1848.] The most ancient Grammar of the Vedas. 17 



1. The real Anunasikya. 



Prat. I. 4, 6 : n at the termination of a word when following a long 

 a, exen when a vowel follows, is lost ; a, in the Anpadd : pada- 

 vritayas. We have as examples agran, gagrasanan, devahutaman, 

 badbadhanan, Indra soman, trshanan, nodeva devan, hanta devan (for 

 example devahutama acvan). 



Note. — Both the other Praticakhyas explain this case by a peculiar 

 process; Prat. II. 3; fol. 38, 4; akaropadho jakaram, i. e. the n 

 terminating a word after a long a becomes before a vowel, a y, and accord- 

 ing to an earlier adhikara, the upadha becomes nasal (likewise Prat. III. 

 1, 9). Mahan becomes indras consequently mahayindras ; after the Sut- 

 raja-vayo : padantayo : svaramadhye lopas (4. fol. 56. b.) the y drops 

 and it remains maha indras. After the same manner the omission of the 

 Visarga is treated in the very same passage (4. fol. 44, b.) kanthya-purvo 

 yakaram ariphitas, (n amely, visarg' aniy as) consequently chitra: adityanam, 

 chitray adityanam chitra adityanam. Panini explains the latter change in 

 the very same manner (VIII. 3, 17, 18, 19,) while he does not use it to 

 explain the omission of the n ; the difference here is certainly much great- 

 er. We have according to the opinion of the Indian Grammarians an- 

 other example for the same occurrence in the word pra-uga. While the 

 same is given in the Praticakhya I. 2, 1, without any further explanation 

 in connexion with puraeta, tita-unri and nama uktibhis, as vivritti within 

 the word ; the second Praticakhya (4. f. 57, a.) has the Sutra prayugam iti 

 yakara lopas. I doubt whether for this word, which in the Sanhita itself 

 is only twice met with (I. 7, 5, 6, and X. 11, 2, 3) we can give another 

 derivation than that of the Praticakhya, which the later grammarians also 

 adopt. The y appears like the v, to have been capable of such a soften- 

 ing (laghuprayatnataras, according to Cakatayana, with Pan. VIII. 

 3, 18,) that nothing of it remained but the hiatus between the vowels, 

 which it had separated (as in the vikara for ai, for instance anvetava u). 

 On that account we might regard the same on one side as a means 

 for explaining a hiatus, on the other side however we might, (as the 

 first Praticakhya) object to this substitution, and treat as hiatus (pada- 

 vrittayas) the same Sandhi, which the second will explain by putting in 

 a semi-vowel. It is however remarkable that the second and third 

 Praticakhyas, as well as Panini and the more ancient teachers quoted by 



D 



