100 CONCERNING SOME OLD SANSKRIT. 



at Djamboe and of that at Pattadakal in the Deecan and also 

 of the oldest Cambodian inscriptions of Bhavavarman. The 

 two first words are quite clear ; they are prathame vayasi i. e. 

 1 in time of youth.' The reading of the next two groups of 

 letters which stand in the same line, is however uncertain. 

 I would read ndvvi since this combination is intelligible. The 

 second line I can make nothing of ; the three last groups of 

 letters might, allowing for defective writing, represent dvivi- 

 dham. 



The two first letter-groups in No. 4 are jaya. In No. 5 

 I read with some diffidence 48. No. 6 might represent siddhi. 



These fragments of inscriptions from Tokun do not, like 

 those from Kedah and from the temple ruins in Province 

 Wellesley, bear a clear stamp of Buddhist origin. The most 

 noteworthy point of this respect is the word arama — the read- 

 ing of which is unfortunately not beyond doubt. Fortunately 

 it is clear from the other inscriptions that Buddhist establish- 

 ments existed in the Malay Peninsula at the period to which 

 the earliest Brahman and Hindu remains in Western Java are 

 referred. 



With the exception of the inscriptions mentioned no others 

 have, I believe, been found in the Malay Peninsula itself, but 

 one which formerly stood on a large rock at the entrance of 

 Singapore Eiver, is worthy of description. In the Journal of 

 the Asiatic Society of Bengal for 1837 (14) there is a drawing 

 of this ancient record which shews that even then it had 

 suffered considerably. Later on, shortly previous to the year 

 1848, the stone was apparently removed and so damaged that 

 a few years later only fragments of it could be found. Mr. 

 Laidlay so far succeeded in deciphering some of the pieces 

 that he was able to give a facsimile. He rightly recognised 

 the Kawi characters in the writing and he came to the conclu- 

 sion that the language of the inscription was also Kawi. This 

 conclusion was certainly legitimate though Mr. Laidlay could 

 not have known that the Kawi alphabet was at one time used 

 in Java for Sanskrit inscriptions. 



(14) (Misc. Papers Indo-China Vol. Igepa 219). 



Jour. Straits Branch 



