1867.] Proceedings of the Asiatic Society. 33 



The following letter from Mr. Thomas on the derivation of Arian 

 Alphabets was read — 



11 1 am glad to find that my notice of the derivation of Arian Alpha- 

 bets attracted attention, and I am most curious to learn the course the 

 discussion took at the meeting of the Asiatic Society of Bengal ; more 

 especially as I am now following out the Indian section of the en- 

 quiry, and have arrived, already, at some unexpected results, tending 

 to confirm the original Dravidian derivation of the Sanskrit Alphabet. 

 The readers of our Journal will not fail to call to mind that Prinsep, 

 irt his early comments upon the Lat alphabet, pointed out that, in 

 many instances, the aspirate letters were formed by a duplication of the 

 lines of their corresponding simple letters. The question was not 

 raised as to when these aspirates had been designed, but the inference 

 was, that they had been formed simultaneously with the simple letters, 

 and out of the same elements. I have a different theory to propose, 

 which I submit for the examination and comments of your members ; 

 it is to assume that all the simple letters were Dravidian, and consti- 

 tuted a complete and sufficient alphabet for that class of languages^ 

 while the aspirates were later additions required for the due expression 

 of Magadhi and other northern dialects, as the Sanskrit in after times 

 added its own sibilants to the latter alphabet. A glance at the subjoined 

 comparative alphabets will shew the 20 consonants (out of the full 21) 

 of the Dravidian system, as opposed to the 31 consonants of the 

 Prakrit of Asoka's edicts. Of the additional aspirates of the latter 

 scheme, two only can in any way claim to be ordinary duplications ; 

 the chhj and th ; while a more simple origin might be sought for 

 the latter in a common circle : dh, dh and ph may fairly be taken as 

 intentional modifications of their corresponding normal letters, but 

 kh, and gh, like th, and th have more in common as fellow aspirates 

 than association with their own leading consonants ; and finally jh and 

 bh seem to have been unfettered adaptations. The s (rta) again differs* 

 from the y ( Jj) only in the reversal of the leading lower limb. As the 

 alphabetical data, upon which alone we have now to rely, are derived 

 from inscriptions embodying a different language, and dating so late 

 as B. C. 250, we can scarcely expect to recover the missing Dravidian 

 consonants, but one at least of the vowel tests is significant in the 

 extreme. The Dravidian vowels, as contrasted with the Sanskrit 

 series by Caldwell, arrange themselves as follows ; 



