तिङन्तजुहोत्यादिप्रकरणम्

Covers SK 2489–2504 (16 sūtras): all present. Every unit has its own kaumudī. The spine is not an Aṣṭādhyāyī run. The SK follows the Dhātupāṭha order of the juhotyādi roots (1083 hu … 1106 gā) and teaches each rule at the root where it first bites, drawing on AS books 2, 3, 6, 7 and 8 (2.4.75 → 6.1.10 → 3.1.39 → 6.4.115 → 7.4.77 … 6.4.42). The AS-contiguous run 6.4.113–118, the ā-rules for abhyasta stems, is split: 6.4.114 id daridrasya was taught at SK 2482 (p45), 6.4.115 at 2492, and 6.4.113 and 6.4.116–118 at 2497–2500. Every AS-ref below is checked against references/ashtadhyayi/sutraani-data.json.

• Anuvṛtti from later in the SK. Several rules here carry words from sūtras the SK teaches only later. Examples: ī haly aghoḥ (6.4.113, SK 2497) gives hali at 2492 and 2498; bhṛñām it (7.4.76, SK 2496) gives it at 2493; nijāṃ trayāṇāṃ guṇaḥ ślau (7.4.75, SK 2502) gives ślau at 2493 and trayāṇām and ślau at 2496; atra lopo 'bhyāsasya (7.4.58, SK 2620) gives abhyāsasya at 2493, 2496 and 2502; viḍvanor anunāsikasyāt (6.4.41, SK 2982) gives āt at 2504. In the other direction, 2503 (7.3.87) had already fed piti and sārvadhātuke to the u-final and brū rules of p45 (2443–2452). The outline names them where BM does.

tiṅantajuhotyādiprakaraṇam — outline (p46, SK 2489–2504, all 16 sūtras)

p45 (tiṅantādādiprakaraṇam) closed at 2488 with "ity adādayaḥ".

• Where the prakaraṇa begins. The print heads it "tiṅante juhotyādayaḥ 43". Our prakaraṇa number 46 differs systemically. sk2489 opens with the Dhātupāṭha entry 1083 hu dānādanayoḥ, then juhotyādibhyaḥ śluḥ (2.4.75). "itaś catvāraḥ parasmaipadinaḥ": hu, ñibhī, hrī and pṝ (1083–1086) are parasmaipadin.
• Where it ends. At 2504 with 1106 gā stutau (Vedic): jigāti, jigītaḥ, jigati. The print closes "iti juhotyādayaḥ" ("Here end the juhotyādi roots"). Ours lacks it, so the English supplies it in brackets (scan page-0411). p47 opens with the print heading "tiṅante divādayaḥ 44" and 1107 divu (SK 2505).
• Contents in outline.
• 2489–2491: hu. ślu for śap, doubling before ślu, optional ām in liṭ with śluvadbhāva, and guṇa over yaṇ by paratva: juhoti, juhvati, juhavāṃcakāra / juhāva, juhudhi, juhavāni, ajuhavuḥ.
• 2492: bhī (and hrī). Optional i before a hal-initial kṅit sārvadhātuka: bibhitaḥ / bibhītaḥ; jihreti, jihriyati.
• 2493–2495: pṝ (and ṛ). i in the abhyāsa, u for ṝ after a labial, and an optional short vowel in the kit liṭ: piparti, pipūrtaḥ, papāra, papratuḥ / paparatuḥ.
• 2496–2497: bhṛñ, māṅ, ohāṅ. i in the abhyāsa, then ī for the ā of an abhyasta: bibharti, bibharāmāsa / babhāra, mimīte, mimate, jihīte.
• 2498–2500: ohāk (jahāti); then dā and dhā. i or ī before a hal-initial kṅit, ā/i/ī before hi, and loss before y: jahitaḥ / jahītaḥ, jahāhi / jahihi / jahīhi, jahyāt; dattaḥ, dehi, adita.
• 2501: dhā. bh(aṣ) for the d of dadh: dhattaḥ, dhatse, dhaddhve.
• 2502–2503: ṇij, vij, viṣ; the Vedic roots. guṇa of the abhyāsa before ślu, and no light-penultimate guṇa before a vowel-initial pit ending: nenekti, nenijāni, veveṣṭi. The run from ghṛ to jana follows: iyarti, babdhaḥ, tutūrtaḥ, dadhantaḥ, jajanti.
• 2504: jan (and gā). ā for the final of jan, san and khan: jajātaḥ, jajñati, jajāyāt / jajanyāt; jigāti.

