped or oblong, is provided with a carefull}" fitted 

 covering piece or envelope bearing the address or "docket" entry. 

 An ingenioush' designed system of fastening with a string and a neatly 

 inserted cla}" seal, prevented unauthorized inspection of the contents. 



The remarkable series of clay seals discovered on these tablets is of 

 exceptional interest, because it furnishes most convincing evidence of 

 the infiuence which classical western art has exercised even in distant 

 Khotan. A frequently recurring seal, probably that of an official, 

 shows the figure of Pallas Athene, with shield and a^gis, treated in 

 archaic fashion. Another fine seal is that of a well-modeled naked 

 figure of pure classical outline, perhaps a seated Eros. On others, 

 again, appear portrait heads showing classical modeling, though bar- 

 barian features, etc. We know well how classical art had established 

 its infiuence in Bactria and on the northwest frontier of India. But 

 there was little to prepare us for such tangible proofs of the fact that 

 it had penetrated so much farther to the east, to halfway ))etween 

 western P]urope and Peking. I may note here, as an interesting dis- 

 covery made while these pages are passing through the press, that 

 Professor Karabacek has traced the remains of a (Ireek legend, appar- 

 ently a magic fornuda, impressed on the edge of one of the claj^ seals, 

 containing in its center the figure of Athene Promachos. 



From the contents of the documents themselves we may confidently 

 expect nuich fresh light upon a chapter of Central Asian history and 

 civilization which until now has seemed almost entirely lost. Owing to 

 the great numl)er of the texts, the cursive character of the script, and 

 peculiar difiiculties connected with the nature of the records, their 

 complete decipherment will recpiire nmch time and lal)or. But it is 

 already certain that, as I recognized in the course of my first examina- 

 tion on the spot, the language of the documents is an early foi'm of 

 Indian Prakrit, with a large admixture of Sanskrit terms. It is highly 

 probable that most of them contain official orders, such as safe con- 

 ducts, correspondence, etc., as well as private memoranda and records. 

 Religious texts, prayers, etc., may be suspected in some of the long- 

 tablets found in what seem to be shrines or monasteries. Many of 

 the documents bear exact dates, in which the years are indicated with 

 reference to the reigns of named rulers. These will (Miable us pi-obably 

 to restore a portion of the historical chronology of this region. 



