40 Proceedings of the Asiatic Society. [Feb. 



their use. Had such been the case, there would have been some trace in 

 the formation of the letters to indicate their origin under different 

 states of civilization. Such, however, is entirely wanting. The as- 

 pirated letters in the simplicity of their configuration differ in no 

 respect from the surds and the sonants. The one set appears to have 

 been produced by the same intellectual effort as the other, and the two 

 are of character exactly alike. I admit that three out of the ten aspi- 

 rates, viz. chh, th and ph appear to be duplications or modifications of 

 the surds ch, t and p., but they constitute only one-fourth of the total 

 of 12 aspirates, the rest of which are perfectly independent in design 

 and shape. Mr. Thomas thinks the bh to be an inconsistent develop- 

 ment upon the basis of the old d, but there is no reason to show 

 why the aspirated sonant of the labial class should be formed on the 

 model of the unaspirated sonant of the dental, instead of the same 

 letter of its own class. I cannot therefore admit the argument to be 

 of any value. Again the s is supposed to be an adaptation of the y 7 

 " produced by the reversal of its leading lower limb." But the question 

 remains unanswered, why^the s should be formed on the model of y to 

 which it bears no phonetic resemblance whatsoever, instead of any 

 other letter ? The hypothesis in this case involves another difficulty ; it 

 assumes that the Sanskrit first coined only one s sound, leaving it to 

 be inferred that the other two sibilants were introduced into the 

 language a long time after, when we know for certain that the San- 

 skrit originally had three sibilants, two of which it lost in the Prakrits. 

 As to the vowel c , nothing can be more natural than that the long and 

 the short sounds of the same kind should be indicated by slight modi- 

 fications of the same figure. I cannot conceive that, to account for them, 

 it is necessary to assume their origin at different times under the in- 

 fluence of different nationalities. Those who can devise a system of 

 alphabetic writing may safely be presumed to have sufficient in- 

 telligence to make the same letter do duty for both a long and a short 

 sound by a slight modification. 



" One other argument in favour of the Tamilian origin of the Sanscrit 

 alphabet I have now to notice : it is the use of what are called cerebral 

 or lingual letters. It has been said that the Arians never used cerebral 

 letters ; we find them not in the Zend, the Greek, the Latin, and the 

 Teutonic ; ergo they should not be found in the Sanscrit ; but since 



