Praudhamanorama p137

p137

On "vyor laghuprayatnataraḥ śākaṭāyanasya" (Pā. Sū. 8.3.18): "laghuprayatnataraḥ" means having an articulation of lighter effort. And that, on account of closest proximity, is "v" for "v", and "y" for "y"; thus he says: "va-yau".
On "oto gārgyasya" (Pā. Sū. 8.3.20): [Regarding the gloss] "okārāt" [i.e., 'after the letter o']: as for the explanation by the older grammarian that it means 'after a base ending in the letter o', that is not correct; because the base is not under consideration, and because it is impossible. For in the examples there is no base ending in the letter "o".
[On the phrase] 'of the letter y' [in the gloss on] hali sarveṣām (Pā. Sū. 8.3.22): the letter v, however, is not continued, because [a v] preceded by bho, bhago, or agho is impossible. Even though [a v] not preceded [by these] is possible, as in vṛkṣav karoti, nevertheless there is no occasion for elision there, because the consonant [hal] is qualified by aś which is continued [from the preceding rule]. But they say that vṛkṣav hasati and the like are ungrammatical due to non-usage. In gavyam, however, there is no end-of-a-pada status; and in gavyūti, it has been stated that there is no elision of v because of the coalescence of the letter v. If it is objected: In devā namyāḥ and the like, this is accomplished simply by "lopo vyoḥ..." (Pā. Sū. 6.1.66)—do not say so; because the substitution of y is unaccomplished with respect to that [rule], and because in devā yānti there is no condition of being followed by a val consonant.
[On] ahorātraḥ: Masculine gender is [assigned] by "rātrāhnāhāḥ puṃsi..." (Pā. Sū. 2.4.29).