Covers SK 2423–2488 (66 sūtras): all present. SK 2433, 2436 and 2439 have no kaumudī of their own (in ours and in the print). 2433 is glossed jointly under 2434, 2436 is covered by the heading at the end of 2435 and the forms under 2437, and 2439 is glossed jointly under 2440. All three are placed below. The spine is not an Aṣṭādhyāyī run. The SK follows the Dhātupāṭha order of the adādi roots (1011 ad … 1082 hnuṅ) and teaches each rule at the root where it first bites, drawing on AS books 1, 2, 3, 6, 7 and 8 (e.g. 2.4.72 → 2.4.40 → 6.4.101 → 7.3.100 … 7.4.53). Every AS-ref below is checked against references/ashtadhyayi/sutraani-data.json.
- Anuvṛtti from later in the SK. Many rules here carry words from sūtras the SK teaches only later. Examples: ado jagdhir lyap ti kiti (2.4.36, SK 3080) at 2424 and 2427; aḍ gārgyagālavayoḥ (SK 2476) at 2426; śā hau (SK 2487) at 2431; rudādibhyaḥ sārvadhātuke (SK 2474) at 2439–2440; adabhyastāt (SK 2479) at 2442; nābhyastasyāci piti sārvadhātuke (7.3.87, SK 2503) at 2443–2452; vija iṭ (1.2.2, SK 2536) at 2447; vido laṭo vā (SK 2464) at 2450; iṅaś ca (2.4.48, SK 2616) at 2459; īd yati (6.4.65, SK 2843) at 2462; sārvadhātuke yak (3.1.67, SK 2756) at 2466; sipi dhāto rur vā (SK 2485) at 2468; ī haly aghoḥ (6.4.113, SK 2497) at 2482. The outline names them where the commentators do.
tiṅantādādiprakaraṇam — outline (p45, SK 2423–2488, all 66 sūtras)
p44 (tiṅantabhvādiprakaraṇam) closed at 2422 with ṛter īyaṅ (3.1.29).
- Where the prakaraṇa begins. It opens with adiprabhṛtibhyaḥ śapaḥ (2.4.72) and root 1011 ad "to eat". The rest of the class follows in Dhātupāṭha order. Open boundary question (for Ajai): the 1915 print (NSP p. 390, scan page-0400) already heads SK 2422 ṛter īyaṅ "tiṅante adādayaḥ 42" and ends it with "1011 ad bhakṣaṇe | dvau parasmaipadinau ‖", whereas ours keeps 2422 in p44 and opens sk2423 with that root entry (.work/sk-run/p45-notes.md).
- Where it ends. At 2488. After the vaś paradigm the gaṇasūtra carkarītaṃ ca puts the yaṅ-luk roots in this class, 1082 hnuṅ follows, and the kaumudī closes "ity adādayaḥ" ("Here end the adādi roots"). The print then heads "tiṅante juhotyādayaḥ 43" and gives 1083 hu. p46 opens with juhotyādibhyaḥ śluḥ (SK 2489).
- Contents in outline.
- 2423–2427: ad. luk of śap, ghasḷ in liṭ (optionally), luṅ and san, dhi for hi, aṭ before an apṛkta ending: atti, jaghāsa/āda, addhi, ādat, aghasat.
- 2428–2434: han. Loss of the nasal, optional ṇ, velar after the abhyāsa, ja before hi, and vadha under the ārdhadhātuke heading: hataḥ, prahaṇmi, jaghāna, jahi, vadhyāt, avadhīt. The heading introduces dviṣ.
- 2435–2438: dviṣ; duh, dih, lih; cakṣiṅ. jus after dviṣ, the h-final sandhi (dogdhi, leḍhi), and cakṣ → khyāñ/khśāñ with aṅ: cakhyau/cakśau/cacakṣe, akhyat. The heading introduces īra and īḍ.
- 2439–2442: iṭ before se and dhve; ās … ṣūṅ; śīṅ. īḍiṣe, īḍidhve, then ruṭ and guṇa for śīṅ: śete, śerate.
- 2443–2449: the u-final roots and ūrṇu. vṛddhi in luk, optional īṭ, and the ūrṇu rules: yauti, ravīti/rauti, ūrṇauti/ūrṇoti, ūrṇunāva, aurṇot, aurṇāvīt; stauti/stavīti.
- 2450–2454: brū. āha, āttha, īṭ (bravīti), vac in the ārdhadhātuka (uvāca), um (avocat).
- 2455–2462: iṇ, iṅ, ik, vī. yaṇ, the long abhyāsa, the short vowel after an upasarga, gā, gāṅ, the ṅit affix after gāṅ, and ī for ā: yanti, īyatuḥ, niriyāt, agāt, adhijage, adhyagīṣṭa.
- 2463: jus after the ā-final roots. ayuḥ/ayān, with the roots yā … vac and the root entry for vid.
- 2464–2468: vid. ṇal in laṭ (veda), vidāṅkurvantu, u after kṛ, u for its a (vidāṅkurutāt, vidāṅkuru), ru for d (aveḥ).
- 2469–2472: as. Loss of a, bhū in the ārdhadhātuka, e before hi (edhi), ṣ after an upasarga (niṣyāt).
- 2473–2479: mṛj, and the five rudādi. vṛddhi of mṛj (mārṣṭi); iṭ, īṭ, aṭ (roditi, arodīt/arodat); ṣ in suṣuṣupatuḥ; ṇ in prāṇiti; at after an abhyasta (jakṣati).
- 2480–2483: jāgṛ and daridrā. guṇa (jajāgaratuḥ, ajāgaruḥ), i or loss of ā (daridritaḥ, daridrati).
- 2484–2488: cakās, śās, the five Vedic roots, hnuṅ. d/ru for a pada-final s (acakāt, acakāḥ), i and śā for śās (śiṣṭaḥ, śādhi), loss of the final of dīdhī/vevī (dīdhitā).
Threads that run through the prakaraṇa:
- luk of śap is the class-mark. adiprabhṛtibhyaḥ śapaḥ (2423) removes the vikaraṇa, so the ending meets the root directly. Much of the prakaraṇa is therefore sandhi between a root-final consonant and a consonant-initial ending, done by the tripādī: dogdhi, leḍhi, dveṣṭi, īṭṭe, mārṣṭi, vaṣṭi, āttha. The gaṇasūtra carkarītaṃ ca (in 2488) brings the yaṅ-luk roots under the same luk (TB).
- The ārdhadhātuke heading (2.4.35, SK 2432) governs the substitutes vadha (2433–2434), khyāñ (2436–2437), vac (2453), gā/gāṅ (2458–2460) and bhū (2470). In AS order ghasḷ (2.4.37, 2.4.40; SK 2427, 2424) also follows it. The kaumudī re-invokes it at the end of 2435, 2452 and 2469. At 2434 Bhaṭṭoji reads it as a viṣayasaptamī: the substitute comes as soon as an ārdhadhātuka is intended (BM). The sārvadhātuka liṅ ("vidhyādi", liṅ in the senses of injunction etc.) therefore keeps the original root: hanyāt, brūyāt, syāt.
- ābhīya asiddhatva (asiddhavad atrābhāt, 6.4.22, SK 2183) decides one form after another:
- jahi (2431): ja does not count, so no luk of hi;
- āyan (2456): the yaṇ does not count, so āṭ;
- vidāṅkuru (2467): the luk of hi does not count, so u still comes;
- edhi, āstām (2471): the e and the a-loss do not count, so dhi and āṭ;
- śādhi (2487): śā does not count, so dhi.
- At 2488 (sadhi) the same argument runs on pūrvatrāsiddham (8.2.1, SK 12).
- "ṅic ca pin na" (Bhāṣya: what is ṅit is not pit) blocks vṛddhi or īṭ before the ṅit yāsuṭ and tātaṅ: yuyāt (2443), ūrṇuyāt (2448), brūtāt (2453). The specially enjoined ṅit of yāsuṭ still bars guṇa (BM).