Threads that run through the prakaraṇa:

• ślu, doubling and abhyasta are the class-mark. juhotyādibhyaḥ śluḥ (2489) deletes śap by ślu, and ślau (2490) doubles the root. The doubled stem is abhyasta (ubhe abhyastam, 6.1.5, SK 426). It therefore takes:
• at for jha (adabhyastāt, 7.1.4, SK 2479): juhvati, bibhyati, bibhrati, mimate, jahati, dadati, dadhati, nenijati, iyrati, bapsati, jigati;
• jus in laṅ (sijabhyastavidibhyaś ca, 3.4.109, SK 2226), with guṇa before jus (jusi ca, 7.3.83, SK 2481): ajuhavuḥ, apiparuḥ, abibharuḥ, adaduḥ, anenijuḥ, aiyaruḥ.
• What the abhyāsa vowel becomes. By default it is shortened and made palatal or unaspirated: ju-hoti, ja-hāti, da-dāti. Four rules set it otherwise:
• i by bhṛñām it (2496) for bhṛñ, māṅ and ohāṅ: bibharti, mimīte, jihīte;
• i by artipipartyoś ca (2493) for ṛ and pṝ: iyarti, piparti;
• guṇa by nijāṃ trayāṇāṃ guṇaḥ ślau (2502) for ṇij, vij and viṣ: nenekti, vevekti, veveṣṭi;
• i in the Veda by bahulaṃ chandasi (7.4.78, SK 3598): jigharmi, jiharti, jigāti.
• All four hold only before ślu. Hence babhāra and nineja in liṭ. Hence also bibharāmāsa, where śluvadbhāva (2491) brings ślu's operations into the ām-liṭ.
• The ā-final abhyasta roots. Before a vowel-initial kṅit sārvadhātuka the ā is lost (śnābhyastayor ātaḥ, 6.4.112, SK 2483): mimāte, jahati, dadati. Before a hal-initial kṅit it becomes ī (2497): mimīte, jihīte, jigītaḥ. The ghu roots dā and dhā (dādhā ghv adāp, 1.1.20, SK 2373) are excepted, so the ā is simply lost: dattaḥ, dhattaḥ. jahāti has its own set: i or ī (2498), ā, i or ī before hi (2499), and loss before y (2500). The ghu roots also take e before hi with loss of the abhyāsa (ghvasor ed dhāv abhyāsalopaś ca, 6.4.119, SK 2471: dehi, dhehi). In luṅ they take luk of sic (adāt) or i with loss of sic (adita, adhita).
• paratva decides two conflicts. guṇa, the later rule, beats the yaṇ of huśnuvoḥ sārvadhātuke (6.4.87, SK 2387) in juhavāni and ajuhavuḥ (2491). guṇa and vṛddhi beat the u of 2494 in piparti and papāra.
• The kaumudī lists forms bare. Lakāra and person labels are the translator's, given in Pāṇinian terms. Derivation stages follow BM and TB as recorded in the unit notes.
• Where the commentary lies. Only the kaumudī blocks are translated (SK-TRANSLATE §25).
• TB on the hu preamble (NSP p. 399, scan page-0409) is in no source.json. It is read on the scan and transcribed in a textual note on sk2489. Its last clause is at the tail of sk.2488.t1.
• TB printed under a neighbouring unit: 2491 at the tail of sk.2488.t1; 2499 in the second half of sk.2498.t1; 2502 at the tail of sk.2501.t1. TB on ḍubhṛñ is at the end of sk.2495.t1, which is correct, since the headword is in 2495's kaumudī. The print has no TB on 2490.
• BM lead-ins printed under the preceding unit: 2490 (tail of sk.2489.b1), 2493 (tail of sk.2492.b1), 2497 (tail of sk.2496.b1), 2498 (tail of sk.2497.b1), and 2502 plus the 2503 lead-in (tail of sk.2501.b1; sk2502 has no BM block).
• BM missing or eye-skipped in our source and read from Wikisource (uttara-2/0018–0029): the hu preamble (2489), ślau (2490), ā ca hau (2499), ohāṅ/ohāk (2497), the 2498 anuvṛttis and the vacanasāmarthya passage (2501).
• Artefacts: sk.2489.b1 carries its gloss twice. The final 'siddham' of sk.2504.t1 is the misfiled last word of the TB on the divādi opening (nṛtī, SK 2505; NSP p. 402), which our source otherwise lacks. It is not an OCR slip. The heading 'atha tiṅante kaṇḍvādayaḥ' at the end of sk.2504.b1 is a digital-BM artefact. Wikisource closes BM with "…bālamanoramāyāṃ śluvikaraṇaṃ samāptam".

