Atharva-Veda Samhita/Book IV/Hymn 37

37. Against various superhuman foes: with an herb.

[Bādarāyaṇi.—dvādaçarcam . ajaçṛn̄gyapsarodevatyam. ānuṣṭubham: 3. 3-av.6-p. triṣṭubh; 5. prastārapan̄kti; 7. paroṣṇih; 11. 6-p. jagatī; 12. nicṛt.]

Found (except vs. 9) in Pāipp. xiii. (in the verse-order 1-4, 7, 6, 5, 12, 8, 10, 11), but in a much defaced condition. Used by Kāuç. with the preceding hymn, as one of the cātanāni (8. 25); but also independently (28. 9) in a remedial rite against possession by evil spirits. And the comm. quotes it from Nakṣ. K. 21 ⌊error for Çānti K., says Bloomfield⌋, as employed in a mahāçānti called gāndharvī. ⌊As to Bādarāyaṇi, see introduction to hymn 40.⌋

Translated: Kuhn, KZ. xiii. 118 (interesting Germanic parallels); Ludwig, p. 352; Griffith, i. 180; Bloomfield, 33, 408; Weber, xviii. 144.

1. By thee of old the Atharvans slew the demons, O herb; by thee did Kaçyapa slay; by thee Kaṇva, Agastya.

The comm. explains that one or other of the specified plants, the sahamānā etc., is here addressed.

2. By thee do we expel (cat) the Apsarases, the Gandharvas; O goat-horned one, drive the demon; make all disappear by [thy] smell.

'Drive' (aja) in c is a play upon the name goat (aja-) in 'goat-horned.' The comm. declares the epithet to be equivalent to viṣāṇin (Odina pinnata), and to be given on account of the shape of the fruit. ⌊Dhanvantari, p. 23, Poona ed., gives meṣaçṛn̄gī and viṣāṇikā as synonyms of ajaçṛn̄gī.⌋ Ppp. has in b cātayāmasi instead of -mahe.

3. Let the Apsarases go to the stream, to the loud (?) down-blowing of the waters: Guggulū, Pīlā, Naladī, Āukṣagandhi, Pramandanī: so go away, ye Apsarases; ye have been recognized.

⌊See Weber's note and reference to Rumpelstilzchen.⌋ Tārá in b is rendered "crossing"; but as this sense is found nowhere else, it seems safer to take the word as the adjective, common later; the comm. glosses it with tārayitāram, a worthless etymological guess. After it, instead of avaçvasam, the comm. reads iva svasam (= suṣṭhu nāupreraṇakuçalaṁ yathā), and, strangely enough, Ppp. has the same. As everywhere else where the word occurs, the mss. vary between gulgulu and guggulu, and SPP. reads the former and our edition the latter; here the decided majority, with Ppp. and the comm., give gulg- (our Bp.H.K. have gugg-). Pādas c and d appear to be made up of names of Apsarases, all formed upon odor-names: guggulū́ is fem. to gúggulu 'bdellium,' and naladī́ to nálada 'nard'; pramandanī́ is related with pramanda 'a certain fragrant plant'; and āukṣágandhi means something like 'ox-smell'; but the comm. declares them to be pañca homadravyāṇi 'five articles of oblation'; Ppp. reads prabandhinī in d. Most of our mss. accent apsarásas in e, but SPP. reports only one of his as doing so; both editions read apsarasas. The comm. makes a different division of the material, reckoning the refrain (our e, f) as a verse with our 4 a, b, and omitting the refrain in 4 (much as it treated 34. 5-7 above); SPP. follows the Anukr. throughout (see under the next verse). The comm. reads in f pratibaddhās; prátibuddhā abhūtana is found also as RV. i. 191. 5 d. Ppp. adds between our d and e yatrā marty apsv antaḥ: samudre turūṇyarī turvaçī puṇḑarīka (not followed by an avasāna-sign). The Anukr. definition of the verse ought to read jagatī instead of triṣṭubh. ⌊in the prior draft, W. notes the suggestions concerning avaçvasám given by BR. i. 490 and OB. i. 126 and implying ava = 'away'; but rejects them.⌋

4. Where [are] the açvatthás, the nyagródhas, great trees, with crests: thither go away, ye Apsarases; ye have been recognized.