- The abhyasta roots of jakṣityādayaḥ ṣaṭ (6.1.6, SK 428) are jakṣ, jāgṛ, daridrā, cakās, śās, dīdhī and vevī. They take at for jh (adabhyastāt, 2479: jakṣati, jāgrati, daridrati, cakāsati, śāsati) and jus in laṅ (sijabhyastavidibhyaś ca, 3.4.109, SK 2226: ajakṣuḥ, ajāgaruḥ, adaridruḥ, aśāsuḥ). vid shares the jus (aviduḥ).
- The kaumudī lists forms bare. Lakāra and person labels here are the translator's, given in Pāṇinian terms (prathama/madhyama/uttama-puruṣa; vidhiliṅ, āśīrliṅ). 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 printed under a neighbouring unit: 2429–2438 in sk.2428.t1; 2441 in sk.2440.t1; 2451 in sk.2450.t1; 2454–2456 in sk.2452.t1; 2465 in sk.2464.t1; 2466 in sk.2465.t1; 2477 in sk.2476.t1.
- No TB is printed on 2423–2424, 2432, 2453, 2458, 2460, 2463's yā forms, 2470 and 2485.
- BM missing from our source and read from Wikisource: 2434, 2452, 2459, 2460, 2467 (fuller anuvṛtti), 2468, 2471. BM for 2477 is inside sk.2476.b1.
- Misfiled BM blocks: sk.2452.b1 is BM on SK 2316, sk.2460.b1 on SK 1152, sk.2478.b1 on SK 2984, and sk.2484.b1 on SK 433. No BM on 2478 or 2484 exists.
Where the p45 audit (.work/sk-run/p45-audit-fixes.md and the seven reports in .work/sk-run/p45-audit/) corrects a reading, this outline follows the corrected reading. The fixers are applying these to the en.yml notes in parallel.
- 2428 (must-fix): no ṣaṇu. The nasal-final tanādi are eight: tanu, kṣaṇu, kṣiṇu, ṛṇu, tṛṇu, ghṛṇu, vanu, manu. This follows our text, BM ("ete 'ṣṭau"; sanoti is excluded because ā is taught for it) and TB (pratīka tanukṣaṇukṣiṇv-, "na sa parigaṇitaḥ"). The print's ṣaṇu goes to the Print Check.
- 2435 (must-fix): the dogdhi chain. guṇa comes first, then the tripādī: duh-ti → doh-ti → dogh-ti → dogh-dhi → dogdhi.
- 2459 (must-fix): iṅaḥ is carried forward from iṅaś ca (SK 2616). BM (Wikisource uttara-1/0210) says so, and gives the same for 2460.
- 2434: "vidhyādi" means liṅ in the senses of injunction etc., not the "injunctive". The number 2300 cited for 'ārdhadhātuke' (ours and print) should point to the heading SK 2432 (BM: "ārdhadhātuke vivakṣite … prāg eva vadhādeśe").
- 2435: 'ante' is carried into idito num dhātoḥ (SK 2262) from goḥ pādānte (BM, TB). sphāyan is the śatṛ of the anudāttet sphāyī vṛddhau.
- 2438: samacakṣiṣṭa means "he avoided" (luṅ; BM, Wikisource).
- 2440: aiḍḍhvam has no iṭ because the sūtra's dhve is itself a modification of dhvam, so it does not take in laṅ's dhvam.
- 2442, 2454: "stautyantāḥ" means the parasmaipadins up to ṣṭuñ (exclusive; BM), and "śāsyantāḥ" means the roots up to śāsu (1075).
- 2466: read vidāṅkarotu as one word. Wikisource BM reads viśvaset as "not established" (asiddham), where our OCR has "beside the point".
- 2467: BM (Wikisource) also carries kṛ from nityaṃ karoteḥ (SK 2548) and kṅiti from gamahana… (SK 2363).
- 2472: TB's example for "parasya" is prādur asti; prādur āsīt is BM's.
- 2473: the laṅ form is amārṭ in our English, but the print may read amārḍ (BM pratīka "amārḍ iti"). Both are valid.
- 2481: in the verse, "ataḥ" means "thus": "…vṛddhi again, prohibition: thus, with yaṇ first, the applicabilities are nine".
- 2484: the shared anuvṛttis are named by BM on 2485: padasya (adhikāra), ante (SK 84), s (SK 162), and daḥ running on from vasusraṃsu… (SK 334).
- 2488: the print has no dhātu number 1081 (the gaṇasūtra is marked "(ga)"); ours supplies it.
Print readings (NSP, scan-confirmed) that the outline follows:
- addhi (2425; ours addhiḥ);
- dugdhi, līḍhi, alikṣat and alihvahi (2435);
- khśādi and asiddhakāṇḍe (2437);
- anidit, pṛṅkte, dayāyāsaś ca 2324 (2440);
- ruvīyāt (2444), ūrṇuvāte (2445), aurṇāviṣṭām and paryastaut (2449);
- brūtāt (2453), vivakṣite (2459), adhipūrvaḥ (2462);
- upasargeṇaḥ … yakāre (2472), svap (2474), aruditām (2476), ṣatvasyāsiddhatvāt (2477);
- antasthādiḥ and nidrākṣaye (2479), daridritaḥ (2482), ṇvuli and adaridrīt (2483), ruḥ (2485), śāsatu (2486), avāśīt/avaśīt (2488).
Still open for Ajai (mūla):
- the 2422/2423 boundary;
- the print's ṣaṇu (2428);
- īritā and īriṣyate, which are in ours and Wikisource but not in the print (2438; BM's pratīka does show īritā);
- the cross-reference numbers 2300 (2434) and 2377 (2435), both for SK 2432;
- aurṇavīt against BM's aurṇuvīt (2448);
- the extra śiṣṭām (2486);
- the repeated uṣṭaḥ after vakṣi, where uṣṭhaḥ is expected (2488).
Ordering: SK order throughout.
A. ad: luk of śap, ghasḷ, dhi, aṭ — running example: ad + śap + ti → ad + ti → atti
- 2423 adiprabhṛtibhyaḥ śapaḥ 2.4.72 — after ad and the roots that follow it in the Dhātupāṭha, luk of śap (luk from ṇyakṣatriyārṣañito yūni luk aṇiñoḥ 2.4.58, SK 1276; BM).
- Root entry: 1011 ad bhakṣaṇe; "dvau parasmaipadinau" means ad and han (1012). ad is aniṭ (BM).
ad + śap + ti → ad + ti (luk) → atti (the d becomes t by cartva). Likewise attaḥ.adanti: this is why the substitute of jho 'ntaḥ (SK 2169) begins with a (BM).- 2424 liṭy anyatarasyām 2.4.40 — before liṭ, ad is optionally replaced by ghasḷ (adaḥ from SK 3080, ghasḷ from 2427; BM).
ghas + ṇal → jaghāsa. ṇal is not kit, so there is no penultimate loss.jaghas + atus → jaghs atus: atus is kit (asaṃyogāl liṭ kit, SK 2242), so the penultimate a is lost by gamahana… (SK 2363; the digital source misprints 1263). Sthānivadbhāva is barred for a car-injunction, so gh → k. The s becomes ṣ by śāsivasighasīnāṃ ca (SK 2410): jakṣatuḥ, jakṣuḥ.- thal: ghas never occurs before tās, so iṭ is obligatory:
jaghasitha. - Without the substitute:
āda, ādatuḥ, and āditha (iṭ obligatory by iḍ atty-arti-vyayatīnām, SK 2384). luṭ attā, lṛṭ atsyati. - 2425 hujhalbhyo her dhiḥ 6.4.101 — after hu and after a jhal-final aṅga, hi → dhi (dhi has more than one sound, so it replaces the whole of hi).
ad + hi → addhi (print).- With tātaṅ, which is the later rule:
attāt. dhi does not come again, because what is set aside in a conflict stays set aside (BM; TB "sakṛdgatau"). adāni (āṭ by āḍ uttamasya pic ca, SK 2204).- Why no dhi in rudihi? BM says iṭ intervenes. TB reports two views: carry 'hali' forward, or read dh for h. §16 neutral.
- 2426 adaḥ sarveṣām 7.3.100 — after ad, the augment aṭ comes to an apṛkta sārvadhātuka, "in the view of all". aṭ is from 2476, apṛkte from 2448 and sārvadhātuke from 2444; "sarveṣām" stops gārgyagālavayoḥ (BM, TB).