Where the p46 audit (.work/sk-run/p46-audit-fixes.md and the two reports in .work/sk-run/p46-audit/) corrects a reading, this outline follows the corrected reading. The fixer is applying these to the en.yml notes in parallel.

• 2498 (must-fix): the anuvṛttis. Wikisource BM (uttara-2/0023) gives four: it from id daridrasya (SK 2482); anyatarasyām (= vā) from bhiyo 'nyatarasyām (SK 2492); kṅiti from gamahana… (SK 2363); hali from ī haly aghoḥ (SK 2497). Our BM's "hali from bhiyo 'nyatarasyām" is an eye-skip.
• 2489: svabhāvāt means "from beginningless, established usage" (anādisiddha-loka-vyavahārāt; BM, Wikisource), not "from the word's own nature".
• 2490: BM (Wikisource) confirms dve from ekāco dve prathamasya (SK 2175) and dhātoḥ from liṭi dhātor anabhyāsasya (SK 2177). It calls the yaṇ of juhvati an apavāda of uvaṅ.
• 2494: raparatva (uraṇ raparaḥ, 1.1.51, SK 70) means the u is followed by r. It is not an r-augment.
• 2495: the rule TB cites is ṝta id dhātoḥ (7.1.100, SK 2390), with long ṝ. The reconciliation of TB's i with BM's u is the translator's.
• 2496: in bibhrati, at replaces the jh of jhi, not jhi.
• 2497: BM (Wikisource) on ohāṅ: ātmanepadin, being ṅit; i by bhṛñām it; forms like māṅ; cutva in the abhyāsa. On ohāk: o and k are it.
• 2499: BM (Wikisource): ā is luptaprathamāka; jahāteḥ runs on from 2498; ca draws in (anukarṣa) the i of id daridrasya and the ī of ī haly aghoḥ.
• 2501: BM (Wikisource) adds that with the ā-loss sthānivat, dadh would be neither jhaṣ-final nor followed by t. It calls the vacanasāmarthya route a "useless trouble" (vṛthā) once pūrvatrāsiddhīye na sthānivat is applied.
• 2504: āśīrliṅ geyāt is our BM's reading only (Wikisource gāyāt).

Print readings (NSP, scan-confirmed) that the outline follows:

• ṛcchaty ṝtām 2383 (2494; ours ṝcchāty- … la);
• apiparuḥ and chāndasatvaṃ (2495);
• jahāter hau (2499);
• syāt and nenigdhi (2502);
• syāt, ityasyaiva, lṛditvād and jigharmy agniṃ (2503);
• the colophon "iti juhotyādayaḥ" (2504).

Still open for Ajai (mūla):

• 2494: ours glosses "ya ṝt tadantasya" (long), where the print and Wikisource have short ṛ. Grammar supports ours: BM carries ṝtaḥ from 7.1.100, and both commentators' counter-example is the long-vowel root ṝ. The English keeps ṝ.

Ordering: SK order throughout.

