0CT> — mar. 1858-59.] for the Indian characters. 



259 



though it does not follow that this defect is owing to the charac- 

 ter. Even if it were owing to this, however, the superiority of 

 the Indian characters to the Roman would not thereby be esta- 

 blished, for the Indian characters also are used with a consider- 

 able latitude of pronunciation. It is commonly said that each In- 

 dian character has only one sound, that when once that sound has 

 been learnt it must invariably be adhered to, and that every word 

 is pronounced precisely as it is written ; but there is a good deal 

 of exaggeration in all this eulogy. In Tamil, at least, every letter 

 has as wide a margin of sound as in any European tongue. 



The Tamil letter t has three sounds, and the proper place for 

 each sound can be learnt only by practice. At the beginning of a 

 word and when doubled, it is pronounced as t — more softly, how- 

 ever, than the English t; after a nasal it is pronounced as d; and 

 in the middle of a word, when followed by a vowel, it takes the 

 sound of the soft English th in " than." Only one character is 

 employed, and yet the least violation of these rules grates unplea- 

 santly on the Native ear. If the Roman t were substituted for the 

 Tamil one, it does not appear to me that the difficulty which now 

 exists would be in the least increased. 



The same latitude of pronunciation characterises the greater 

 number of the Tamil consonants. The vowels also have at least 

 as many different shades of pronunciation as the English vowels. 

 In pronouncing the vowel a, for instance there are four distinct 

 modes of pronunciation, each of which characterises certain classes 

 of words. In like manner, every Tamil vowel, long or short, might 

 be shown to have two or three different pronunciations, which are 

 in reality so many diphthongs. They are not distinguished how- 

 ever from one another by any difference of character, nor is any 

 notice taken of them in Native grammars, though the intelligible- 

 ness of what is spoken depends in a great degree upon the accu- 

 rate observance of these differences. 



The objection which I am now discussing has really no validity 

 at all, for. whatever be the latitude with which particular letters 

 are pronounced in the European languages, they will have that la- 

 titude only in the Indian languages which their Indian equivalents 

 have already. In this respect the substitution of the one charac- 