- laṅ:
ād + t → ād + a + t → ādat, ādaḥ, ādam. - No aṭ where the ending is not apṛkta: āttām, ādan, ātta, ādva, ādma.
- vidhiliṅ
adyāt: the stem is not a-final after luk, so ato yeyaḥ (SK 2212) does not apply. āśīrliṅ adyāstām, adyāsuḥ. - 2427 luṅsanor ghasḷ 2.4.37 — in luṅ and before san, ad → ghasḷ.
- It is ḷdit, so aṅ replaces cli (puṣādidyutādyḷditaḥ parasmaipadeṣu, SK 2343):
aghasat. TB gives the san form: jighatsati. - Root entry 1012 han hiṃsāgatyoḥ:
praṇihanti (ṇ of ni by nergada…, SK 2285).
B. han: loss of the nasal, ṇ, the velar, ja and vadha — running example: han + tas → hataḥ; han + hi → ja + hi → jahi; vadha + yās + t → vadhyāt
- 2428 anudāttopadeśavanatitanotyādīnām anunāsikalopo jhali kṅiti 6.4.37 — "anunāsika" is luptaṣaṣṭhīka and qualifies the anudāttopadeśa and tanādi roots, not vanati, which is always nasal-final (BM, TB). The final of these roots (alo 'ntyasya) is lost before a jhal-initial kit/ṅit.
- The nasal-final anudāttopadeśa roots are yam, ram, nam, gam, han and man (manyati).
- The nasal-final tanādi are eight: tanu, kṣaṇu, kṣiṇu, ṛṇu, tṛṇu, ghṛṇu, vanu, manu. sanoti is excluded (see the audit note above).
han + tas (ṅit by sārvadhātukam apit) → ha + tas → hataḥ.ghnanti: jhi is not jhal-initial. Instead the penultimate is lost by gamahana… (SK 2363) and h → gh by ho hanteḥ (SK 358) (BM).- 2429 vamor vā 8.4.23 — the n of han, after an r/ṣ standing in an upasarga, optionally → ṇ before v or m. upasargāt comes from SK 2287, raṣābhyām… from SK 235 and hanteḥ from SK 359 (BM).
prahaṇmi / prahanmi, prahaṇvaḥ / prahanvaḥ.- liṭ
jaghāna, jaghnatuḥ, jaghnuḥ. The velar comes by ho hanter ñṇinneṣu (7.3.54, SK 358), which is antaraṅga compared with 2430 (BM, TB). - 2430 abhyāsāc ca 7.3.55 — the h of han after an abhyāsa → velar (ho hanteḥ from SK 358, ku from cajoḥ ku ghiṇṇyatoḥ, SK 2863; BM).
- thal
jaghanitha / jaghantha: here ho hanteḥ cannot apply, since the ending is not ñit/ṇit and no n follows the h (BM). - hantā;
haniṣyati (iṭ by ṛddhanoḥ sye, SK 2366); hantu, hatāt, ghnantu. - 2431 hanter jaḥ 6.4.36 — han → ja before hi (hau from śā hau, SK 2487).
- ja is ābhīya and so asiddha, which means no luk of hi by ato heḥ (SK 2202):
jahi. - hanāni: āṭ is pit, so there is no penultimate loss.
- laṅ ahan, ahatām, aghnan, ahanam.
- 2432 ārdhadhātuke 2.4.35 — adhikāra ("ity adhikṛtya"): the substitutes that follow are taught for an ārdhadhātuka. BM supplies "ado jagdhiḥ and the rest will be stated". See the threads for its scope.
- 2433 hano vadha liṅi 2.4.42 — no kaumudī (ours and print). han → vadha in the ārdhadhātuka liṅ. "vadha" is luptaprathamāka (BM). Its forms are given under 2434.
- 2434 luṅi ca 2.4.43 — also in luṅ. The substitute vadha is a-final.
- 'ārdhadhātuke' is a viṣayasaptamī, so vadha stands before the ārdhadhātuka is added and is a-final at its upadeśa. ato lopaḥ (6.4.48, SK 2308) then removes the a:
vadha + yās + t → vadhyāt, vadhyāstām (āśīrliṅ, the forms of 2433). - The kaumudī cites 'ārdhadhātuke' as 2300; see the audit note.
- In the vidhiliṅ etc. (sārvadhātuka) there is no vadha:
hanyāt, and prahaṇyāt (ṇ by hanter atpūrvasya, SK 359). - luṅ
avadhīt. vadha is polysyllabic at its upadeśa, so sic takes iṭ. The a-loss counts as the original (sthānivat), so there is no vṛddhi by ato halādeḥ laghoḥ (SK 2284) (BM, TB). - Then "atha catvāraḥ svaritetaḥ": 1013 dviṣ aprītau, ubhayapadin and aniṭ.
- laṭ
dveṣṭi (guṇa before pit tip, ṣṭutva) / dviṣṭe; luṭ dveṣṭā; lṛṭ dvekṣyati (ṣ → k before s), dvekṣyate. - loṭ dveṣṭu, dviṣṭāt,
dviḍḍhi (dhi, then jaśtva), dveṣāṇi, dveṣai, dveṣāvahai. - laṅ
adveṭ.
C. dviṣ, the h-final roots duh/dih/lih, and cakṣiṅ → khyāñ — running example: duh + ti → doh-ti → dogh-ti → dogh-dhi → dogdhi; cakṣ + liṭ → cakhyau / cakśau / cacakṣe
- 2435 dviṣaś ca 3.4.112 — after dviṣ, jus optionally replaces laṅ's jhi. jher jus comes from SK 2213 and laṅaḥ śākaṭāyanasya from SK 2463 (BM).
adviṣuḥ / adviṣan, adveṣam; dviṣīta.dvikṣīṣṭa: kit by liṅsicāv ātmanepadeṣu (SK 2300), so no guṇa.advikṣat: ksa by śala igupadhād aniṭaḥ ksaḥ (SK 2336).- 1014 duha prapūraṇe ("drawing out", i.e. milking: BM pūraṇābhāva, TB tyājana).
duh-ti → doh-ti (guṇa, tip being pit) → dogh-ti (dāder dhātor ghaḥ 8.2.32, SK 325) → dogh-dhi (jhaṣas tathor dho 'dhaḥ 8.2.40, SK 2280) → dogdhi (jaśtva). dugdhaḥ.dhokṣi: h → gh, d → dh (bhaṣ), gh → k, ṣ.- dugdhe, dhukṣe, dhugdhve; dogdhu.
dugdhi (print): hi → dhi by 2425. bhaṣ applies only before s and dhv.- dohāni, dhukṣva, dhugdhvam, dohai; laṅ
adhok, adoham, adhugdhvam. - luṅ adhukṣat, adhukṣata. With luk of ksa by lugvā duhadihaliha… (7.3.73, SK 2365) before ta, thās, dhvam, vahi, the forms are like laṅ: adugdha, adugdhāḥ, adhugdhvam, aduhvahi (BM, TB).
- 1015 diha upacaye (upacaya = growth):
praṇidegdhi (ṇ by SK 2285). - 1016 liha āsvādane.
leḍhi: h → ḍh (ho ḍhaḥ, SK 324), t → dh (SK 2280), ṣṭutva, then the first ḍh is lost (ḍho ḍhe lopaḥ, SK 2335). līḍhaḥ, lihanti.lekṣi: ḍh → k by ṣaḍhoḥ kaḥ si, SK 295.- līḍhe, likṣe, līḍhve; leḍhu, līḍhi (print), lehāni.
- laṅ aleṭ; luṅ alikṣat (print), alikṣata, alīḍha (luk of ksa), alikṣāvahi / alihvahi (print).
- 1017 cakṣiṅ vyaktāyāṃ vāci, also "seeing".
- Its i is an anudātta it, there for yuc (anudāttetaś ca halādeḥ, SK 3129): vicakṣaṇa.
- There is no num, because idito num dhātoḥ (SK 2262) takes 'ante' from goḥ pādānte (BM, TB).
