तिङन्तस्वादिप्रकरणम्

Covers SK 2523–2533 (11 sūtras): all present. Every unit has its own kaumudī. The spine is not an Aṣṭādhyāyī run. The SK follows the Dhātupāṭha order of the svādi roots (1247 ṣuñ … 1280 dṛ, our numbering) and teaches each rule at the root where it first bites, drawing on AS books 3, 6, 7 and 8 (3.1.73 → 8.3.117 → 7.3.58 → 7.2.43 → 7.2.64 → 7.2.42 → 7.2.39 → 8.4.15 → 7.3.56 → 6.4.123 → 7.4.72). Two AS runs are split and taken out of order:

• the iṭ rules 7.2.38–44: 7.2.43 at SK 2526, 7.2.42 at 2528, 7.2.39 at 2529. The others were taught earlier or come later: 7.2.38 vṝto vā (SK 2391), 7.2.40 sici ca parasmaipadeṣu (SK 2392), 7.2.41 iṭ sani vā (SK 2625), 7.2.44 svaratisūti… (SK 2279);
• the velar rules 7.3.52–58: 7.3.58 at SK 2525 and 7.3.56 at 2531. The others are 7.3.52 cajoḥ ku ghiṇṇyatoḥ (SK 2863), 7.3.54 ho hanter ñṇinneṣu (SK 358), 7.3.55 abhyāsāc ca (SK 2430) and 7.3.57 sanliṭor jeḥ (SK 2331).

Every AS-ref below is checked against references/ashtadhyayi/sutraani-data.json.

• Anuvṛtti from later in the SK. Most rules here carry words from sūtras the SK teaches only later:
• sārvadhātuke from sārvadhātuke yak (3.1.67, SK 2756) at 2523;
• na from na raparasṛpi… (8.3.110, SK 3168) at 2524;
• ku from cajoḥ ku ghiṇṇyatoḥ (SK 2863) at 2525 and 2531;
• iḍ vā from iṭ sani vā (SK 2625) at 2526 and 2528;
• liṅsicor ātmanepadeṣu from 2528 at 2526, two units earlier;
• dīrghaḥ from graho 'liṭi dīrghaḥ (7.2.37, SK 2562) at 2529;
• abhyāsasya from atra lopo 'bhyāsasya (7.4.58, SK 2620) at 2533.
• In the other direction, 2532 draws et and abhyāsalopaḥ back from ghvasor ed dhāv abhyāsalopaś ca (6.4.119, SK 2471 in p45). The outline names them where BM does.

tiṅantasvādiprakaraṇam — outline (p48, SK 2523–2533, all 11 sūtras)

p47 (tiṅantadivādiprakaraṇam) closed at 2522 with 1246 gṛdhu, "vṛt" and "iti divādayaḥ".

• Where the prakaraṇa begins. The print heads it "tiṅante svādayaḥ 45". Our prakaraṇa number 48 differs systemically. The print sets the dhātu entry ṣuñ abhiṣave (ours 1247), with its gloss, before the sūtra svādibhyaḥ śnuḥ (3.1.73). BM (Wikisource only) adds that from here on the roots are ñit and so ubhayapadin.
• Where it ends. At 2533, with the Vedic run 1272 aha … 1280 dṛ and "vṛt". The print closes "iti svādayaḥ" ("Here end the svādi roots"). Ours lacks it, so the English supplies it in brackets (scan page-0418). p49 opens with tudādibhyaḥ śaḥ (SK 2534) and 1281 tuda.
• Contents in outline.
• 2523: su. śnu for śap: sunoti, sunvanti, sunavāni, asāvīt, abhiṣuṇoti.
• 2524: su before sya and san; ṣiñ, śiñ, ḍumiñ, ciñ. No ṣ for the s of su: visoṣyati. mamau (ā), praṇicinoti.
• 2525: ci (and stṛñ). An optional velar after the abhyāsa in san and liṭ: praṇicikāya / cicāya. staryāt.
• 2526: stṛ (and kṛñ). An optional iṭ in liṅ and sic before taṅ for a conjunct-initial root in ṛ: stariṣīṣṭa / stṛṣīṣṭa, astariṣṭa / astṛta. kṛṇoti, cakartha, akṛta.
• 2527–2529: vṛ (and dhuñ).
• iṭ before thal outside the Veda: vavaritha.
• An optional iṭ in liṅ and sic before taṅ for vṛ and ṝ-final roots.
• No lengthening of liṅ's iṭ: variṣīṣṭa / vṛṣīṣṭa, avariṣṭa / avarīṣṭa / avṛta.
• dhunoti, dudhuviva.
• 2530–2531: hi (then pṛ … sādha). ṇ after an upasarga and a velar after the abhyāsa: prahiṇoti, jighāya. āpnuvanti, āpnuhi, āpat.
• 2532: rādh "injure". e and loss of the abhyāsa: aparedhatuḥ, redhitha.
• 2533: aś; dambh; the Vedic tail. nuṭ after the long abhyāsa: ānaśe. By a vārttika, e and loss of the abhyāsa for dambh: debhatuḥ / dadambhatuḥ.

