Vol. VIII, No. 9.] The Bakhshali Manuscript. 351 
[W.S.} 
Soma-varma and Asata!; the Mil-kihar fountain inscription * ; 
the Sai fountain inscription, 8 etc. 
(6) It is written in a mixture of Prakrit and Sanskrit and 
in the Gatha dialect whose ‘‘ literar form consisted in what 
n he ec 
characteristics of the underlying vernacular, etc.’’ Hoernle 
gives examples of many of these peculiarities which occur in the 
Bakhshali manuscript. Almost every peculiarity he cites has 
n 
many ‘of the same “peculiarities—incorrect Samdhi, interchange 
of sibilants, etc.; in the Brahmor copper-plate inscription of 
Yugakara Varman already referred to are trespasses against 
Samdhi rules, final oe are omitted, éa is substituted for 
sa, na and na are confused exactly as in the Bakhshali manu- 
script ; the copper-plate siaeionion of Vidagdha® likewise 
contains anumber of Sanskritized vernacular terms, it confuses 
n and n, interchanges re Ni etc.; the Kalait copper-plate 
inscription of Soma-varman ® puts ri for fz, cea pe sibi- 
lants, breaks samdhi rules, oh. ; and the inscription of Soma- 
varman and Asata’? and the Luj fountain sonaher aa of 
4.D. 1105-6 contain similar irregularities. 
(c) Hoernle refers to eae use of the words dinara and 
dramma as evidence of of composition when Greek in- 
fluence was still in force and goes from their use that the 
work must have been com osed ‘‘ within the first three cen- 
turies of the Christian era.’ The same argument would place 
Bhaskara’ s works (which were actually written in the twelfth 
bharata, it has been argued that that work was composed before 
the introduction of the denarius into India. Mr. B. 
Mazumdar quotes an interesting passage from the Databuuira- 
carita, in which the term sodasa sahasrani dinaranam occurs. 
1 ib., p. 192. oo 2 — : ib., p. 236. + ib., p. 197. 
6 ib., p. 165. ib., ib., p. 187. 
8 Lilavati, §§ 72, ": Fijavanit, Mer ete. 
9 i, 43; vi, 814; 106}, © 
10 Epigr raphia ia Indica, iy 167. 
“ Vogel, p. 204 
¢ J.R.A g. 1907, pp. 408 and 681. 