- The ṅ is a jñāpaka that "ātmanepada caused by an anudātta it is not invariable". Hence sphāyan (śatṛ from the anudāttet sphāyī). BM demurs: the ṅ already has a use, so treat sphāyan as pṛṣodarādi.
caṣṭe: k lost by skoḥ saṃyogādyor ante ca (SK 380), then ṣṭutva. cakṣāte.- The block closes: "with 'ārdhadhātuke' as heading" (cited as 2377 in ours and the print; the heading is SK 2432).
- 2436 cakṣiṅaḥ khyāñ 2.4.54 — no kaumudī (ours and print). cakṣiṅ → khyāñ before an ārdhadhātuka (BM). The Bhāṣya's point that the substitute is really khśāñ is given under 2437.
- 2437 vā liṭi 2.4.55 — in liṭ the substitution is optional.
- In the Bhāṣya the substitute is khś-initial (khśādi, print). In the asiddha section, after the ṇatva section (TB), comes the vārttika śasya yo vā: y optionally replaces its ś. BM and TB draw consequences for puṃkhyāna, paryākhyāna and sauprakhyīya.
- It is ñit, so both padas:
cakhyau, cakhye (y) / cakśau, cakśe (ś). The vārttika cayo dvitīyāḥ… does not turn the k into kh, because cartva is asiddha. cacakṣe (no substitute).- Outside liṭ the substitute is obligatory: khyātā / kśātā; khyāsyati, -te / kśāsyati, -te; āśīrliṅ khyāyāt / khyeyāt, kśāyāt / kśeyāt.
- The sārvadhātuka forms are of cakṣ itself: acaṣṭa, cakṣīta.
- 2438 asyativaktikhyātibhyo 'ṅ 3.1.52 — aṅ replaces cli (cleḥ from SK 2222) after as "throw", vac (brū's substitute and the root vac; TB) and khyā.
- The y-form takes aṅ:
akhyat, akhyata. - The ś-form does not, because the rule names "khyāti":
akśāsīt, akśāsta (BM). - vārttika varjane khśāñ neṣṭaḥ:
samacakṣiṣṭa ("he avoided"). - Then the anudāttet roots up to pṛc:
- 1018 īra: īrte, īrāñcakre, īrtām, īrṣva, īrdhvam, airiṣṭa. Ours adds īritā and īriṣyate, which the print lacks.
- 1019 īḍa:
īṭṭe (ṣṭutva, then cartva).
D. iṭ before se and dhve; ās … ṣūṅ; guṇa and ruṭ for śīṅ — running example: īḍ + se → īḍ-i-se → īḍiṣe; śī + ate → śe + r + ate → śerate
- 2439 īśaḥ se 7.2.77 — no kaumudī (ours and print). After īś, iṭ (from iḍ atty-arti…, SK 2384) comes to the sārvadhātuka se (sārvadhātuke from SK 2474; BM). It is glossed jointly with 2440.
- 2440 īḍajanor dhve ca 7.2.78 — joint gloss: after īś, īḍ and jan, iṭ comes to the sārvadhātuka se and dhve. The rule is split "for variety": se runs forward and dhve is drawn back (BM, TB).
īḍiṣe, īḍidhve.- A thing modified in one part is not another thing (ekadeśavikṛtam ananyavat), so the loṭ
īḍiṣva, īḍidhvam also take iṭ. - But mentioning the modified form does not include the original. The sūtra's dhve is itself dhvam modified, so the laṅ dhvam gets no iṭ:
aiḍḍhvam. - The roots 1020–1032:
- 1020 īś: īṣṭe, īśiṣe, īśidhve.
- 1021 ās: āste; āsāṃcakre (ām by dayāyāsaś ca 3.1.37, SK 2324, print number); āssva; ādhvam (s lost by dhi ca; BM); āsiṣṭa.
- 1022 āṅaḥ śāsu: āśāste. The "āṅ" is only as a rule (prāyika), so "namovākaṃ praśāsmahe" stands.
- 1023 vas "cover": vaste, vasse, vadhve; vavase (v-initial, so no e and no loss of the abhyāsa); vasitā.
- 1024 kasi: kaṃste, kaṃsāte, kaṃsate (num, being idit). Some read it anidit (print): kaste. Read with the palatal ś and also anidit: kaṣṭe, kaśāte, kakṣe, kaḍḍhve.
- 1025 ṇisi: niṃste. It ends in the dental s; the Ābharaṇa's palatal reading is an error.
- 1026 ṇiji: niṅkte, niṅkṣe, niñjitā (seṭ).
- 1027 śiji: śiṅkte.
- 1028 piji: piṅkte. Some read pṛji: pṛṅkte (print).
- 1029 vṛjī: ī-marked, so no num: vṛkte, vṛjāte, vṛkṣe. Others read it i-marked: vṛṅkte.
- 1030 pṛcī: pṛkte.
- 1031 ṣūṅ: sūte; suṣuve; suṣuviṣe (iṭ invariable by the krādi restriction; BM, TB); sotā / savitā; saviṣīṣṭa / soṣīṣṭa; asaviṣṭa / asoṣṭa.
suvai: guṇa is barred by bhūsuvos tiṅi (SK 2224).- 1032 śīṅ svapne.
- 2441 śīṅaḥ sārvadhātuke guṇaḥ 7.4.21 — guṇa for śīṅ before a sārvadhātuka. It is an apavāda of kṅiti ca (SK 2217), so guṇa comes even before a kṅit ending:
śete, śī + āte → śe + āte → śayāte. - 2442 śīṅo ruṭ 7.1.6 — after śīṅ, the at that replaces jha takes the augment ruṭ. jhaḥ comes from SK 2169, and at from SK 2479, read as a sixth (BM, TB).
śe + ate → śe + r + ate → śerate. śeṣe; śedhve (no ḍh); śaye, śevahe.- śiśye (yaṇ by er anekācaḥ), śayitā, aśayiṣṭa.
- Then "atha stautyantāḥ parasmaipadinaḥ": the parasmaipadins up to ṣṭuñ (1043), exclusive. ūrṇu takes both padas. 1033 yu miśraṇe 'miśraṇe ca.
E. The u-final roots and ūrṇu: vṛddhi, īṭ, guṇa, the doubling nu — running example: yu + ti → yauti; ru + ti → ravīti / rauti; ūrṇu + ti → ūrṇauti / ūrṇoti
- 2443 uto vṛddhir luki hali 7.3.89 — in the domain of luk (luki is a viṣayasaptamī), u → vṛddhi before a hal-initial pit sārvadhātuka, but not in an abhyasta. The rest comes from nābhyastasyāci piti sārvadhātuke (SK 2503), without aci (BM, TB).
yu + ti → yauti.- yutaḥ has no vṛddhi, since the ending is not pit. yuvanti takes uvaṅ.
- liṭ yuyāva; luṭ yavitā.
- vidhiliṅ
yuyāt: no vṛddhi, by the Bhāṣya's "ṅic ca pin na". The specially enjoined ṅit of yāsuṭ prevails over its pit (BM). - āśīrliṅ
yūyāt: ārdhadhātuka, lengthened by akṛtsārvadhātukayor dīrghaḥ (SK 2298). - luṅ ayāvīt. Root entry 1034 ru.
- 2444 turustuśamyamaḥ sārvadhātuke 7.3.95 — after tu, ru, stu, śam, am, a hal-initial sārvadhātuka tiṅ optionally takes īṭ. hali comes from SK 2443, tiṅi from SK 2224, īṭ from SK 2452 and vā from yaṅo vā (SK 2651) (BM).
- "sārvadhātuke" is repeated so that 'piti' is not carried forward (apidartham).
ravīti / rauti, ruvītaḥ / rutaḥ.- No īṭ in these: ruvanti (not hal-initial), śāmyati (śyan is not a tiṅ), rūyāt (āśīrliṅ, ārdhadhātuka).
- vidhiliṅ ruyāt / ruvīyāt (print); arāvīt, araviṣyat.
- tu is a sautra root "remembered" as lugvikaraṇa: tavīti / tauti, tuvītaḥ / tutaḥ, totā, toṣyati. TB and BM say śam and am take the rule only in the Veda.