Threads that run through the prakaraṇa:

• śnu is the class-mark. śnu (ś is it) is sārvadhātuka by being śit, and so ṅit by sārvadhātukam apit (1.2.4, SK 2234). The root before it takes no guṇa, while śnu itself takes guṇa before a pit ending: sunoti, sunavāni, sunavai. Three rules then shape the u of śnu:
• yaṇ, not uvaṅ, before a vowel, by huśnuvoḥ sārvadhātuke (6.4.87, SK 2387): sunvanti. Where the u follows a conjunct there is no yaṇ: āpnuvanti; BM's aśnuvate.
• optional loss of the u before m and v, by lopaś cāsyānyatarasyāṃ mvoḥ (6.4.107, SK 2333): sunvaḥ / sunuvaḥ, sunvahe / sunuvahe. The u stays after a conjunct: āpnuvaḥ (BM, Wikisource).
• luk of hi by utaś ca pratyayād asaṃyogapūrvāt (6.4.106, SK 2334): sunu. After a conjunct there is no luk: āpnuhi.
• paratva decides one conflict. guṇa before the pit āṭ beats the yaṇ of huśnuvoḥ: sunavāni, sunavai (BM, TB).
• iṭ for ṛ- and ṝ-final roots in the ātmanepada liṅ and sic. 2526 gives the optional iṭ to a conjunct-initial root in ṛ (stṛ), and 2528 to vṛ and the ṝ-final roots. The two options then play out as follows:
• Without iṭ, the ending is jhal-initial and so kit by uś ca (1.2.12, SK 2368). There is therefore no guṇa (stṛṣīṣṭa, vṛṣīṣṭa, kṛṣīṣṭa), and sic is lost by hrasvād aṅgāt (8.2.27, SK 2369) (astṛta, avṛta, akṛta).
• With iṭ, the root takes ordinary guṇa (stariṣīṣṭa, astariṣṭa, variṣīṣṭa). vṝto vā (7.2.38, SK 2391) may lengthen the iṭ, but 2529 bars that in liṅ, so variṣīṣṭa has only a short iṭ. It is still optional in the ātmanepada sic (avariṣṭa / avarīṣṭa) and barred in the parasmaipada sic by sici ca parasmaipadeṣu (7.2.40, SK 2392) (avārīt).
• A velar after the abhyāsa. 2525 gives an optional ku for ci in san and liṭ (cikāya / cicāya). 2531 gives it for the h of hi except before caṅ (jighāya). Both draw ku from 7.3.52 and abhyāsāt from 7.3.55.
• ṣ and ṇ after an upasarga. ṣ comes by upasargāt sunoti… (8.3.65, SK 2270): abhiṣuṇoti, abhyaṣuṇot, but abhisuṣāva. 2524 bars it before sya and san (visoṣyati), and si is outside that list (visinoti). ṇ comes by ner gada… (8.4.17, SK 2285) (praṇicinoti) and by 2530 (prahiṇoti). tṛp takes no ṇ, being in kṣubhnādi (tṛpnoti).
• e with loss of the abhyāsa beyond ata ekahalmadhye (6.4.120, SK 2260). The rule is extended twice:
• to rādh "injure" (2532), whose vowel is long, since TB takes "ataḥ" as the vowel a as such;
• to dambh by the vārttika dambheś ca (2533), since there the n-loss is ābhīya and so asiddha.
• Dhātupāṭha segments. ṣuñ … dhuñ (1247–1255) are ñit ubhayapadin. The kaumudī's "now the parasmaipadin" covers ṭudu … sādha (1256–1263); BM bounds it at rādha sādha. aśū and ṣṭigha (1264–1265) are anudāttet. tika … dṛ (1266–1280) are parasmaipadin "ā gaṇāntāt". Within that stretch the gaṇasūtra chandasi makes aha … dṛ (1272–1280) Vedic. pṛ, spṛ and smṛ are also called Vedic ("trayaś chāndasāḥ").
• The kaumudī lists forms bare. The lakāra and puruṣa labels are the translator's, given in Pāṇinian terms. Derivation stages follow BM and TB as recorded in the unit notes. Dhātu numbers are ours throughout; ours run one lower than the print (ours 1247 ṣuñ = print 1248 … ours 1280 dṛ = print 1281).
• Where the commentary lies. Only the kaumudī blocks are translated (SK-TRANSLATE §25).
• TB printed under a neighbouring unit. TB on 2523 is at the tail of sk.2522.t1, and TB on 2525 at the tail of sk.2524.t1. sk.2525.t1 "vibhāṣā pati" is only the OCR of TB's last word "[cicī]ṣati". TB on 2528 is at the tail of sk.2527.t1. The print has no TB on 2529.
• BM misfiled or missing. sk.2524.b1 is BM on soḍhaḥ (SK 2358). The real BM on 2524 and on ṣiñ … ciñ, and the BM preamble to 2523, exist only in Wikisource (uttara-2/0047–0049).
• BM lead-ins printed under the preceding unit. 2526's lead-in ("stṛ + sīṣṭa") is at the tail of sk.2525.b1. BM on vṛñ (vavāra, vavruḥ) is at the tail of sk.2526.b1. 2530's lead-in on prahiṇoti is at the tail of sk.2529.b1.
• Artefacts. sk.2533.t1 after "iti svādayaḥ" is TB on tudādi, and sk.2533.n1–n2 are footnotes on bhrasj. Both belong to p49. sk.2533.b1 ends with a stray heading, "atha svārthikāḥ".

