Praudhamanorama p206

p206

[On] 'ramā': the affixes su etc. occur because it is a nominal stem by the metarule regarding that which is qualified by gender. As for [the statement that it is] from that which ends in āp; and as for what they explained: 'The meaning is that although it is not a nominal stem because of ending in an affix, [it gets su etc.] because ṅī and āp are mentioned separately'—that is incorrect, since it has already been stated that the mention of ṅī and āp has a different purpose.
The words "[from an elision after] eṅ or a short vowel" [are stated for this reason]: because it is subsequent and expressly stated, when the substitution of e is made by "sambuddhau ca" (Pā. Sū. 7.3.106), elision by "halṅyābbhyo..." (Pā. Sū. 6.1.68) would not apply, since, although the state of [ending in] āp is present through treatment like the original element, the form of the phoneme ā is absent—because, having analyzed [the rule] by coalescing ā and āp, the mention of dīrgha is rejected. The purport is that even according to the literal wording, only the elision by "eṅ-hrasvāt..." (Pā. Sū. 6.1.69) is justified, on account of its being subsequent.
On "yāḍāpaḥ" (Pā. Sū. 7.3.113): Although it would indeed be proper to say merely yaṭ—and one cannot object that the subsequent form would result by "ato guṇe" (Pā. Sū. 6.1.97), since that is blocked by the very capacity of the pronunciation of the short a—nevertheless, it might be presumed that the short a was only for ease of pronunciation; therefore, yāṭ was formulated for the sake of clarity. Regarding "vṛddhireci": As for what was cited by the older grammarian, namely "āṭaśca" (Pā. Sū. 6.1.90), that is incorrect, because there is no āṭ here, since the indicatory letter ṭ of yāṭ and syāṭ is an indicatory letter of the entire aggregate.
On "sarvanāmnaḥ syāḍḍhrasvaśca" (Pā. Sū. 7.3.114): The pronunciation of the long vowel is as before. On this, [some] say: Let the rule be formulated simply as "sarvanāmno'suṅ ca". And thus, when yāṭ follows, asuṅ being substituted for the final [phoneme] because it has the marker ṅ, forms like sarvasyai are established; [hence] the purpose of making syāṭ and prescribing a short vowel is to be considered. Here, this is the consideration: if one says "asuṅ ca", forms like amuṣmai would not be established. But in fact, it is suitable to formulate the rule as "āmi sarvanāmnaḥ suṭ yāṭo hrasvaśca". And it should not be objected that in this way the shortening would apply only to yāṭ, because of the capacity of pronouncing the long vowel; for even if a short vowel were pronounced, "ato guṇe" could indeed be averted merely by the capacity of the vowel a—this is to be kept in mind.