- The roots 1035–1039:
- 1035 ṇu: nauti, navitā.
- 1036 ṭukṣu: kṣauti.
- 1037 kṣṇu: kṣṇauti.
- 1038 ṣṇu: snauti, suṣṇāva, snavitā, snuyāt, snūyāt.
- 1039 ūrṇuñ ācchādane.
- 2445 ūrṇoter vibhāṣā 7.3.90 — for ūrṇu, the vṛddhi of 2443 is optional before a hal-initial pit sārvadhātuka.
ūrṇauti / ūrṇoti, ūrṇutaḥ, ūrṇuvanti.- ātmanepada ūrṇute, ūrṇuvāte (print), ūrṇuvate.
- vārttika ūrṇoter ām neti vācyam: no ām in liṭ. Declared nu-like, ūrṇu neither begins with an ic nor is anekāc (BM).
- 2446 na ndrāḥ saṃyogādayaḥ 6.1.3 — an n, d or r after a vowel that heads a consonant cluster is not reduplicated (dve from SK 2175, ajādeḥ from SK 2176).
- In ūrṇu the ṇ is asiddha for doubling, so the syllable nu is doubled.
- The paribhāṣā pūrvatrāsiddhīyam advirvacane is anitya: ubhau sābhyāsasya (SK 2606) is the liṅga.
ūrṇu + a → ūr-ṇu-nu-a → ūrṇunāva, ūrṇunuvatuḥ, ūrṇunuvuḥ.- 2447 vibhāṣorṇoḥ 1.2.3 — after ūrṇu, an iṭ-initial affix is optionally ṅit. iṭ comes from vija iṭ (SK 2536) and ṅit from SK 2461 (BM, TB).
ūrṇunuvitha (ṅit, no guṇa, uvaṅ) / ūrṇunavitha (guṇa); ūrṇuvitā / ūrṇavitā.- loṭ ūrṇautu / ūrṇotu (by 2445).
- ūrṇavāni, ūrṇavai: āṭ is pit and vowel-initial, so neither rule reaches them.
- 2448 guṇo 'pṛkte 7.3.91 — ūrṇu takes guṇa before an apṛkta hal-initial pit sārvadhātuka. It is an apavāda of 2445's vṛddhi (BM).
aurṇu + t → aurṇot, aurṇoḥ.- ūrṇuyāt, ūrṇuyāḥ: no vṛddhi (ṅic ca pin na).
- ūrṇūyāt; ūrṇaviṣīṣṭa / ūrṇuviṣīṣṭa.
- luṅ aurṇavīt (ours and the print; BM's pratīka reads aurṇuvīt), aurṇuviṣṭām.
- 2449 ūrṇoter vibhāṣā 7.2.6 — for ūrṇu, vṛddhi is optional before an iṭ-initial sic in parasmaipada. sici vṛddhiḥ comes from SK 2297, and iṭi from neṭi (SK 2268).
aurṇāvīt, aurṇāviṣṭām (print), aurṇāviṣuḥ; with guṇa aurṇavīt.- The roots 1040–1044:
- 1040 dyu: dyauti, dyotā.
- 1041 ṣu: sotā, asauṣīt.
- 1042 ku: kotā.
- 1043 ṣṭuñ:
stauti / stavīti, stutaḥ / stuvītaḥ, stute / stuvīte. astāvīt: iṭ by stusudhūñbhyaḥ parasmaipadeṣu, SK 2385.abhyaṣṭaut: ṣ across aṭ by prāk sitād aḍvyavāye 'pi, SK 2276.paryaṣṭaut / paryastaut (print): optional by sivādīnāṃ vā…, SK 2359.- 1044 brūñ vyaktāyāṃ vāci.
F. brū: āha, āttha, īṭ, vac, um — running example: brū + tip → āh + ṇal → āha; brū + sip → āh + thal → āth + tha → āttha; brū + ti → brū + ī + ti → bravīti
- 2450 bruvaḥ pañcānām ādita āho bruvaḥ 3.4.84 — after brū, the first five parasmaipada endings of laṭ (tip, tas, jhi, sip, thas) are optionally replaced by ṇal, atus, us, thal, athus, and brū by āha.
- "laṭo vā" comes from 2464 (BM). The first bruvaḥ is a fifth case, the second a sixth (TB).
- The final a of āha is for pronunciation, so the effective substitute is āh (BM on 2451: an a-final āha would give *āṭttha).
āha, āhatuḥ, āhuḥ.- 2451 āhas thaḥ 8.2.35 — the final of āh → th before a jhal (from jhalo jhali, SK 2281).
āh + tha → āth + tha → āttha: cartva of the first th (khari ca; BM).- āhathuḥ: athus is not jhal-initial.
- TB: "āhas taḥ" would have been more economical.
- 2452 bruva īṭ 7.3.93 — after brū, a hal-initial pit sārvadhātuka takes īṭ. piti (and sārvadhātuke) come from SK 2503, hali from SK 2443 (BM, Wikisource).
- No īṭ in āttha. āh would bring it in by sthānivadbhāva, but th is taught for a following jhal, and īṭ would destroy the jhal.
bravīti, brūtaḥ, bruvanti; ātmanepada brūte. āha is only parasmaipada (BM).- Then 'ārdhadhātuke' as heading.
- 2453 bruvo vaciḥ 2.4.53 — brū → vac before an ārdhadhātuka.
uvāca: saṃprasāraṇa of the abhyāsa by SK 2408.ūcatuḥ: saṃprasāraṇa by vacisvapiyajādīnāṃ kiti (SK 2409) before doubling. ūcuḥ.- uvacitha / uvaktha (no īṭ, since the ending is not sārvadhātuka); ūce; vaktā.
- The sārvadhātuka forms are of brū: bravītu,
brūtāt (print). tātaṅ is not pit (ṅic ca pin na), so no īṭ. - bravāṇi, bravai; brūyāt.
- āśīrliṅ ucyāt; luṅ aṅ by 2438.
- 2454 vaca um 7.4.20 — vac takes the augment um before aṅ (aṅi from ṛdṛśo 'ṅi guṇaḥ, SK 2406).
a-vac-at → a-va-u-c-at → avocat: um comes after the last vowel (midaco 'ntyāt paraḥ, SK 37), then ād guṇaḥ (SK 69). avocata.- Heading: the roots up to śāsu (1075) are parasmaipadin, but iṅ is ātmanepadin.
- 1045 iṇ gatau: eti (guṇa before pit), itaḥ.
G. iṇ, iṅ, ik, vī: yaṇ, the long abhyāsa, the short vowel, gā/gāṅ, ṅit, ī for ā — running example: i + anti → y + anti → yanti; i i atus → ī yatus → īyatuḥ; iṇ + luṅ → a-gā-t → agāt; adhi + iṅ + liṭ → adhi-gāṅ → adhijage
- 2455 iṇo yaṇ 6.4.81 — yaṇ for iṇ before a vowel-initial affix (aci from SK 271, tadādi). It is an apavāda of iyaṅ, though the later guṇa/vṛddhi still prevail: ayanam, āyakaḥ (BM, TB).
yanti.iyāya: vṛddhi of the second part, then iyaṅ for the abhyāsa by abhyāsasyāsavarṇe (SK 2290).- 2456 dīrgha iṇaḥ kiti 7.4.69 — the abhyāsa of iṇ is lengthened before a kit liṭ. abhyāsasya comes from SK 2620, liṭi from SK 2353 (BM, TB).
i i atus → i y atus (2455) → īyatuḥ, īyuḥ.- iyayitha / iyetha; etā; itāt, ihi, ayāni.
- laṅ ait, aitām,
āyan: the yaṇ is ābhīya and asiddha, so āṭ comes (āḍ ajādīnām, SK 2254). - iyāt / āśīrliṅ īyāt.
- 2457 eter liṅi 7.4.24 — the aṇ of iṇ after an upasarga is shortened before a kit liṅ. upasargād hrasvaḥ comes from SK 2702, aṇaḥ from ke 'ṇaḥ (SK 834), and kiti from SK 2649 (BM).
nis + īyāt → niriyāt.abhīyāt: a single substitute used on both sides at once is not treated as end-and-beginning (na antādivat).sameyāt: e is not aṇ. samīyāt is the bhvādi root i.- Bhaṭṭoji's gloss adds "ārdhadhātuka". BM calls it a slip; TB says it only describes the liṅ. §16 neutral.
