Professor Buhler on the Sanskrit Linguals. 131 



cause whatever. We find, for instance, darvaghata, carva- 

 ghata for — ghata, the roots bhan for bhan (comp. <poW<i>), 

 pan ' to praise' for pan, ven for ven. In these and numerals 

 other words the form with the dental generally belongs 

 to the Yedic language. The existence of the double 

 forms, etymology and the results of comparative 

 philology prove that the dental sounds are the more an- 

 cient, and also that no external cause produced their trans- 

 formation. But we may safely conjecture that the gra- 

 dually increasing predilection of the Hindus for lingual 

 sounds is the true reason of these changes, since it is a 

 fact, very frequently observable in the history of languages, 

 that a phonetic innovation outgrows in course of time its 

 original and legitimate limits. Thus in Prakrit every 

 final m or n, which is not dropped, must be changed to 

 anusvara, whilst in Sanskrit this change takes place only if 

 certain letters follow. Again in French an original ' c follow- 

 ed by e a' becomes a sibilant, sh, though the conditions, 

 which caused the analogous transition of c c' to 9, ch, 

 before i and e, do not exist here. It is the general pre- 

 dilection of the Romance languages for palatals which pro- 

 duced this otherwise inexplicable phenomenon. 



So much for the facts offered by the Sanskrit language 

 in favour of my theory. We have seen that the ancient 

 linguals ' v and sh produced lingual mutes and nasals, 

 either independently or assisted by the universal laws of 

 assimilation, and that also ri and xi, the two lingual vowels, 

 the former of which at least belongs to the Pre-sanskritic 

 period, brought about the same result. Hence the Hindu 

 contracted a liking for these sounds and changed not only 

 * h/ which on account of this changeable nature easily 

 lent itself to this proceeding, to ' dh,' but also dentals to 

 the corresponding linguals. Moreover I have pointed out 

 repeatedly, how the predilection for linguals becomes 

 stronger and stronger in course of time, how in the daugh- 



