Atharva-Veda Samhita/Book XIX/Hymn 16

16. For safety and protection.

[Atharvan.—tṛcam.* mantroktabahudevatyam. 1. anuṣṭubh; 2.3-av. 7-p. bṛhatīgarbhā ’tiçakvarī.⌋ *⌊So the London ms.; the Berlin ms. says in fact dvyṛcam: see under vs. 2.⌋

This and the following hymns, to 23 inclusive, are wanting in Pāipp. The comm. has 16-19 used in the night, in a ceremony to be performed by the purohita, on the entrance of a king into his sleeping-house (according to Pariçiṣṭa iv. 5). The hymn is repeated below as 27. 14, 15.

Translated: Griffith, ii. 276.

1. Freedom from rivals in front, behind us [is] fearlessness made; Savitar [protect] me on the south, the lord of Çachī me on the north.

The comm. takes kṛtám in b as 2d du. impv., = kurutam, in spite of the accent, trying to find a dual subject in the two gods mentioned in c, d; and SPP., in obedience to this, even reads kṛtam, although twelve of his thirteen authorities (with all of ours) have kṛtám, the thirteenth evidently disagreeing with the rest purely by the accidental omission of an accent-mark.* It would not be impossible to take mā in c and d as object of dakṣiṇatás and uttarā́t. *⌊For the use of kṛtám (the participle), cf. tāír me kṛtáṁ svastyáyanam, above, 9. 12 c.—In his Collation Book, W. refers to RV. khila, 3. 4, which reads asapatnám purástān naḥ çiváṁ dakṣíṇataḥ kṛdhi: abháyaṁ sátatam paçcā́d bhadrám uttarató gṛhé.

2. From the sky let the Ādityas defend me; from the earth let the fires defend; let Indra-and-Agni defend me in front; let the Açvins yield (yam) refuge round about; crosswise let the inviolable [cow], let Jātavedas, defend [me]; let the being-makers be my defense (várman) on all sides.

In e the mss. all read tiraçcī́naghnyā́, which the pada-text resolves into tiraçcī́n: aghnyā́, and this SPP. retains, though tiraçcī́n is not a possible word. Our emendation to -cī́nā ’ghnyā́ is a very simple one (implying -cī́nā: aghnyā́); the translation is founded on it; but a more radical alteration of the pāda would be acceptable: something like, for instance, tiryák cā ’gnī́ rakṣatu jātávedāḥ; the jātávedās leads naturally to the suspicion that agnís is somehow hidden in the ’ghnyā́; the comm. indeed reads tiraçcīn agnī r-; but he is able to regard tiraçcīn as a masc. accus., implying asmān; or else as by Vedic license for -cīs, and this for -cībhyas, implying digbhyas (!); and such assumptions are forbidden us. The pada-mss. all read rakṣantu in e.

The Anukr. in its metrical definition treats this all as one verse, and the same treatment is implied by the summation at the end of the amivāka (see p. 928); but the comm. and one of our mss. make what follows the second avasāna into a separate or third verse; ⌊a like contradiction obtains as between the Anukr. and the comm. in the repeated passage, below, 27. 14, 15 (see the note); here, moreover, as noted above, the mss. of the Anukr. are at variance as to whether the hymn is to be reckoned as of 2 vss. or of 3⌋. The addition of bṛhatīgarbhā to the metrical definition is quite uncalled-for; ⌊doubtless because pāda b scans better as 8 syllables than as 9: no less uncalled-for is the addition of saptapadā, unless, dividing what follows the second avasāna into 3 pādas, we begin the seventh with an enclitic⌋.

⌊I suspect that our text consists of 6 pādas (8 + 8: ll + 11: 11 + 11 = 60, "atiçakvarī"), call them 1 vs. or 2, as you will. Pādas c and e and f have good triṣṭubh cadences; c is good triṣṭubh if we resolve indraagnī; so is e, with W's tiryák cā ’gnī́ r-; the presence of me in f is all that spoils f; and the absence of me after yachatām is all that spoils the cadence of d, if, substituting the grammatical equivalent, we pronounce açvínā ’bhítaḥ at the beginning.⌋