- 2458 iṇo gā luṅi 2.4.45 — iṇ → gā in luṅ. sic takes luk by gātisthāghupābhūbhyaḥ… (SK 2223):
agāt, agātām, aguḥ. - 1046 iṅ adhyayane. The Dhātupāṭha note "nityam adhipūrvaḥ" says it is always used with adhi (BM).
adhīte, adhīyāte, adhīyate.- 2459 gāṅ liṭi 2.4.49 — for iṅ (from iṅaś ca, SK 2616; BM, Wikisource), gāṅ comes in liṭ. It comes either at the stage of the la (the vārttika view) or as soon as liṭ is intended (the Bhāṣya view; print "vivakṣite") (TB, BM).
adhijage, adhijagāte, adhijagire. luṭ adhyetā, lṛṭ adhyeṣyate.- loṭ
adhyayai: guṇa and ay come first, then yaṇ of adhi. Savarṇa-lengthening, which would otherwise be antaraṅga, is blocked by the wording ṇer adhyayane vṛttam (SK 3066). - laṅ adhyaita;
adhyaiyātām (iyaṅ comes before āṭ because it is the later rule; BM, Wikisource), adhyaiyata, adhyaiyi, adhyaivahi. - adhīyīta …; āśīrliṅ adhyeṣīṣṭa.
- 2460 vibhāṣā luṅlṛṅoḥ 2.4.50 — in luṅ and lṛṅ, iṅ → gāṅ optionally. iṅaḥ and gāṅ are carried forward from SK 2616 and 2459 (BM, Wikisource). The forms are given under 2462.
- 2461 gāṅkuṭādibhyo 'ñṇin ṅit 1.2.1 — after the substitute gāṅ and after kuṭ and the rest (a tudādi subgroup; TB), an affix that is not ñit or ṇit is ṅit.
- The gā of 2458 and the root gāṅ "to go" are not included: the substitute's ṅ is otherwise idle (BM, TB).
- This makes the sic of adhyagīṣṭa ṅit.
- 2462 ghumāsthāgāpājahātisāṃ hali 6.4.66 — for the ghu roots, mā, sthā, gā, pā, hā and sā, ī replaces the ā before a hal-initial kit/ṅit ārdhadhātuka.
- ātaḥ comes from SK 2372 and īt from SK 2843. kṅiti comes from SK 2428 per BM, or from dīṅo yuḍaci kṅiti (SK 2507) per TB.
adhi a gā s ta → adhi a gī s ta → adhyagīṣṭa / without gāṅ adhyaiṣṭa. lṛṅ adhyagīṣyata / adhyaiṣyata (sya ṅit by 2461).- 1047 ik smaraṇe is also always with adhi. The liṅga is adhīgarthadayeśāṃ karmaṇi (2.3.52, SK 613), which would otherwise say just "igartha".
- vārttika iṇvad ikaḥ:
adhiyanti (yaṇ by 2455), adhyagāt (gā by 2458). - Some (kecit) extend only what is taught under the ārdhadhātuka heading, so there is no yaṇ: Bhaṭṭi's "adhīyan" (BM, TB).
- 1048 vī: veti, vītaḥ, viyanti, veṣi, vemi, vīhi, avet, avītām, aviyan.
- Some say that with aṭ the stem is anekāc, so yaṇ: avyan. "kecit" shows dissatisfaction (BM, TB).
- A separate root ī is embedded in the entry: eti, ītaḥ, iyanti, īyāt, aiṣīt.
- 1049 yā prāpaṇe (here = going): praṇiyāti, yātaḥ, yānti.
H. jus after the ā-final roots; yā … vac — running example: a-yā + jhi → a-yā + us → ayuḥ / ayān
- 2463 laṅaḥ śākaṭāyanasyaiva 3.4.111 — after an ā-final root, jus optionally replaces laṅ's jhi. jher jus comes from SK 2213 and ātaḥ from ātaḥ (SK 2227). The option comes from naming Śākaṭāyana (BM).
- The "eva" is for the next rules. The rule does not reach loṭ via loṭo laṅvat, because laṅ is named (BM).
ayuḥ / ayān. yāyāt (vidhiliṅ and āśīrliṅ), yāyātām / yāyāstām.- It does not reach daridrā (TB, under 2483).
- The roots 1050–1064:
- 1050 vā, 1051 bhā, 1052 ṣṇā, 1053 śrā, 1054 drā, 1055 psā.
- 1056 pā rakṣaṇe: pāyāstām, apāsīt. er liṅi and gātisthā… take only pibati (BM).
- 1057 rā and 1058 lā. Candra gives both the sense of giving.
- 1059 dāp lavane:
praṇidāti / pranidāti (optional by śeṣe vibhāṣā…, SK 2232), dāyāstām, adāsīt. It is not a ghu root (BM, TB). - 1060 khyā prakathane is used only in the sārvadhātuka.
- The liṅga is the vārttika sasthānatvaṃ namaḥ khyātre with its Bhāṣya. The khś-initial substitute exists to keep the jihvāmūlīya out of namaḥ khyātre, which would fail if khyātṛ came from this root (BM, TB).
- The Nyāsakāra says khyā is not used with sam.
- 1061 prā.
- 1062 mā māne is intransitive ("tanau mamus tatra na kaiṭabhadviṣaḥ"). With an upasarga in another sense it is transitive ("udaraṃ parimāti muṣṭinā"). It is not in nergada…: praṇimāti / pranimāti.
- 1063 vac paribhāṣaṇe: vakti, vaktaḥ.
- It is not used before anti. Others say not before a plural ending, and yet others not before jhi (BM, TB on the extent of each).
- vagdhi, vacyāt, ucyāt, avocat.
- Root entry 1064 vid jñāne.
I. vid: ṇal in laṭ, vidāṅkurvantu, u after kṛ, u for its a, ru for d — running example: vid + tip → vid + ṇal → veda; vid + loṭ → vidām + kṛ-u-tu → vidāṅkarotu; vidāṅkaru + tāt → vidāṅkurutāt
- 2464 vido laṭo vā 3.4.83 — after the adādi vid, ṇal etc. optionally replace the parasmaipada endings of laṭ (from parasmaipadānāṃ ṇalatusus…, 3.4.82, SK 2173).
- vindati and vidyati are excluded, since śa and śyan separate the laṭ from the root (BM, TB).
veda, vidatuḥ, viduḥ, vettha, vidathuḥ, vida, veda, vidva, vidma. Otherwise vetti, vittaḥ ….- liṭ viveda, vividatuḥ.
- With ām by uṣavidajāgṛbhyo 'nyatarasyām (SK 2341), the a-final vida is fixed by nipātana, so there is no guṇa of the light penultimate:
vidāñcakāra. luṭ veditā. - 2465 vidāṅkurvantv ity anyatarasyām 3.1.41 — in loṭ, for vid, the following are optionally fixed by nipātana:
- ām;
- no guṇa;
- luk of the loṭ;
- anuprayoga of kṛ in loṭ.
- "iti" shows that the person and number of "kurvantu" are not intended (BM; NSP fn).
- TB notes a view that the pada too is not intended (vyatividāṅkurutām).
- 2466 tanādikṛñbhya uḥ 3.1.79 — after the tanādi and kṛñ, the affix u, as an apavāda of śap. sārvadhātuke comes from SK 2756 and kartari from kartari śap (SK 2167) (BM).
- Naming kṛñ, already a tanādi root, is a liṅga that group operations (gaṇakārya) are anitya. Hence "na viśvased aviśvastam". BM reports that the Bhāṣya rejects the mention of kṛñ, so the form is doubtful; the Wikisource reading is "asiddham". TB accounts for aviśvasta separately.
vidām + kṛ + u + tu → vidāṅkarotu (eruḥ, guṇa of ṛ and of u; BM).- 2467 ata ut sārvadhātuke 6.4.110 — the a of u-final kṛ → u before a kṅit sārvadhātuka. upratyayāntasya comes from utaś ca pratyayād asaṃyogapūrvāt (SK 2334). BM (Wikisource) also carries kṛ from nityaṃ karoteḥ (SK 2548) and kṅiti from SK 2363.
