Praudhamanorama p221

p221

[According to the lexicon,] "snūḥ and prasthaḥ, [and] sānu is non-feminine"; thus the word "sānu" has two genders. He gives its form when it is neuter: "snūnī sānūnī". In the masculine, it is like "śambhu".
[On] "priyakroṣṭūni": As for what was said by the older grammarian, namely "priyakroṣṭṝṇī", that contradicts the Mahābhāṣya and the vārttika which uphold the conflict of rules where the prior prevails.
As for what they explained—that the conflict of rules pertains only to this domain of "vibhāṣā tṛtīyādiṣvaci" (Pāṇini 7.1.97), because in the Mahābhāṣya only "priyakroṣṭune araṇyāya" is cited as an example—that is not so, because there is no authority for an example to restrict [the rule].
As they say: "Nor is an example to be taken as restrictive." Moreover, if the conflict of rules where the prior prevails pertained only to the optional treatment like tṛc, then the mention of treatment like tṛc in that context would be entirely useless, because [the form] "priyakroṣṭunā", etc., is easily established in the alternative even without a conflict of rules. Nor can it be said that the cessation of "priyakroṣṭrā", etc., is its purpose, because in the option of masculine-like treatment that is unavoidable, and because it was exemplified in that very way by [the Bhāṣyakāra] himself. Nor should it be said that in a collective dvigu compound such as "pañcakroṣṭunā", due to the absence of masculine-like treatment, the cessation of treatment like tṛc there is the purpose. For even so, there would be an inconsistency with the example in the Mahābhāṣya. If it is claimed that it is an indicator, alas, then let it simply indicate "priyakroṣṭūni" in order to accommodate the nominal stem portion exemplified in the Mahābhāṣya: this is the direction of thought.