Where the p47–53 audit (.work/sk-run/p47-53-audit-fixes.md and the report .work/sk-run/p47-53-audit/sk2523-2533.md) corrects a reading, this outline follows the corrected reading. The fixer has applied the 2526 must-fix to the en.yml.

• 2526 (must-fix): astariṣṭa. BM's "'ṛtaś ca saṃyogādeḥ' iti iṭpakṣe guṇaḥ" cites the present sūtra (7.2.43) as the source of the iṭ: "in the option where iṭ comes by this rule, guṇa". The guṇa is the ordinary guṇa, since sic with iṭ is not jhal-initial and so not kit by uś ca (the translator's reasoning; BM names no rule for the guṇa). It is not ṛtaś ca saṃyogāder guṇaḥ (7.4.10, SK 2379), which holds in liṭ only; BM quotes that rule, with guṇaḥ, only for tastaratuḥ. Without iṭ, sic is lost by hrasvād aṅgāt.
• 2526: order. In the Aṣṭādhyāyī this sūtra (7.2.43) follows 7.2.42 (SK 2528) and continues its words. Bhaṭṭoji takes it first because stṛ is the root at hand.
• 2524: ḍumiñ. The ā by mīnātiminoti… (6.1.50, SK 2508) comes only in the ec-context, before a non-śit affix (BM, Wikisource, "ejviṣaye aśiti"). Hence mamau and mātā, but mimye and mīyāt keep i, the ending being kit.
• 2523: the bracketed restatement of the sūtra in k1 is redundant; it has no effect on the outline.
• Commentary readings the outline uses (audit COMMENTARY OCR):
• TB on 2523 reads "paratvād guṇaḥ" (OCR bhujaḥ).
• BM on 2531 has three separate glosses (āpnuvanti, āpnuvaḥ, āpnuhi), per Wikisource; ours conflates them by an eye-skip.
• BM on 2532 reads "krādiniyamān nityam iṭ" (Wikisource; ours "niṭ").