A. hu: ślu, doubling, ām with śluvadbhāva, guṇa by paratva — running example: hu + śap + ti → hu + ti → hu-hu-ti → ju-ho-ti → juhoti; juhu + jhi → juhu + ati → juhvati

• 2489 juhotyādibhyaḥ śluḥ 2.4.75 — after hu and the roots that follow it, ślu replaces śap. śapaḥ comes from adiprabhṛtibhyaḥ śapaḥ (2.4.72, SK 2423; BM). ślu is a name for deletion (pratyayasya lukślulupaḥ, 1.1.61, SK 260).
• Root entry 1083 hu "giving and eating". Some add "taking", and the Bhāṣya adds "pleasing".
• "Giving" here is casting (prakṣepa). That it is into a receptacle enjoined by rule, and of an oblation, comes from established usage (svabhāvāt; BM, Wikisource).
• TB (scan only): a stone thrown into the āhavanīya is a casting but not a homa. hoṣye is incorrect (asādhu), but juhvānāḥ is correct as a cānaś form.
• hu + śap + ti → hu + ti: this leads into 2490 (BM).
• 2490 ślau 6.1.10 — before ślu, the dhātu is doubled. dve comes from SK 2175 and dhātoḥ from SK 2177 (BM, Wikisource).
• hu + ti → hu-hu-ti. The abhyāsa operations follow (BM, Wikisource), then guṇa before the pit tip: juhoti, juhutaḥ.
• juhu + jhi: at for the jh by adabhyastāt (SK 2479), then yaṇ by huśnuvoḥ… (SK 2387), an apavāda of uvaṅ: juhvati.
• 2491 bhīhrībhṛhuvāṃ śluvac ca 3.1.39 — after bhī, hrī, bhṛ and hu, ām is optional in liṭ. Before that ām, the operations are as before ślu.
• The sūtra's sixth case stands for the fifth. ām and liṭi come from kāspratyayād ām amantre liṭi (3.1.35, SK 2306; BM). The translator takes vā from SK 2341.
• hu + ām → hu-hu-ām → juhavām (doubling as before ślu, guṇa) → juhavāṃcakāra. Without ām: juhāva.
• hotā, hoṣyati; juhotu, juhutāt.
• juhu + hi → juhudhi: hi → dhi by hujhalbhyo her dhiḥ (SK 2425).
• juhavāni: āṭ (āḍ uttamasya pic ca, SK 2204) is pit, so the ending is not ṅit. guṇa is the later rule and beats the yaṇ of huśnuvoḥ (BM; TB at the tail of sk.2488.t1).
• ajuhavuḥ: jus by SK 2226, then guṇa by jusi ca (SK 2481), which again beats the yaṇ by paratva.
• juhuyāt; āśīrliṅ hūyāt; luṅ ahauṣīt (vṛddhi before sic).
• Root entry 1084 ñibhī bhaye: bibheti.

B. bhī and hrī: optional i before a hal-initial kṅit — running example: bi-bhī + tas → bibhitaḥ / bibhītaḥ

• 2492 bhiyo 'nyatarasyām 6.4.115 — for bhī, i optionally replaces the final before a hal-initial kṅit sārvadhātuka.
• The anuvṛttis are: it from id daridrasya (6.4.114, SK 2482), hali from SK 2497, kṅiti from gamahana… (SK 2363) and sārvadhātuke from ata ut sārvadhātuke (SK 2467) (BM).
• bibhitaḥ / bibhītaḥ.
• The conditions are shown by TB's counter-examples. Not sārvadhātuka: bebhīyate, bibhiye. Not hal-initial: bibhyati (at, then yaṇ). Not kṅit: bibheti.
• liṭ bibhayāṃcakāra (ām and ślu-doubling by 2491) / bibhāya; luṭ bhetā.
• 1085 hrī lajjāyām: jihreti, jihrītaḥ, jihriyati (at, iyaṅ; BM); jihrayāṃcakāra / jihrāya.
• Root entry 1086 pṝ pālanapūraṇayoḥ. The scan cannot show whether the vowel is long; ours reads pṝ.

