264 



Substitution of the Roman [No. 8, new seeies. 



Prof. Monier Williams' use of tlie acute accent as the sign, not 

 only of long vowels, but of the first s of the Sanskrit sibilants, e. g. 

 dar'si. A dot over the s and a horizontal line or acute accent over 

 the i would involve no perplexity. 



The Sanskrit vowels ri and ri, the vowel Iri, which is found in 

 only one word in the language, and M, which never occurs at all, 

 may safely be dispensed with. The difference between the so 

 called vowel ri, and the syllable which is composed of consonantal 

 r and * is more a fanciful than a real one, and so far as it is real, 

 it consists merely in a slight peculiarity in the pronunciation of i 

 after r in certain words. It does not appear to me that any dis- 

 tinguishing sign is required ; the ordinary characters r and i will 

 suffice. Williams distinguishes this vowel r by a dot underneath. 

 The abandonment of the dot in this connection will enable us to 

 represent in this manner one of the peculiar r's of the south In- 

 dian languages. 



Thus, the help of a single simple, unmistakeable sign enables 

 the Roman character to express each of the vowel sounds of the 

 Indian languages. Certainly none of the Indian systems can vie 

 with this in simplicity. In Tamil alone there are 1 7 different signs 

 of length employed for distinguishing the long forms of the five 

 simple vowels ! 



(b.) The second class of sounds for which the Roman alphabet 

 requires to make some special provision, is that of the " cerebral" 

 or lingual sounds. These are the t, th, d, dh and n of the third 

 row of Sanskrit consonants, with the corresponding letters in the 

 other Indian alphabets, to which I would add the peculiar I or rl 

 of the Vedas and the Dravidian languages, and the harsh lingual 

 r of the same family of tongues. 



The six lingual consonants differ from the dentals, the conso- 

 nants of the next row, only by a certain peculiarity of pronuncia- 

 tion. The most natural way of distinguishing them from the den- 

 tals would be by adding to them some distinguishing mark, and 

 this distinguishing mark should be affixed in the same way to them 

 all. Instead of adopting some such plan as this, the Indian 

 alphabets use two totally distinct sets of characters, thereby in- 



