arrived at his interpretations. It will be observed that where any- difference arises 

 between the Sanskrit and the Tibetan ^ the English translation is usually on the side 

 of the Tibetan. 



A great many phrases have to-day become absolutely stereotyped among writers 

 on Buddhism, which were altogether new to Csoma and his contemporaries, and 

 many words which he thought fit to explain are now-a-days familiar to all students 

 of that religion. In cases where we have improved on Csoma's English without 

 affecting his meaning, we have not thought it necessary to make any remark; but 

 where we have altered or added to what Csoma wrote, the portion for which we are 

 responsible has been placed within crochets. The editors are responsible for all the 

 footnotes. 



I have to thank Babu Surendra Nath Kumar, I^ibrarian of the Asiatic Society, 

 for his constant help in reading the proofs of the Sanskrit portion, while Dr. Satis 

 Chandra Vidyabhusana was absent from Calcutta ; and I must also acknowledge my 

 indebtedness to my teacher J^ama lyob-Zang Ge-Gen ' who read all the Tibetan portion 

 before it was dismissed to press. 



Finally, I have to thank my wife for many hours of patient labour spent in trans- 

 cribing Csoma's MS. and in noting the variant readings in Minayeff's version. 



The present instalment represents about one-third of Csoma's manuscript. When 

 the whole work has been printed we propose to add Aphabetical Indices to all the 

 Sanskrit and all the Tibetan words and phrases contained in the Vocabulary. 



Calcutta: E. Dknison Ross. 



October, 1910. 



[Note.— The Roman numbers given to the sectional headings have been added by the editors. The 

 numbers in brackets are those of Csoma's manuscript. In the Tanjur the groups are not numbered ; but 

 Csoma's numbers represent the order in which tlie groups occur in the Tibetan original. 



In the body of the work the English translations of the headings have been allowed to stand very 

 much as Csoma gives them ; but in the table of contents these headings have for the sake of convenience 

 been somewhat curtailed or modified. The variant readings which have been found in Minayeff's work 

 are prefixed by the letter M.]. ^ 



'1 fqSQ-^^^i'^l^^l 