C. pṝ (and ṛ): i in the abhyāsa, u for ṝ after a labial, the short vowel in liṭ — running example: pṝ-pṝ-ti → pir-pṝ-ti → pi-pṝ-ti → pi-par-ti → piparti; pi-pṝ-tas → pi-pur-tas → pipūrtaḥ; pa-pṝ-atus → papratuḥ / paparatuḥ

• 2493 artipipartyoś ca 7.4.77 — for ṛ (arti) and pṝ (piparti), i replaces the final of the abhyāsa before ślu.
• abhyāsasya comes from SK 2620, it from bhṛñām it (SK 2496) and ślau from nijāṃ trayāṇām… (SK 2502) (BM).
• BM: the abhyāsa's ṛ → i with r after it, the r is lost by halādiḥ śeṣaḥ, and the second part takes guṇa with r: piparti (the form is given under 2494).
• TB: arti's form, iyarti, comes under 2503.
• 2494 ud oṣṭhyapūrvasya 7.1.102 — u replaces the final of an aṅga ending in ṝ when a labial that is part of the aṅga precedes the ṝ.
• ṝtaḥ comes from ṝta id dhātoḥ (7.1.100, SK 2390). The heading aṅgasya is read twice: once for the labial's being part of the aṅga, once for "ending in ṝ" (BM).
• "Part of the aṅga" excludes samīrṇaḥ (sam + ṝ + kta). There the m is not part of the aṅga, so i comes, not u (BM, TB).
• guṇa and vṛddhi are later and prevail: piparti (guṇa before tip), papāra (vṛddhi before ṇal).
• pi-pṝ + tas (ṅit, so no guṇa) → pi-pu-r-tas (u, followed by r) → pipūrtaḥ (lengthening by hali ca, SK 354).
• pipurati (at; ṅit, so u).
• In the kit liṭ, ṛcchaty ṝtām (SK 2383, print) makes guṇa obligatory. That leads into 2495.
• 2495 śṝdṝprāṃ hrasvo vā 7.4.12 — for śṝ, dṝ and pṝ, a short vowel optionally replaces the ṝ before a kit liṭ.
• liṭi comes from dayater digi liṭi (SK 2388). kiti is supplied because the rule is an apavāda of the guṇa of ṛcchaty ṝtām, which is for the kit liṭ (BM). śṝ and dṝ are kryādi roots (TB).
• papratuḥ, papruḥ: short ṛ, then yaṇ. Otherwise guṇa: paparatuḥ, paparuḥ.
• Why not simply make the guṇa optional? Without guṇa, i (TB, by 7.1.100) or u (BM, by 2494) would come, and an aṅga rule beats the varṇa rule yaṇ.
• luṭ paritā / parītā: optional lengthening of iṭ by vṝto vā (SK 2391).
• laṅ apipaḥ, apipūrtām, apiparuḥ (print; jus, guṇa by jusi ca). pipūryāt; āśīrliṅ pūryāt; luṅ apārīt, apāriṣṭām.
• Some teachers read the root with short ṛ: piparti, pipṛtaḥ, piprati, pipṛyāt, priyāt, apārṣīt (aniṭ). On Pāṇini's ṝ-final reading, "taṃ rodasī pipṛtam" can only be Vedic (chāndasatvaṃ, print).
• Root entry 1087 ḍubhṛñ dhāraṇapoṣaṇayoḥ. It is ubhayapadin (BM). TB: the ḍu gives bhṛtrimam by ḍvitaḥ ktriḥ.

D. bhṛñ, māṅ, ohāṅ: i in the abhyāsa; ī for the ā of an abhyasta — running example: bhṛ-bhṛ-ti → bi-bhṛ-ti → bibharti; mā-mā-te → mi-mā-te → mimīte

