CONCERNING SOME OLD SANSKRIT. 101 



I have attempted to decipher the three fragments publish- 

 ed by Mr. Laidlay and to determine the language of the in- 

 scription but I must confess that I have not succeeded. Most 

 of the characters can be recognised singly but the gaps are so 

 numerous that no words can be positively recognised. Thus 

 I read in the third line of figure 1 the letter-groups saldgala- 

 lasayananara : in the second line of figure 2 ya-dmdnavana ; 

 in the third line kesarabharala in the sixth line of figure 3 

 yadalama. Granted that no vowel-marks and Anusswara's 

 have been omitted in the facsimile, I see no chance of so divid- 

 ing these letter-groups as to make an unmistakeable Javanese 

 word. I cannot however assert that the inscription is written 

 in any other language. 



In a work entitled ' The Malayan Peninsula ' by Captain 

 Begbie quoted by Mr. Laidlay, reasons are given for believing 

 that inscription dates from the reign of Cri-Baja Wikrama 

 (1223-1236). Palaeography is not opposed to the conjecture. 



As regards the question, which of the Kawi types — that of 

 Java or of Sumatra, the characters on the Singapore inscrip- 

 tion most resemble, some letters, notably ma, which in Java- 

 nese Kawi differ markedly from those found in Sumatra, re- 

 appear in their Javanese form on the Singapore stone and I 

 therefore believe that we must assign the inscription to the 

 Javanese type. Ma is the most characteristic letter in these 

 alphabets since it is different both in later Cambodian in the 

 time of suryavarman (15) and in the Sumatran Kawi. On the 

 other hand the form for sa is common to both Sumatran and 

 Javanese Kawi and different in the later Cambodian. 



It is to be feared that the Singapore record has been 

 damaged beyond hope of restoration ; so much the more reason 

 for fixing our attention on the little of it that remains in 

 transcription. 



(15) In my paper on the Koetei inscription, I assumed on the 

 strength of one date that Suryawarman reigned in the 8th century of 

 Caka; it appears however from the investigations of Messrs. Aymonier 

 and Bergaigne that this date is two centuries too early, see the re- 

 marks of the latter savant in the Journal Asiatique (Eebruary March 

 1882) Note 4. 



R. A. Soc, No. 49, 1907. 



