206 Literary Intelligence, Ffj 0t 2 



agho u aghoy'%, aghoatra aghoy'atra, bho atrabhoy'atra, bhago atra 

 bhagoy'atra, gluci gita iti sandhipratishedhah. 



Note. — In the cases marked by # the MS. has y and v instead 

 of f, v'. 

 Translation. 



And if v and y (in this position) are followed by a vowel or diph- 

 thong, then the elision is not clearly audible ; (i. e. the pronunciation 

 of the v and y is unarticulated and the letters are hardly audible), 



I add the explanation of the word aspashtah given in the above- 

 mentioned compendium. There we read : 

 ...aspashtah aspashta§rutih pracithila sthanakaranaparispanda§ca asan- 



nah vakaro yakaracca 



Again it must be observed that Panini says VIII. 3. 22. hali sarves- 

 ham — All the (old) grammarians prescribe the loss of such a v and y, 

 if it is followed by consonants ; and this rule is certainly contained in 

 Qakatayana's Sutra, I. 1, 153. 



After this, I think, there can be hardly any doubt that the author 

 of the Qabdanucasana was the predecessor of Panini. 



But, in order to make doubly sure, I will adduce another proof for 

 this relation, which seems to me to be still more conclusive. 

 Panini teaches V. 2. 124 : vaco gminih. 



The word vac takes the affix gmini (in the meaning of matu). 

 The Calcutta Pandits who prepared the first edition of Panini 

 understood, the Sutra so, that the real form of the affix was gmin, and 

 consequently formed the monster vaggmin (with double g.). They 

 even misled Dr. Boethlingk (see his note to the Sutra). Benfey # 

 and Aufrechtf understood the Sutra rightly and formed vagmin. The 

 latter form alone occurs in literature, and is the only correct one. 

 The obscurity of the Sutra is caused by Panini' s negligence. He has 

 omitted to state that the letter ' g' is prefixed to mm only in order 

 to indicate that the final of vdc does not become nasal, as it ought, 

 according to the Sandhi rules. He has taken the Sutra, with a slight 

 alteration, from Qakatayana's grammar, where according to the Com- 

 pendium, it is read thus : vaco gmin. 



It is perfectly intelligible in Qakatayana's system, as there a pre- 

 fixed ' g' constantly means " no Sandhi." The author of the Com- 

 pendium says in commenting on the Sutra : 



f. Vollst. Sankt. gr. aff. min 9 f Unadisutras glossary s. v. vagmin* 