Print readings (NSP, scan-confirmed) that the outline follows. All are mūla items escalated to Ajai under the SK source-correction policy:

• ṣuñ abhiṣave (2523; ours aṅiṣave);
• śryukaḥ 2381 (2529; ours 2389, which is sthāghvor ic ca);
• hinoter hasya (2531; ours hinor);
• ā gaṇāntāt (2533, twice; ours āgaṇantāt);
• the Nyāsakāra sentence "apidviṣayakam iti nyāsakārādimate tu" (2533 k3; ours lacks it; supplied in brackets);
• the colophon "iti svādayaḥ" (2533; supplied).

Still open for Ajai (mūla):

• 2533: the vārttika division. The print puts the sign before "dambheś ca" only, while ours and Wikisource tag "etvābhyāsalopau vaktavyau" as part of it. The English stays neutral.
• 2527: ours and the print end "varitā" alone. BM's pratīka and Wikisource presuppose "varitā, varītā" (optional lengthening by vṝto vā). This may be a gap in the print tradition. Not added.

Ordering: SK order throughout. Groups D and E follow the SK in taking 7.2.43 (2526) before 7.2.42 (2528).

A. su: śnu for śap — running example: su + ti → su + śnu + ti → su-no-ti → sunoti; sunu + anti → sunvanti

• 2523 svādibhyaḥ śnuḥ 3.1.73 — after su and the roots that follow it, śnu comes before a sārvadhātuka denoting the agent. It is an apavāda of śap (BM). kartari comes from kartari śap (3.1.68, SK 2167) and sārvadhātuke from SK 2756; the sources are the translator's identification, and BM supplies only the words.
• Root entry (1247) ṣuñ abhiṣave (print): bathing someone, pressing, bathing oneself, distilling liquor. In the sense of bathing oneself it is intransitive.
• su-no-ti → sunoti: śnu is ṅit, so the root takes no guṇa; śnu takes guṇa before the pit tip.
• sunutaḥ: tas is ṅit, so no guṇa.
• sunu + anti → sunvanti: yaṇ by huśnuvoḥ (SK 2387), not the uvaṅ of aci śnudhātu… (BM).
• sunvaḥ / sunuvaḥ and sunvahe / sunuvahe: optional u-loss by lopaś cāsya… (SK 2333).
• liṭ suṣāva, suṣuve; luṭ sotā, showing that the root is aniṭ.
• loṭ sunu: luk of hi by utaś ca pratyayāt (SK 2334).
• sunavāni, sunavai: āṭ is pit, so the ending is not ṅit, and guṇa, the later rule, beats the yaṇ (BM; TB at the tail of sk.2522.t1).
• sunuyāt (yāsuṭ is ṅit). āśīrliṅ sūyāt: lengthening by akṛtsārvadhātukayor dīrghaḥ (7.4.25, SK 2298).
• luṅ asāvīt, where iṭ comes by stusudhūñbhyaḥ… (7.2.72, SK 2385) against the prohibition; ātmanepada asoṣṭa.
• With an upasarga:
• abhiṣuṇoti: ṣ by upasargāt sunoti… (SK 2270), then ṇ after the ṣ.
• abhyaṣuṇot: ṣ across the aṭ by prāk sitād aḍvyavāye 'pi (8.3.63, SK 2276).
• abhisuṣāva: no ṣ in the abhyāsa, by the restriction in sthādiṣv abhyāsena cābhyāsasya (8.3.64, SK 2277). The second part takes ṣ by ādeśapratyayayoḥ (8.3.59, SK 212).