• 2496 bhṛñām it 7.4.76 — i replaces the abhyāsa of bhṛñ and the two roots after it (māṅ, ohāṅ) before ślu.
• The plural gives "bhṛñ and the roots after it". abhyāsasya comes from SK 2620, and trayāṇām and ślau from nijāṃ trayāṇāṃ guṇaḥ ślau (SK 2502) (BM, TB). Some get "three" from the kapiñjala maxim (TB).
• bibharti, bibhṛtaḥ, bibhrati (at for the jh of jhi, then yaṇ); ātmanepada bibhṛdhve.
• bibharāmāsa: by śluvadbhāva (2491) the ām-liṭ gets doubling and i.
• babhāra: no ślu, so no i. babhartha and babhṛva take no iṭ (BM).
• loṭ bibhṛhi (hi is apit, so ṅit, so no guṇa) and bibharāṇi (āṭ is pit, so guṇa).
• laṅ abibhaḥ, abibhṛtām, abibharuḥ (jus; jusi ca). bibhṛyāt.
• āśīrliṅ bhriyāt (riṅ by riṅ śayagliṅkṣu, SK 2367) and bhṛṣīṣṭa (kit by uś ca, SK 2368).
• luṅ abhārṣīt, and abhṛta (sic lost by hrasvād aṅgāt, SK 2369).
• Root entry 1088 māṅ māne śabde ca. It is ātmanepadin, and the stem is mi-mā-te (BM).
• 2497 ī haly aghoḥ 6.4.113 — ī replaces the ā of śnā and of an abhyasta before a hal-initial kṅit sārvadhātuka, except in a ghu root.
• ī is luptaprathamāka. śnābhyastayor ātaḥ comes from SK 2483, kṅiti from SK 2363 and sārvadhātuke from SK 2467 (BM).
• mi-mā + te → mimīte.
• Before a vowel-initial ending the ā is lost by 2483: mimāte, and mimate (at).
• TB: ghumāsthā… (SK 2462) does not act here, since 'ārdhadhātuke' runs into it. TB's counter-examples: lunīte (śnā), lunāti (not kṅit), dhattaḥ, dattaḥ (ghu).
• luṅ praṇyamāsta.
• 1089 ohāṅ gatau: ātmanepadin, i by 2496, cutva in the abhyāsa (BM, Wikisource). jihīte, jihāte, jihate; jahe, hātā, hāsyate.
• Root entry 1090 ohāk tyāge, parasmaipadin. Its o and k are it (BM, Wikisource). jahāti.

E. ohāk (jahāti): i or ī, ā/i/ī before hi, loss before y; then dā and dhā — running example: ja-hā + tas → jahitaḥ / jahītaḥ; ja-hā + hi → jahāhi / jahihi / jahīhi; ja-hā + yāt → jahyāt

• 2498 jahāteś ca 6.4.116 — for jahāti too, i optionally replaces the ā before a hal-initial kṅit sārvadhātuka. Otherwise ī comes by 2497.
• The anuvṛttis follow the audit: it from SK 2482, anyatarasyām from SK 2492, kṅiti from SK 2363 and hali from SK 2497 (BM, Wikisource).
• jahitaḥ / jahītaḥ.
• jahati: at, then the ā is lost by 2483.
• liṭ jahau. TB's counter-examples: jahāti (not kṅit), hīyate (not sārvadhātuka).
• BM's pratīka jahitāt is not in ours or the print.
• 2499 ā ca hau 6.4.117 — before hi, the ā of jahāti becomes ā, and by ca also i and ī.
• ā is luptaprathamāka. jahāteḥ runs on from 2498, and ca draws in the i of SK 2482 and the ī of SK 2497 (BM, Wikisource).
• jahāhi, jahihi, jahīhi. TB: hence Bhaṭṭi's "jahihi jahīhi jahāhi rāmabhāryām".
• laṅ ajahāt, ajahuḥ. The print reads "jahāter hau".
• 2500 lopo yi 6.4.118 — the ā of jahāti is lost before a y-initial sārvadhātuka. jahāteḥ comes from SK 2498, ātaḥ from SK 2483 and sārvadhātuke from SK 2467. It is an apavāda of 2498 (BM).
• jahyāt.
• āśīrliṅ heyāt (TB: not sārvadhātuka). It takes e by er liṅi (SK 2374).
• luṅ ahāsīt.
• 1091 ḍudāñ dāne:
• praṇidadāti (ṇ by nergada…, SK 2285).
• dattaḥ: ghu (SK 2373), so no ī by 2497. The ā is lost by 2483 (BM, TB). dadati, datte; dadau.
• dehi: e and loss of the abhyāsa by ghvasoḥ… (SK 2471).
• adadāt, adaduḥ; dadyāt; deyāt.
• luṅ adāt, aduḥ (luk of sic by gātisthā…, SK 2223), and adita (i by sthāghvor ic ca, SK 2389; sic lost by SK 2369).
• 1092 ḍudhāñ dhāraṇapoṣaṇayoḥ (some add "giving"): praṇidadhāti. BM gives the stem dadh + tas, which leads into 2501.

