Ppp. has suhavā instead of subhagā in c. The comm. explains abhiṣṭaye by abhyeṣaṇāya (taking it from root iṣ) or abhimataphalaprāptaye. In c, devó in our edition is a misprint for devī́.
3. Let the divine Açvins, lords of beauty, protect us; let dawn-and-night also make broad for us; O child of the waters, in case of any detriment to [our] household; O divine Tvashṭar, increase [us] in order to our completeness.
Ppp. has, in a, sudaṅsasā for çubhas patī; and in c, d it reads vihvṛtī kayasya cid devo ‘sūvandadhite çarma yacha naḥ. The comm. partly agrees with it in reading abhihvṛtī and kayasya (explaining it as for kasya); it also has at the beginning pātam, and takes the nouns that follow as vocatives. It understands ábhihvṛtī (p. -tī íti) as a locative (= -hvṛtāu or -hvaraṇe), as is done in our translation ⌊cf. JAOS. x. 389⌋. The accent requires amendment, to abhíhrutī. Florenz suggests the substitution of abhihrútas, which would be an easier reading.
4. To various divinities: for protection.
[Atharvan.—nānādāivatam. 1. pathyābṛhatī; 2. saṁstārapan̄kti; 3. 3-p. virāḍ gāyatrī.]
Found also in Pāipp. xix. Used by Kāuç. (23. 9) in a rite for prosperity in connection with the division of inherited property; and twice (124. 6; 135. 10) in the chapter of portents, when two crowns appear on some one's head and when the house-beam breaks; and it is reckoned to the puṣṭika mantras (note to 19. 1) and to the svastyayana gaṇa ⌊note to 25. 36⌋. For its employment with vi. 1, 3 etc., see under vi. 1; and in Vāit. with vi. 3 etc., see under vi. 3.
Translated: Florenz, 252 or 4; Griffith, i. 246.
1. Tvashṭar [protect] my address (vácas) to the gods, [also] Parjanya, Brahmaṇaspati; with sons, with brothers, let Aditi now protect our hard to surpass [and] saving power.
The verse is found also in SV. (i. 299), which has no for me in a, and, at the end, trā́maṇaṁ vácaḥ, thus rectifying the meter of the last pāda. Ppp. ends with trāmaṇe çavā. The form trā́maṇaṁ seems to be a bastard neut., corresponding to the masc. trā́māṇaṁ ⌊cf. JAOS. x. 522, 530: there seem to be no man-stems used as adjectives in the neuter⌋, and to have been avoided in AV. by the substitution of trā́yamāṇaṁ; of the resulting metrical disturbance the Anukr. takes no notice.
2. Let Aṅça, Bhaga, Varuṇa, Mitra, Aryaman, Aditi—let the Maruts protect [us]; may the hatred of that injurer pass away; repel the foe from near by.
The accent of pā́ntu (if correct) shows that only marútas is felt to be its subject; but Ppp. reads instead aditiḥ pātv aṅhasaḥ. In c, the comm. has abhihvṛtas. The last pāda is obscure, and at least in part corrupt. The want of accent of yāvayat is wrong, and its form is unmotived; emendation to yāváya or -yan can hardly be avoided. A'ntitam (for which Ppp. gives anthitam) is read by all the mss., and occurs again at viii. 5. 11, so that it must be regarded as the real AV. reading. It is emended in our edition to ánti tám; and the comm. also so understands it (tam eva çatrum antikāt). It is translated as if emended to ántitas; or ántitam might be an anomalous equivalent of antikam. The verse (12 + 8: 12 + 8 = 40) is not properly a saṁstārapan̄kti.