B. su keeps its s before sya and san; then ṣiñ, śiñ, ḍumiñ, ciñ — running example: vi + so + sya + ti → visoṣyati (not *viṣoṣyati); mi + ṇal → mā + ṇal → mamau

• 2524 sunoteḥ syasanoḥ 8.3.117 — before sya and san, ṣ does not replace the s of su. The heading is apadāntasya mūrdhanyaḥ (8.3.55, SK 210), and na comes from SK 3168 (BM, Wikisource; TB).
• visoṣyati: the ṣ that upasargāt sunoti… would give after vi does not come.
• With san: abhisusūḥ. susūṣati is no example, since the restriction stautiṇyor eva ṣaṇy abhyāsāt (8.3.61, SK 2627) already bars the ṣ there (TB, BM). Recorded, not added.
• (1248) ṣiñ bandhane: sinoti.
• visinoti: no ṣ. sātpadādyoḥ (8.3.111, SK 2123) bars it, and si is not in upasargāt sunoti… (BM, Wikisource).
• siṣāya, siṣye.
• (1249) śiñ niśāne: palatal-initial; śetā.
• (1250) ḍumiñ prakṣepaṇe. ā replaces the final i by mīnātiminotidīṅāṃ lyapi ca (6.1.50, SK 2508), in the ec-context, before a non-śit affix (BM, per audit).
• mamau (ā, then au for ṇal); mamitha / mamātha.
• mimye and mīyāt keep i, the ending being kit.
• mātā, māsīṣṭa; amāsīt, amāsiṣṭām, amāsta.
• (1251) ciñ cayane: praṇicinoti, with ṇ by ner gada… (8.4.17, SK 2285). This leads into 2525.

C. ci: an optional velar after the abhyāsa in liṭ and san; then stṛñ — running example: praṇi + ci-ci + ṇal → praṇi-ci-kāy-a → praṇicikāya / cicāya

• 2525 vibhāṣā ceḥ 7.3.58 — for ci after an abhyāsa, a velar (ku) optionally replaces the c before san and liṭ.
• ku comes from cajoḥ ku ghiṇṇyatoḥ (SK 2863), abhyāsāt from abhyāsāc ca (SK 2430) and sanliṭoḥ from sanliṭor jeḥ (SK 2331) (BM, TB).
• praṇicikāya / cicāya; ātmanepada cikye / cicye.
• With san: cikīṣati / cicīṣati (TB). Recorded, not added.
• luṅ acaiṣīt, aceṣṭa.
• (1252) stṛñ ācchādane: stṛṇoti, stṛṇute.
• āśīrliṅ staryāt: yāsuṭ is kit and would bar guṇa, so guṇa comes by guṇo 'rtisaṃyogādyoḥ (7.4.29, SK 2380) (BM).
• BM also gives liṭ tastāra and tastaratuḥ, the latter with guṇa by ṛtaś ca saṃyogāder guṇaḥ (SK 2379, a liṭ rule).
• The lead-in "stṛ + sīṣṭa" leads into 2526.

D. stṛ: an optional iṭ in liṅ and sic before taṅ; then kṛñ — running example: stṛ + iṭ + sīṣṭa → star-i-ṣīṣṭa → stariṣīṣṭa / stṛ + sīṣṭa → stṛṣīṣṭa; a-stṛ-iṭ-s-ta → astariṣṭa / a-stṛ-s-ta → astṛta