F. dhā: bh for the d of dadh — running example: da-dhā + tas → dadh + tas → dhadh + tas → dhattaḥ

• 2501 dadhas tathoś ca 8.2.38 — for dadh (dhā once doubled) ending in a jhaṣ, bhaṣ replaces the baś before t or th, and by ca also before s and dhv.
• jhaṣantasya baśo bhaṣ comes from ekāco baśo bhaṣ jhaṣantasya sdhvoḥ (8.2.37, SK 326), which also supplies sdhvoḥ through ca (BM).
• "dadha" is the doubled root, not the root dadha dhāraṇe (TB, read on the scan).
• dadh + tas: the abhyāsa's d → dh, and the root's dh → t by cartva (BM): dhattaḥ, dhatthaḥ.
• dadhati (at, ā-loss).
• dadhvaḥ: no bhaṣ, since no t, th, s or dhv follows.
• Ātmanepada dhatte, dhatse (before s), dhaddhve (before dhv).
• dhehi (2471); luṅ adhita.
• Why the ā-loss does not block it. The ā-loss (2483) looks sthānivat, which would leave dadh not jhaṣ-final and not followed by t. Vāmana and Mādhava answer by the force of the statement (vacanasāmarthya). Bhaṭṭoji's answer is that a pūrvatrāsiddha operation has no sthānivadbhāva, and BM (Wikisource) calls the other route needless.
• Heading: three svaritet roots. 1093 ṇijir śaucapoṣaṇayoḥ.

G. ṇij, vij, viṣ: guṇa of the abhyāsa, no light-penultimate guṇa; the Vedic run ghṛ … jana — running example: nij-nij-ti → ni-nij-ti → ni-nej-ti → ne-nej-ti → nenekti; nenij + āni → nenijāni (not *nenejāni)