- Being tapara, the new u takes no guṇa:
vidāṅkaru + tāt → vidāṅkurutāt, vidāṅkurutām. - hi takes luk by utaś ca… (SK 2334). That luk is ābhīya and asiddha, so the u still comes:
vidāṅkuru. - vidāṅkaravāṇi: āṭ is pit, so there is guṇa.
- laṅ avet, avittām,
aviduḥ (jus by SK 2226). - 2468 daś ca 8.2.75 — for a pada-final d of a dhātu, ru optionally before sip. sipi dhāto rur vā comes from SK 2485, padasya from the adhikāra, and ante from jhalāṃ jaśo 'nte (SK 84) (TB; also BM, Wikisource).
avid + s (s lost by hal-ṅyādi) → aveḥ / avet.- 1065 as bhuvi: asti.
J. as: loss of a, bhū, e before hi, ṣ after an upasarga — running example: as + tas → s + tas → staḥ; as + hi → e + hi → e + dhi → edhi
- 2469 śnasor allopaḥ 6.4.111 — the a of śna and of as is lost before a kṅit sārvadhātuka. sārvadhātuke comes from SK 2467 and kṅiti from SK 2363 (BM).
staḥ, santi.asi: the s of as is lost before si by tāsastyor lopaḥ (SK 2191).- sthaḥ, stha, asmi, svaḥ, smaḥ.
- Then 'ārdhadhātuke' (SK 2432) as heading.
- 2470 asterbhūḥ 2.4.52 — as → bhū before an ārdhadhātuka: babhūva, bhavitā. The loṭ is sārvadhātuka and keeps as: astu, stāt, stām, santu.
- 2471 ghvasor ed dhāv abhyāsalopaś ca 6.4.119 — for a ghu root and as, e replaces the final before hi, and the abhyāsa is lost as a whole (TB: dehi, dhehi).
as + hi → e + hi. The e is ābhīya and asiddha, so hi → dhi (2425). The a is lost by 2469: edhi.- With tātaṅ, the later rule, there is no e (sakṛdgati): stāt, stam, sta; asāni ….
- laṅ
āsīt (īṭ by astisico 'pṛkte, SK 2225). - āstām, āsan: the a-loss is ābhīya and asiddha, so āṭ.
- syāt; bhūyāt; abhūt. There is no īṭ, because that rule wants as and sic actually present.
- 2472 upasargaprādurbhyām astir yacparaḥ 8.3.87 — the s of as, after iṇ in an upasarga or after prādus, → ṣ when y or a vowel follows.
- saḥ comes from SK 335. iṇkoḥ (SK 211) and apadāntasya mūrdhanyaḥ (SK 210) are headings (BM).
ni + syāt → niṣyāt, prāduḥṣyāt; niṣanti, prāduḥṣanti. BM writes prāduṣṣ-.abhistaḥ: t follows.- Root entry 1066 mṛjū śuddhau.
K. mṛj; the five rudādi: iṭ, īṭ/aṭ, ṣ in supi, ṇ in an, at for jha — running example: mṛj + ti → mārj-ti → mārṣ-ṭi → mārṣṭi; rud + ti → rud-i-ti → roditi; a-rod + t → arodīt / arodat
- 2473 mṛjer vṛddhiḥ 7.2.114 — the ik of mṛj → vṛddhi before an affix taught after a root (dhātupratyaya), so not in parimṛḍbhyām (BM, TB). It is an apavāda of guṇa.
- vārttika: before a vowel-initial kṅit, optionally.
- j → ṣ by vraścabhrasja… (SK 294).
- laṭ
mārṣṭi, mṛṣṭaḥ, mṛjanti / mārjanti. - liṭ mamārja, mamārjatuḥ / mamṛjatuḥ, mamārjitha / mamārṣṭha. luṭ mārjitā / mārṣṭā.
- loṭ
mṛḍḍhi: hi → dhi, then ṣ, ṣṭutva, jaśtva. - laṅ amārṭ (see audit note), amārjam; luṅ amārjīt / amārkṣīt.
- Root entry 1067 rudir.
- 2474 rudādibhyaḥ sārvadhātuke 7.2.76 — after rud, svap (print), śvas, an, jakṣ, a vaḷ-initial sārvadhātuka takes iṭ. iḍ valādeḥ comes from SK 2184 (BM).
roditi, ruditaḥ.- Before hi, iṭ is the later rule, so no dhi:
rudihi. - 2475 rudaś ca pañcabhyaḥ 7.3.98 — after these five, a hal-initial pit apṛkta sārvadhātuka takes īṭ. īṭ comes from SK 2452, hali from SK 2443, piti sārvadhātuke from SK 2503 and apṛkte from SK 2448 (BM). The forms are given under 2476.
- TB: "pañcabhyaḥ" keeps it off jāgṛ (ajāgaḥ).
- 2476 aḍ gārgyagālavayoḥ 7.3.99 — for Gārgya and Gālava, aṭ comes instead, under the same conditions.
arodīt / arodat, aruditām (print), arudan, arodīḥ / arodaḥ.- yāsuṭ is antaraṅga over aṭ and īṭ, which depend on a particular root and ending:
rudyāt. - luṅ arudat (aṅ by irito vā) / arodīt (īṭ by astisicaḥ, since sic intervenes).
- 1068 ñiṣvap:
- svapiti (iṭ by 2474);
- suṣvāpa;
- suṣupatuḥ, suṣupuḥ: saṃprasāraṇa by SK 2409 before doubling;
- suṣvapitha / suṣvaptha.
- 2477 suvinirdurbhyaḥ supisūtisamāḥ 8.3.88 — after su, vi, nir, dur, the s of supi (svap in saṃprasāraṇa form), sūti and sama → ṣ. saḥ comes from SK 335 under the mūrdhanya heading (BM).
- kit liṭ. The root is joined first with the upasarga. Saṃprasāraṇa (later, SK 2409) and ṣatva come before doubling, and ṣatva is not asiddha for doubling (pūrvatrāsiddhīyam advirvacane):
su + ṣup + ṣup + atus → suṣuṣupatuḥ, suṣuṣupuḥ. - non-kit ṇal. Doubling comes first, then saṃprasāraṇa of the abhyāsa (SK 2408). ṣatva is asiddha (print) and halādiḥ śeṣaḥ is nitya, so the abhyāsa loses its v first. No "supi" shape is left, so no ṣ:
susuṣvāpa. - susvaptā; asvapīt / asvapat; svapyāt; āśīrliṅ supyāt / suṣupyāt; asvāpsīt.
- 1069 śvas: śvasiti, śvasitā, aśvasīt / aśvasat, śvasyātām, śvasyāstām. luṅ aśvasīt: no vṛddhi, by hmyantakṣaṇaśvasajāgṛṇiśvyeditām (SK 2299).
- 1070 an: aniti, āna, anitā, ānīt / ānat.
- 2478 aniteḥ 8.4.19 — the n of an after a cause in an upasarga → ṇ. upasargāt comes from SK 2287 and naḥ ṇaḥ from SK 235 (TB):
pra + aniti → prāṇiti. - 1071 jakṣ: jakṣiti, jakṣitaḥ (iṭ by 2474).
- 2479 adabhyastāt 7.1.4 — after an abhyasta, at replaces jha (jhaḥ from SK 2169). It is an apavāda of ant.
jakṣ + jhi → jakṣ + at-i → jakṣati: jakṣ is abhyasta by SK 428.ajakṣuḥ: jus by SK 2226, in turn an apavāda of at (TB).- Ujjvaladatta's reading of jakṣ with an initial y is an error.
- "The five rudādi are finished."
- 1072 jāgṛ: jāgarti, jāgṛtaḥ,
jāgrati (at, yaṇ). jāgarāṃcakāra (optional ām, SK 2341) / jajāgāra.