• 2526 ṛtaś ca saṃyogādeḥ 7.2.43 — after a conjunct-initial root ending in ṛ, iṭ is optional for liṅ and sic before an ātmanepada (taṅ) ending. iḍ vā comes from iṭ sani vā (7.2.41, SK 2625), and liṅsicor ātmanepadeṣu from 7.2.42, which the SK teaches later at 2528 (BM, TB).
• stariṣīṣṭa / stṛṣīṣṭa: with iṭ the root takes guṇa. Without iṭ the ending is kit by uś ca, so there is no guṇa (compare kṛṣīṣṭa below).
• astariṣṭa / astṛta (per audit):
• In the iṭ option (iṭ by this rule), guṇa is ordinary, since sic with iṭ is not jhal-initial and so not kit. It is not by SK 2379, which is liṭ-only.
• Without iṭ, sic is lost by hrasvād aṅgāt (SK 2369).
• Why "before taṅ"? astārṣīt (TB).
• (1253) kṛñ hiṃsāyām: kṛṇoti, kṛṇute; cakāra.
• cakartha: iṭ is barred even before thal by kṛsṛbhṛvṛ… (7.2.13, SK 2293).
• cakre; kriyāt (riṅ by riṅ śayagliṅkṣu, 7.4.28, SK 2367).
• kṛṣīṣṭa: kit by uś ca (SK 2368), so no guṇa.
• akārṣīt (vṛddhi before sic, with r).
• akṛta: sic lost by hrasvād aṅgāt.
• (1254) vṛñ varaṇe. It is seṭ (BM), and leads into 2527.

E. vṛ: iṭ before thal outside the Veda; the optional iṭ of liṅ and sic before taṅ, never lengthened in liṅ; then dhuñ — running example: vavṛ + iṭ + tha → vavaritha (Vedic vavartha); vṛ + iṭ + sīṣṭa → variṣīṣṭa (not *varīṣīṣṭa) / vṛṣīṣṭa; avariṣṭa / avarīṣṭa / avṛta

• 2527 babhūthātatanthajagṛmbhavavartheti nigame 7.2.64 — in the Veda, babhūtha, ātatantha, jagṛmbha and vavartha are established without iṭ (nipātana).
• vavartha would already follow from kṛsṛbhṛvṛ…. Naming it is a restriction (niyama): the thal-prohibition for vṛ holds only in the Veda (BM, TB).
• Hence in ordinary language iṭ comes before thal: vavaritha.
• vavṛva, vavṛvahe: no iṭ, since kṛsṛbhṛvṛ… still holds outside thal (translator's note).
• luṭ varitā. The print has this alone; BM presupposes varitā / varītā. Open item, not added.
• TB's Vedic citations and ordinary forms: babhūvitha, tenitha, jagṛhima.
• 2528 liṅsicor ātmanepadeṣu 7.2.42 — after vṛṅ, vṛñ and roots ending in ṝ, iṭ is optional for liṅ and sic before an ātmanepada ending. iḍ vā comes from SK 2625, and vṝtaḥ from vṝto vā (SK 2391) (BM, TB).
• The examples come under 2529.
• Why "ātmanepadeṣu"? avārīt, avāriṣṭām, avāriṣuḥ, with no lengthening by sici ca parasmaipadeṣu (SK 2392) (TB, at the tail of sk.2527.t1).
• 2529 na liṅi 7.2.39 — after vṝ, the iṭ of liṅ is not lengthened.
• dīrghaḥ comes from graho 'liṭi dīrghaḥ (SK 2562), iṭ from ārdhadhātukasyeḍ valādeḥ (7.2.35, SK 2184), turned into a sixth-case word, and vṝtaḥ from SK 2391. liṅi is a seventh-case word in the sense of the sixth (BM).
• variṣīṣṭa: the lengthening due by vṝto vā does not come.
• vṛṣīṣṭa: without iṭ the ending is kit by uś ca, so there is no guṇa.
• avārīt: vṛddhi before sic, with no lengthening by SK 2392.
• avariṣṭa / avarīṣṭa / avṛta: with iṭ, lengthening is optional by vṝto vā. Without iṭ, sic is lost by hrasvād aṅgāt.
• (1255) dhuñ kampane. The short-vowel root is aniṭ: dhunoti, dhunute, adhauṣīt, adhoṣyat.
• It is also read with a long vowel: dhūnoti, dhūnute.
• dudhavitha / dudhotha: iṭ is optional by svaratisūtisūyatidhūñūdito vā (7.2.44, SK 2279).
• dudhuviva: the krādi restriction makes iṭ obligatory in the kit liṭ, overriding śryukaḥ kiti (7.2.11, SK 2381, print).
• luṅ adhāvīt: iṭ is obligatory by SK 2385. Ātmanepada adhaviṣṭa / adhoṣṭa.
• "Now the parasmaipadin roots": (1256) ṭudu upatāpe, dunoti.
• (1257) hi gatau vṛddhau ca. In prahiṇoti ṇ would not come, since pra and the verb are separate padas. That leads into 2530 (BM).