• 2502 nijāṃ trayāṇāṃ guṇaḥ ślau 7.4.75 — guṇa replaces the vowel of the abhyāsa of ṇij, vij and viṣ before ślu. The plural means "those beginning with nij" (tadādi), and abhyāsasya comes from SK 2620 (BM, at the tail of sk.2501.b1).
• After ślu, doubling and halādiḥ śeṣaḥ, the second part takes guṇa before tip. The abhyāsa takes guṇa by this rule, and the j takes kutva (coḥ kuḥ): nenekti (BM). neniktaḥ; nenijati (at).
• liṭ nineja: no ślu, so no guṇa of the abhyāsa (TB). nektā, nekṣyati; nenektu.
• loṭ nenigdhi (print): hi is apit, so ṅit, so no guṇa, and hi → dhi.
• 2503 nābhyastasyāci piti sārvadhātuke 7.3.87 — an abhyasta takes no light-penultimate guṇa before a vowel-initial pit sārvadhātuka.
• guṇaḥ comes from mider guṇaḥ (SK 2346) and laghūpadhasya from pugantalaghūpadhasya ca (SK 2189) (BM).
• TB's counter-examples: dveṣāṇi (not abhyasta), veveṣṭi (not vowel-initial), nineja (not sārvadhātuka). TB says piti is there for uto vṛddhir luki hali (SK 2443 in p45).
• nenij + āni: āṭ is pit, so the ending is not ṅit, and the guṇa that would be due is barred: nenijāni.
• laṅ anenek, aneniktām, anenijuḥ. nenijyāt; āśīrliṅ nijyāt.
• luṅ anijat (aṅ, being ir-marked) / anaikṣīt; ātmanepada anikta (sic lost by jhalo jhali, SK 2281).
• 1094 vijir: vevekti, vevikte.
• vivejitha (liṭ with iṭ) takes guṇa. vija iṭ (SK 2536) does not make the iṭ-initial ending ṅit, because that rule takes only ovijī.
• ṇiji and viji are also rudhādi.
• 1095 viṣḷ: veveṣṭi, veviṣṭe; luṅ aviṣat (aṅ, being ḷ-marked).
• In the ātmanepada, ksa (SK 2336), whose a is lost before a vowel by ksasyāci (SK 2337): avikṣata, avikṣātām, avikṣanta.
• "ā gaṇāntāt": the roots up to the end of the juhotyādi gaṇa are parasmaipadin and Vedic (BM).
• 1096 ghṛ: "jigharmy agniṃ haviṣā ghṛtena" (print). The i comes from bhṛñām it read with bahulaṃ chandasi (SK 3598). The ordinary form is jagharti (BM).
• 1097 hṛ: "ayaṃ sruvo abhijiharti homān", with a Vedic i likewise.
• 1098 ṛ and 1099 sṛ. The i that 2493 gives ṛ would already come from bahulaṃ chandasi, so naming arti shows that ṛ with ślu is used in ordinary speech too (BM, citing the Bhāṣya).
• ṛ-ṛ-ti → i-ṛ-ti (2493) → iyarti: iyaṅ for the abhyāsa (abhyāsasyāsavarṇe, SK 2290), guṇa for the second part. iyṛtaḥ (ṅit); iyrati (at, yaṇ).
• liṭ āra, āratuḥ, āritha (iṭ invariable, SK 2384); artā, ariṣyati (SK 2366); iyarāṇi.
• laṅ aiyaḥ, aiyṛtām, aiyaruḥ; iyṛyāt.
• āśīrliṅ aryāt (guṇo 'rtisaṃyogādyoḥ, SK 2380).
• luṅ ārat (aṅ by SK 2382, guṇa by SK 2406).
• From sṛ: sasarti.
• 1100 bhasa: babhasti.
• babhas + tas: the penultimate a is lost by ghasibhasor hali ca (SK 3550) and the s by jhalo jhali. The t → dh (jhaṣas tathoḥ, SK 2280) and bh → b give babdhaḥ (BM; TB read on the scan).
• bapsati.
• 1101 ki: ciketi.
• 1102 tura: tutorti, tutūrtaḥ (hali ca), tuturati.
• 1103 dhiṣa: didheṣṭi, didhiṣṭaḥ.
• 1104 dhana, "acquiring grain": dadhanti, dadhantaḥ. There is no lengthening before the nasal. BM grounds this on a parallel (cakṣaṃse), and TB says some treat it as Vedic (§16 neutral). dadhanati.
• 1105 jana janane: jajanti.

H. jan (and gā): ā for the final before a jhal-initial san or kṅit — running example: ja-jan + tas → ja-jā-tas → jajātaḥ; ja-jan + ati → jajn-ati → jajñati

• 2504 janasanakhanāṃ sañjhaloḥ 6.4.42 — for jan, san and khan, ā replaces the final before a jhal-initial san and before a jhal-initial kṅit.
• āt comes from viḍvanor anunāsikasyāt (SK 2982), and jhali kṅiti from anudāttopadeśa… (6.4.37, SK 2428). The carried jhal qualifies san, and the sūtra's jhal qualifies kṅit (BM). BM reports that the Bhāṣya rejects the mention of jhal.
• Why "jhal-initial san"? jijaniṣati, sisaniṣati, cikhaniṣati (BM). With a jhal-initial san: siṣāsati (TB).
• jajātaḥ.
• jajñati: ati is ṅit but not jhal-initial, so no ā. The penultimate is lost by gamahana… (SK 2363), and n → ñ by ścutva (BM, TB).
• jajaṃsi (anusvāra, SK 123); liṭ jajāna.
• jajanyāt / jajāyāt and āśīrliṅ janyāt / jāyāt: before y the ā is optional by ye vibhāṣā (6.4.43, SK 2319).
• 1106 gā stutau (Vedic):
• "devāñ jigāti sumnayuḥ": i in the abhyāsa by bahulaṃ chandasi.
• jigītaḥ: ī by 2497.
• jigati: at, then ā-loss by 2483.
• BM adds āśīrliṅ geyāt (Wikisource gāyāt) and luṅ agāsīt.
• Close: "iti juhotyādayaḥ" (Here end the juhotyādi roots). It is supplied from the print; BM and TB close alike. p47 opens with divādi, 1107 divu (SK 2505).