on the Sanskrit Manuscripts in Madras. 



83 



-irmnasya for nrimnasya, maha for mahna, janasa for janasa 

 danda for indrah : in line 4 jata for jatah, rava for eva, 

 manakhan for manasvan, devan for devan : in line 5 rodase 

 dati for rodasi iti, nrinnasya for nrimnasya, maha for mah- 

 na, janasha for janasah, danda for indrah. Now it is very 

 curious that all those letters or groups of letters, which Mr, 

 Taylor mistakes one for the other, kte and kre and 



ma and sa ^T^f anc * ^T) sva-and kha an d ^«t^ 



da and i and ra and e and J?^ mi and mo 



and +\ \ ^ hear in Devanagari some resemblances to 

 each other. • This fact, I think, strengthens our hypothesis 

 regarding the origin of these little mistakes. 



. Now for the second charge, Mr. Taylor does not seem to 

 be aware that the words " tatraisha sukte prathama" do 

 not belong to the text of the Veda. They are the commen- 

 tator Sayana's and mean : " Here follows the first verse of 

 the hymn." 



Secondly, as Mr. Taylor speaks of " the alliteration (as to 

 words and letters) of a peculiar kind," it appears that he 

 was quite ignorant of the fact that in this, as well as the two 

 other passages, he has copied the same verse twice ovee, 

 first in the Samhitapatha, where the words are connected, 

 and next in the Padapatha, where the words are exhibited 

 separately. Sapienti sat ! 



Another class of mistakes occurs in the spelling of the 

 names and Sanskrit words. We find quite absurd and un- 

 grammatical forms, like Brahmam for Brahma (base Brah- 

 man) karmam for karma (base karman), patam (neuter) for 

 pada (masc.) vatam (neuter) for vada (masc), Bhattoji dik- 

 shada for Bhattojidikshita, kuvalaiyanantam (neuter) for 

 kuvalayananda (masc), Appaya dikshada for Apyaya dik- 

 shita etc. On the first five pages of Vol. II, we counted 