L. jāgṛ and daridrā: guṇa, and i or loss of ā — running example: jāgṛ + atus → jajāgaratuḥ; a-jāgṛ + us → ajāgaruḥ; daridrā + tas → daridri + tas → daridritaḥ; daridrā + ati → daridr + ati → daridrati
- 2480 jāgro 'viciṇṇalṅitsu 7.3.85 — jāgṛ takes guṇa (from mider guṇaḥ, SK 2346) before any affix other than vi, ciṇ, ṇal and a ṅit (a paryudāsa; BM). It applies both where vṛddhi would be due (jāgarakaḥ) and where guṇa would be barred.
jajāgaratuḥ: guṇa though kit.- ajāgaḥ; ajāgṛtām (ṅit, no guṇa).
- jāgṛ is abhyasta, so jus comes.
- 2481 jusi ca 7.3.83 — an ik-final aṅga takes guṇa before a vowel-initial jus (aci from ksasyāci, SK 2337).
a-jāgṛ + us → ajāgaruḥ.- vidhiliṅ jāgṛyuḥ: the us is not vowel-initial there.
- āśīrliṅ jāgaryāt, jāgaryāstām, jāgaryāsuḥ (by 2480).
- luṅ
ajāgarīt: nine applicabilities, yaṇ first:
1. yaṇ (iko yaṇ aci);
2. sārvadhātukārdhadhātuka guṇa;
3. sic-vṛddhi;
4. the jāgṛ-guṇa (2480);
5. the hal-final vṛddhi;
6. its prohibition by neṭi (SK 2268);
7. the option of ato halādeḥ (SK 2284);
8. vṛddhi again by ato lrāntasya (SK 2330);
9. its prohibition by hmyanta… (SK 2299).
- Root entry 1073 daridrā durgatau: daridrāti.
- 2482 id daridrasya 6.4.114 — the final ā of daridrā → i before a hal-initial kṅit sārvadhātuka. hali comes from SK 2497, kṅiti from SK 2363 and sārvadhātuke from SK 2467 (BM). It is an apavāda of 2483.
daridritaḥ (print).- 2483 śnābhyastayor ātaḥ 6.4.112 — the ā of śnā and of an abhyasta is lost before a kṅit sārvadhātuka (lopa and kṅiti from SK 2363, sārvadhātuke from SK 2467).
daridrati (at by 2479).- liṭ, anekāc so ām:
daridrāṃcakāra. - In āta au ṇalaḥ (SK 2371), o would have sufficed. au is taught so that au is heard after daridrā's ā is lost, and some take that as a jñāpaka for no ām: dadaridrau, dadaridratuḥ.
- dadaridra for ṇal is baseless (TB defends the rejection).
- Three vārttikas:
1. daridrā loses its ā as soon as an ārdhadhātuka is intended, which gives daridra with ac rather than *daridrāya (BM, TB);
2. in luṅ, optionally;
3. not before san, ṇvul, lyuṭ (print ṇvuli): didaridrāsati, daridrāyakaḥ, daridrāṇam.
- luṭ daridritā.
- laṅ adaridrāt, adaridritām,
adaridruḥ. The ā-option of 2463 does not reach it, because the ā is already gone (TB). - vidhiliṅ daridriyāt (by 2482); āśīrliṅ daridryāt.
- luṅ adaridrīt / adaridrāsīt (iṭ and sak).
- 1074 cakāsṛ: cakāsti,
cakāsati (at), cakāsāṃcakāra. - Some say dhi ca removes only the s of sic: cakāddhi. The Bhāṣya has only cakādhi.
M. cakās, śās, the Vedic dīdhī … vaś, hnuṅ: d/ru for s, i and śā for śās, loss of the final of dīdhī — running example: a-cakās-t → a-cakās → acakāt / acakād; śās + tas → śiṣ + ṭas → śiṣṭaḥ; śās + hi → śā + dhi → śādhi
- 2484 tipy anasteḥ 8.2.73 — a pada-final s → d before tip, except in as.
a-cakās + t (t lost by hal-ṅyādi) → acakād / acakāt. acakāsuḥ (jus).- TB's counter-examples: cakāsti (not pada-final), avaṭ (not s), cakāḥ with kvip (not tip), and Vedic ā from as.
- 2485 sipi dhāto rur vā 8.2.74 — a pada-final s of a dhātu → ru optionally before sip; otherwise d by jaśtva.
- s comes from sasajuṣo ruḥ (SK 162) and ante from SK 84; padasya is the adhikāra (BM). "dhātoḥ" is there for 2468.
acakāḥ / acakāt.- 1075 śāsu anuśiṣṭau: śāsti.
- 2486 śāsa id aṅhaloḥ 6.4.34 — the penultimate of śās → i before aṅ and before a hal-initial kṅit. upadhāyāḥ and kṅiti come from aniditāṃ hala upadhāyāḥ kṅiti (SK 415) (BM).
śās + tas → śis + tas → śiṣṭaḥ (ṣ by SK 2410, ṣṭutva).- śāsati (at, abhyasta); śaśāsatuḥ.
- loṭ śāstu, śiṣṭāt, śāsatu (print). Ours adds śiṣṭām.
- Only the parasmaipadin śās, the one that can take aṅ, is meant: āśāste keeps its ā (BM, TB's first view).
- 2487 śā hau 6.4.35 — śās as a whole (sthānaṣaṣṭhī now; TB) → śā before hi. It is an apavāda of the i.
- śā is ābhīya and asiddha, so hi → dhi:
śādhi. - laṅ
aśāt (d by 2484), aśiṣṭām, aśāsuḥ, aśāḥ / aśāt (2485). - āśīrliṅ śiṣyāt.
- luṅ
aśiṣat (aṅ by sartiśāstyartibhyaś ca, SK 2382). - aśāsiṣyat.
- 1076 dīdhīṅ: "these five are Vedic" (Mādhava; BM qualifies this at 2488). dīdhīte,
dīdhyāte (yaṇ by er anekācaḥ…, SK 272). - 2488 yīvarṇayor dīdhīvevyoḥ 7.4.53 — the final of dīdhī and vevī is lost before y or an i-vowel (lopaḥ from tāsastyor lopaḥ, SK 2191; BM).
dīdhī + i: the lopa is set aside, and the nitya e for the ṭi (ṭita ātmanepadānāṃ ṭer e, SK 2233) gives dīdhī-e → dīdhye.- dīdhyāṃcakre: guṇa barred by dīdhīvevīṭām, SK 2190.
dīdhitā, dīdhiṣyate: the ī is lost before iṭ.- 1077 vevīṅ (in the sense of vī).
- 1078 ṣasa and 1079 ṣasti, parasmaipadin.
- ṣasa: sasti, sastaḥ, sasanti (not abhyasta); sasāsa, sesatuḥ; sastu.
sadhi: the s-loss by dhi ca is asiddha by pūrvatrāsiddham (SK 12), so no luk by ato heḥ.- asat, asastām; asaḥ / asat; sasyāt; asāsīt / asasīt.
- ṣasti:
saṃs-t-ti → santi: s lost by skoḥ (SK 380), t optionally by jharo jhari savarṇe (SK 71). santaḥ, saṃstanti. Some give saṃsti, saṃstaḥ: when many consonants meet, two alone are not a saṃyoga. - 1080 vaś kāntau:
vaṣṭi, uṣṭaḥ (saṃprasāraṇa), uśanti; vakṣi; the second uṣṭaḥ is as transmitted.- uvāśa, ūśatuḥ; vaśitā.
- vaṣṭu, uṣṭāt,
uḍḍhi. - avaṭ / avaḍ, auṣṭām, auśan, avaśam.
- uśyātām / uśyāstām; luṅ avāśīt / avaśīt (print).
- Gaṇasūtra carkarītaṃ ca: the yaṅ-luk root (carkarīta) belongs to the adādi.
- BM draws parasmaipada-only from the heading. TB passes over the prācīna's parasmaipada statement as not relevant here.
- The print has no dhātu number for it; ours has 1081.
- 1082 hnuṅ: hnute, juhnuve, hnuvīta, hnoṣīṣṭa, ahnoṣṭa.
- Close: "ity adādayaḥ" (Here end the adādi roots). p46 opens with juhotyādibhyaḥ śluḥ (SK 2489).