Praudhamanorama p191

p191

[On the word] 'kit': The idea is that thereby elision of ā takes place by 'āto lopa iṭi ca' (Pā. sū. 6.4.64). [On the words] 'yayīpapyādayaḥ': As for what the older grammarian said, that similarly devayajī, atilakṣmī, and so on [are formed]—that is not so. For when analyzed as 'devān yajati' ("he worships the gods"), the word devayaji ending only in a short vowel was established by Ujjvaladatta and others by the uṇādi rule 'in' (Uṇ. 4.136). For Bhaṭṭi's usage 'admo dvijān devayajatīn nihanma' has been explained precisely as concerning a short vowel. Nor is there a long ī-affix expressly taught for yaj. If it is said that it is posited by virtue of the mention of 'bahulam'—it would indeed be so if an indicator not otherwise accountable were found. In the Veda too, only the short vowel is used: 'āyajiṃ tvā manave jātavedase'. But as for the word atilakṣmī, because it has the designation nadī, when shortening, the augment āṭ, the augment nuṭ, and the case ending ām are applied, there is a very great difference, and so it has no similarity with vātapramī. And this will become clear in the original text itself. [On the phrase] 'kvibanta--': And this is clear in Mādhava's work under the verbal root 'māṅ māne śabde ca'. [On the word] 'bahuśreyasī': There is no affix kap by 'īyasaśca' (Pā. sū. 5.4.154). Nor is there shortening of a subordinate member, because of the prohibition 'not in a bahuvrīhi ending in īyas'.
[On] 'yū stryākhyau' (Pā. sū. 1.4.3): [On the gloss] "īdūdantau": Here there is also a view that the technical name applies only to the two speech sounds, from the natural drift of the rule 'ācchīnadyornum' (Pā. sū. 7.1.80). But their expressing the feminine gender is due to their naturally belonging to the feminine suffixes like ṅī, etc., or somewhere by entering as a constituent part into an aggregate of speech sounds expressive of the feminine, since being the locus in which the expressive power resides fully is not intended to be conveyed. Thereby, common usage such as 'nadyantāt' ("after that which ends in a nadī") is not contradicted. Likewise in the designation ghi also, it should be understood that there are two views: either the two sounds i and u or that which ends in them is designated ghi. In the view that it is a designation of the speech sound, the meaning of 'asakhi' is "a part of that which is other than sakhi", not "excluding a part of the word sakhi". Therefore, it should be noted that usages such as 'atisakhinā' are unhindered. [On the word] 'pūrvam': The meaning is "prior to compound formation". [On the word] 'vaktavyam': When the technical name applies at the place of its initial enunciation (yathoddeśapakṣa), this is established by logical reasoning; but in the view that it applies at the time of the operation (kāryakālapakṣa), it is established purely by explicit statement.
On "aci śnu-" (Pāṇini 6.4.77): here [in the sūtra], "yvoḥ" is an attribute of the verbal root alone, not of śnu and bhrū, because they do not deviate [from being u and ī/ū]; therefore he says: "of a verbal root ending in the letter i or the letter u." But if it were explained as: "the letters i and u of śnu, etc.", then the marker ṅ in iyaṅ and uvaṅ would be useless, and having different loci [between the genitive terms] would be a flaw. Moreover, there would be over-application in kṣipati, adyutat, didivatuḥ, ūyatuḥ, etc. But to qualify the letter i and the letter u as followed by a vowel in order to prevent that would involve clumsiness—this is the point. "Due to proximity": due to place of articulation—this is the point.
On "eranekācaḥ-" (Pāṇini 6.4.82): although there is mention of a dvandva in the sūtra "aci śnudhātu-", only the verbal root is connected here, not śnu and bhrū, because the letter i is impossible in them. And the mention of the verbal root, by repetition, qualifies both: the conjunct consonant and the stem; explaining thus, he says: "a conjunct consonant forming part of a verbal root", etc. There, because the conjunct consonant is qualified by the verbal root, yaṇ applies here: unnyau, unnyaḥ. But the qualification of the stem is for the sake of establishing forms like harim, harī, harīn. For if the stem were not qualified by the verbal root, then here too—just as in praddhyam, praddhayaḥ, etc.—yaṇ would occur, barring the single substitute identical with the preceding vowel and the homogeneous long substitute identical with the preceding vowel. For the stem ends in a letter i that is not preceded by a conjunct consonant forming part of a verbal root.