F. hi: ṇ after an upasarga, a velar after the abhyāsa; then pṛ … sādha — running example: pra + hi-nu + ti → pra-hi-ṇo-ti → prahiṇoti; hi-hi + ṇal → ji-hāy-a → ji-ghāy-a → jighāya

• 2530 hinumīnā 8.4.15 — ṇ replaces the n of hinu and mīnā after a cause (r or ṣ) standing in an upasarga.
• raṣābhyāṃ no ṇaḥ comes from 8.4.1 (SK 235), and upasargāt from upasargād asamāse 'pi ṇopadeśasya (8.4.14, SK 2287). "hinumīnā" is a sixth-case dual with ārṣa luk (BM).
• prahiṇoti. For mīnā: pramīṇāti (TB; not added).
• The n is that of the vikaraṇa: śnu for hi and śnā for mī (translator's note).
• 2531 her acaṅi 7.3.56 — a velar replaces the h of hi after an abhyāsa, but not before caṅ.
• ku comes from SK 2863, abhyāsāt from SK 2430, and haḥ, as a sixth-case word, from ho hanter ñṇinneṣu (7.3.54, SK 358) (BM).
• jighāya ("hinoter hasya", print). BM adds jighyatuḥ, jighayitha / jighetha.
• Why "acaṅi"? ajīhayat. The TB takes the prohibition as a jñāpaka that ku reaches hi extended by ṇi too: jighāpayiṣati. TB also gives jeghīyate. Recorded.
• (1258) pṛ prītau: pṛṇoti, partā.
• (1259) spṛ prītipālanayoḥ (others: prīticalanayoḥ; calana is "living", says Svāmin): spṛṇoti, paspāra.
• Some read smṛ: smṛṇoti.
• pṛ, spṛ and smṛ are called Vedic.
• (1260) āpḷ vyāptau: āpnoti, āpnutaḥ.
• āpnuvanti: no yaṇ by huśnuvoḥ, since the u follows a conjunct.
• āpnuvaḥ, which keeps its u likewise (BM, Wikisource).
• āpa, āptā.
• āpnuhi: no luk of hi, likewise (BM, TB).
• luṅ āpat: aṅ, because the root is ḷ-marked.
• (1261) śakḷ śaktau: aśakat (aṅ likewise).
• (1262) rādha and (1263) sādha saṃsiddhau: rādhnoti. The ā is long, so ata ekahalmadhye… does not reach it. That leads into 2532 (BM).

G. rādh "injure": e for its vowel and loss of the abhyāsa — running example: apa + rā-rādh + atus → apa + redh + atus → aparedhatuḥ

• 2532 rādho hiṃsāyām 6.4.123 — for rādh in the sense of injuring, e replaces the a and the abhyāsa is lost, before a kit liṭ and before thal with iṭ.
• The anuvṛttis (BM):
• et and abhyāsalopaḥ from ghvasor ed dhāv abhyāsalopaś ca (6.4.119, SK 2471);
• kiti from gamahana… (6.4.98, SK 2363);
• liṭi from ata ekahalmadhye 'nādeśāder liṭi (6.4.120, SK 2260);
• the whole of thali ca seṭi (6.4.121, SK 2261).
• ataḥ is carried too, but by implication (sāmarthyāt) it names the vowel a as such, so the ā of rādh is covered (TB).
• aparedhatuḥ, redhuḥ. The sense "injuring" comes from the upasarga; otherwise the form is rarādhatuḥ (BM).
• redhitha: iṭ is obligatory before thal by the krādi restriction ("nityam iṭ", Wikisource BM).
• luṭ rāddhā: dh by jhaṣas tathor dho 'dhaḥ (8.2.40, SK 2280) (BM).
• sādh: sādhnoti, sāddhā, asātsīt, asāddhām.
• "Now two anudāttet roots." (1264) aśū vyāptau saṃghāte ca: aśnute. That leads into 2533.

