Praudhamanorama p202

p202

[On] 'oto ṇiditi vācyam' (one should say that [an affix] following o is treated as having an indicatory letter ṇ): the meaning is that one should say 'otaḥ', discarding 'gotaḥ'. And thus, raising the objection that, since vṛddhi would prevail by being subsequent in 'gām', the substitution of ā would not occur, the Bhāṣyakāra stated under the rule 'auto'mśasoḥ' (Pāṇini 6.1.93) that vṛddhi is barred by the substitution of ā because [the latter] is without any other scope. Kaiyaṭa also says there: 'It should be read in the rule as otaḥ, or gotaḥ should be explained as an implied indication of an o-ending.'
[On] 'smṛtau': as for what [some] say—'By the word go, through resemblance to an uṇādi stem, only the word dyo is implied, not anything ending in o in general; therefore only the forms smṛtoḥ, smṛtavau, smṛtavaḥ occur'—that contradicts the Mahābhāṣya. For if treating [an affix] as having an indicatory letter ṇ were accepted only after go and dyo, then 'smṛtām' and the like would be the scope of the substitution of ā by 'auto'mśasoḥ'; and thus, since vṛddhi in 'gām' would be unavoidable by being subsequent, the barring power and the absence of other scope of the substitution of ā stated in the cited Mahābhāṣya would not hold. This is the guiding principle. .. Thus end the stems ending in o ..