Praudhamanorama p138

p138

[On] 'of ahan and the like': In both places, the word ādi denotes kind. Therefore, [passages] such as "svarcakṣo rathinaḥ satyaśuṣmaḥ" [and] "vipraḥ kaviḥ kāvyena svarcano" are to be included.
[On] tṛḍhaḥ, vṛḍhaḥ: Because they are marked with an indicatory u and therefore optionally take iṭ, there is no iṭ before a niṣṭhā suffix according to "yasya vibhāṣā" (Pā. Sū. 7.2.15). [Then occur] substitution of ḍh, substitution of dh, ṣṭutva, and elision of ḍh. As for the counter-example dṛḍhaḥ cited by the older grammarian, that is incorrect. For there is simply no elision of ḍh there. To explain: by the rule "dṛḍhaḥ sthūlabalayoḥ" (Pā. Sū. 7.2.20), elision of the n and h of dṛṃh- and substitution of ḍh for t are stated as irregular ready-made forms; the substitution of ḍh located in the section of unaccomplished rules does not apply here, because it is set aside by the substitution of gh by "dāderdhātoḥ..." (Pā. Sū. 8.2.32). Nor can it be said that the absence of iṭ and the substitution of ḍh are irregularly provided while substitution of dh, ṣṭutva, and elision of ḍh will indeed take place; for if that were so, the substitute ay for ṇi by "lyapi laghupūrvāt" (Pā. Sū. 6.4.56) would not occur in paridraḍhayya, and ṣyaṅ by "aṇiñoḥ..." (Pā. Sū. 4.1.78) would occur in pāridṛḍhī kanyā, since the penultimate would be heavy due to the elision of ḍh being unaccomplished—this is established in the Mahābhāṣya. As for what the older grammarian explained: 'This irregular form belongs only to dṛṃh-; but dṛha vṛddhau is a distinct verbal root, and when its iṭ is absent before a niṣṭhā suffix due to the non-obligatoriness of rules teaching augments, this is to be regarded as a counter-example when the substitution of gh is absent in that alternative by splitting the rule at "vā" in "vā druha..." (Pā. Sū. 8.2.33) and elision of ḍh occurs'—that is incorrect. For it is accepted that both are irregular ready-made forms. So says Vāmana: 'Now, dṛhi is a distinct verbal root; this irregular ready-made form applies to it also, excluding the elision of n.' And this is correct. For otherwise, there would be no substitute ay for ṇi after this, and ṣyaṅ would occur.
[On] 'only substitution of u': Since the accomplished and the unaccomplished are not of equal strength, "vipratiṣedhe param..." (Pā. Sū. 1.4.2) does not apply; because it is unopposed, only substitution of u takes place—this is the meaning. So it has been said: 'In [the domain of] "pūrvatrāsiddham", there is no conflict of rules, because the subsequent rule does not exist.' By this, [the statement] 'the operation of the case ending occurs first, and subsequently the substitution of u and the substitution of m' etc. is explained.
Indicating that su in the rule is an independent word whose genitive ending has been dropped, he says: 'that su which belongs to etad and tad'. The meaning is: that su which expresses the number residing in the meaning of etad and tad. But if it were explained as 'ordained after etad and tad', it would not occur in paramasa dadāti, paramaiṣa dadāti, and the like. The indicator for resorting to a relation mediated through meaning is to be understood as 'anañsamāse'.