78 Journal oj the Asiatic Society of Bengal. [N.S., XII, 



e, o only. Old Western Rajastham i, u are occasionally seen 

 to pass into 6, 6, and so are the vocalic groups ea, oa, both 

 when in hiatus and when separated by yasruti or vasruti. As 

 illustrations of the former case I may quote Maravarl *if, 



which is from Old Western Rajastham *<f)\ the conjunctive 

 participle of ^re generally used in the function of a conjune- 



Western 



<=5\ 



Maravarl— Gujarat! vm from tatsama w . In connection 'with 

 this change of i, u into e, 6, it is, however, to be remarked 

 that it is not a peculiarity of Maravarl and GujaratI but in- 

 stances of a confusion between these four vowels are not un- 

 common in Old Western RaiasthanT anrl 4naU. n ^x h .- ~_ii 



4, 6 are : 



;es," § 7 (2).) Illustrations of ea, oa passing into 



Mar. %<* " to give " < O.W. Raj. ^m or fqw. 

 Ding. Jrsrm ** Duryodhana " < O.W. Raj. ;«wr or ?afr*v, 

 Mar. ntfx " Mandora " < O.W. Raj. *m?r^ or W rV, 

 Mar. *ret " Malade " < O.W. Raj *jt^w. 



The last example is an irregularity, inasmuch as the va in 

 t* is not a vasruti, but consonantal ys and vs are often treat- 

 ed asyasruhs and vasrutis even in Old Western Rajastham. 



-there remain still a few observations to be made in regard 

 to the pronunciation of the e, d vowels. They are not always 

 pronounced in the same way. This had been already remarked 

 by Sir George Gnerson in his Linguistic Survey of India, Vol. IX, 



? *•' & A am ? ty Mr - Divatia in Ws paper cited above. The 

 fact is that both Maravarl and GujaratI show a tendency to pro- 

 nounce e and o less wide when they are final in a word, than 

 when they are medial Here by final I comprehend also an i 

 or o forming part of the penultimate syllable of a plurisyllable 

 word ending in a quiescent a. Thus the d in *ft ' ' son " is never 

 pronounced as wide as the d in *h " mirror," nor are the es in 

 «ft%"is sleeping" and *wi "distinguished soldier" pro- 



^Z C f ^ Wid f ** the i in J* " near '" Na y, ^d in some 

 cases final e and o are actually heard as narrow, as for instance 



in ifcprc, which word-though ^ is f rom ^-is always 

 pronounced as ?Nfm In Maravarl popular son^s % " is ' ' 

 is frequently pronounced % . It would therefore se'em that the 

 language has a tendency to prefer narrow vowels, and that the 



work. 



^ff into mx 



nftto The P rac ^cal conclusion I wish to draw from the above 



H;ilS n fV? n g * T t t0 , orth °g ra P h y- I have incidentally 

 mentioned that GujaratI does not distinguish e, 6 from e 6 in 



