[419] "Let the affix ṇini be used after a root in the sense of what is habitual, when the upapada, or subordinate word, is not a word meaning 'genus' and ends in a case."
[420] "Let vṛiddhi be the substitute of a base ending in a vowel, when that which has an indicatory ñ or ṇ follows;" ṇini has an indicatory ṇ.
[421] Sc. anuśaya + ini = anuśayin.
[422] Ini and ṭhan, which respectively leave in and ika; thus daṇḍa gives daṇḍin and daṇḍika. The line is quoted by Boehtlingk, vol. ii. p. 217, on Páṇ. v. 2, 115, and is explained in the Káśiká, ad loc. The different prohibitions are illustrated by the examples:—(1.) svaván, khaván; (2.) kárakaván; (3.) vyághraván, siṃhaván; (4.) daṇḍavatí śálá (i.e., daṇḍá asyáṃ santi).
[423] By iii. 3, 56.
[424] It is curious to see the great grammarian's favourite study obtruding itself here on such a slender pretext.
[425] See the Káśiká on Páṇ. v. 2, 115. For vivakshártha (meaning "general currency"), compare Commentary on Páṇ. ii. 2, 27. The edition in the Benares Pandit reads vishayaniyamártha.
[426] i.e., Thus nirodha is not vṛitter abhávaḥ, but abhávasyáśryaḥ.
[427] I read in p. 168, last line, prakáśapravṛittiniyamarúpa, from Bhoja's comment on i. 12.
[428] See Káśiká, ii. 3, 36.