H. aś: nuṭ after the long abhyāsa; dambh by vārttika; the Vedic tail — running example: aś-aś + e → a-aś-e → ā-aś-e → ā-n-aś-e → ānaśe; da-dambh + atus → da-dabh-atus → debh-atus → debhatuḥ

• 2533 aśnoteś ca 7.4.72 — for aś of the śnu class, the augment nuṭ comes after the long a of the abhyāsa.
• The anuvṛttis: abhyāsasya from atra lopo 'bhyāsasya (SK 2620), and tasmān nuṭ from tasmān nuḍ dvihalaḥ (7.4.71, SK 2288). tasmāt means the a already lengthened by ata ādeḥ (7.4.70, SK 2248) (BM, TB).
• Why "aśnoteḥ"? So that the rule does not reach the kryādi aśa bhojane: āśa, āśatuḥ (TB).
• ānaśe.
• luṭ aśitā / aṣṭā: ṣ by vraścabhrasja… (8.2.36, SK 294), then ṣṭutva.
• lṛṭ aśiṣyate / akṣyate; vidhiliṅ aśnuvīta.
• āśīrliṅ akṣīṣṭa / aśiṣīṣṭa: iṭ is optional, the root being ūdit.
• luṅ āśiṣṭa / āṣṭa: sic is lost by jhalo jhali (8.2.26, SK 2281). ākṣātām.
• (1265) ṣṭigha āskandane: stighnute, tiṣṭighe, steghitā.
• "Now, up to the end of the gaṇa, parasmaipadin" ("ā gaṇāntāt", print).
• (1266) tika and (1267) tiga gatau ca: tiknoti, tignoti.
• (1268) ṣagha: saghnoti.
• (1269) ñidhṛṣā: dhṛṣṇoti, dadharṣa, dharṣitā.
• (1270) dambhu dambhane: dabhnoti, dadambha.
• The view of another grammar, adopted under bhvādi, is that liṭ is optionally kit for śranthi, granthi, dambhi and svañji.
• In the kit option the n is lost by aniditām hala upadhāyāḥ kṅiti (6.4.24, SK 415).
• That loss is ābhīya (asiddhavad atrābhāt, 6.4.22, SK 2183) and so asiddha. Hence ata ekahalmadhye… cannot give e and loss of the abhyāsa (BM: "tasyābhīyatvāt"). The vārttika therefore follows:
• vārttika dambheś ca (sk.2533.k2) — "e-substitution and loss of the abhyāsa are to be stated for dambh too".
• debhatuḥ / dadambhatuḥ.
• Before tip, sip and mip:
• Sudhākara and others extend the kit status to pit endings too: debha, debhitha, debha.
• The Nyāsakāra and others limit it to apit endings (sentence supplied from the print): dadambha, dadambhitha, dadambha only.
• dabhyāt: n lost by aniditām….
• (1271) ṛdhu vṛddhau. Some read tṛpa prīṇane: tṛpnoti, with no ṇ, being in kṣubhnādi (kṣubhnādiṣu ca, 8.4.39, SK 792).
• The gaṇasūtra chandasi is a heading to the end of the gaṇa.
• (1272) aha: ahnoti. (1273) dagha: daghnoti. (1274) camu: camnoti.
• (1275–1280) ri, kṣi, ciri, jiri, dāśa, dṛ hiṃsāyām: riṇoti, kṣiṇoti, ciriṇoti, jiriṇoti, dāśnoti.
• kṣi is used in ordinary language too: "na tad yaśaḥ śastrabhṛtāṃ kṣiṇoti".
• Others take ṛkṣi as one vowel-initial root (ṛkṣiṇoti), which makes five roots, not six (TB).
• Close: "vṛt" and "iti svādayaḥ" (Here end the svādi roots). The colophon is supplied from the print; BM closes alike. p49 opens with tudādi, 1281 tuda (SK 2534).