The division and numbering in our edition of this verse and the two next following is faulty, owing to the unclearness of the mss. first used; the correct division, agreeing with the Anukr., is given by SPP., and our translation follows it ⌊and makes clear what it is⌋; vss. 3-5 all end with the refrain tát páre ’tā- etc., and this, with the number 4, needs to be added in our text after çikhaṇḑínaḥ. In Ppp., the place of this verse is taken by the addition reported above, under vs. 3. The comm. takes çikhaṇḍin as meaning "peacocks"; he quotes TS. iii. 4. 84 to the effect that certain trees, including açvattha and nyagrodha, are the houses of Gandharvas and Apsarases.

5. Where [are] your swings, green and whitish; where cymbals [and] lutes sound together—thither go away, ye Apsarases; ye have been recognized.

⌊Change the number 4 to 5 at the end of the first line of p. 74 of the edition.⌋ The accent karkaryás as nom. plur. is false, and must be emended to -ryàs, as read by SPP. with half of his mss., and a part (O.Op.D.) of ours (our P.M.W. give karkayás). Again nearly all our mss., with some of SPP's, accent apsarásas in the refrain. Ppp. is quite corrupt: yatra vokhsā haritārjunā ghātāṣ karkarī asaṁvadanti. The first half-verse (13 + 11: 8 + 8) is irregular.

6. Hither hath come this mighty one (vīryàvant) of the herbs, of the plants; let the goat-horned arāṭakī́, the sharp-horned, push out.

⌊Put a simple avasāna-mark in place of the number 5.⌋ Arāṭakī seems to be used here as specific name of the herb in question; but the comm. takes it as epithet, deriving it from a-rā 'non-giving' with a derivative from aṭ 'go,' and meaning hiṅsakān uccāṭayati! Ppp. adds two more pādas: ape ’te ’to ‘psaraso gandharvā yatra vo gṛhāḥ.

7. Of the hither-dancing, crested Gandharva, Apsaras-lord, I split the testicles, I bind fast (?) the member.

All the mss. read in c ápi yāmi, which SPP. accordingly retains; our emendation to dyāmi is unsatisfactory, both as regards the sense and because dā is not elsewhere used with api (neither is yā); one might also guess abhí yāmi "attack"; the comm. explains it by apigataṁ niruddhaṁ karomi, which is worthless. The comm. gives "peacock" as an alternative equivalent of çikhaṇḑin with an implied comparison: "dancing like a peacock." Ppp. begins with īyaṁ vīruc chikh-. ⌊If the definition of the Anukr. (8 + 8: 12) is right, pāda c lacks a syllable.⌋

8. Terrible are Indra's missiles (hetí), a hundred spears of iron; with them let him push out the oblation-eating, ávakā-eating Gandharvas.

Half our mss., and the large majority of SPP's, read at the end of this verse ṛṣata; both editions give ṛṣatu, as in the next verse. Avakā is defined as a certain grass-like marsh-plant, Blyxa octandra, the same with çāivala or çāivāla; the comm. defines it as jaloparisthāḥ çāivālaviçeṣāḥ, but attempts no explanation of why the Gandharvas should be supposed to eat it. He reads in b çatapṛṣṭīs (one feels tempted to emend rather to çatábhṛṣṭīs), and in c abhihradān (for haviradān). The Anukr. takes no notice of the redundant syllable in c (also in 9 c).

9. Terrible are Indra's missiles, a hundred spears of gold; with them let him push out the oblation-eating, ávakā-eating Gandharvas.

This very slightly varied repetition of vs. 8 is wanting in Ppp. All the mss. have ṛṣatu at the end here.

10. The ávakā-eating ones, scorching, making light (?) in the waters—all the piçācás, O herb, do thou slaughter and overpower.

All our pada—mss. read in b jyotaya॰māmakā́n as a compound, and it seems very strange that SPP. gives in both forms of text jyotaya māmakā́n, as two independent words, and reports nothing different as found in any of his authorities; it is perhaps an oversight on his part. Either reading being plainly untranslatable, the rendering given implies emendation to jyotayamānakā́n, as the simplest and most probable alteration; several cases of such expansions of a participle in māna by an added -ka occur ⌊Skt. Gram. §1222 g, f; cf. Bloomfield's note⌋, one of them (pravartamānaká) even in RV. Ppp. reads as follows: avakāçaṁ abhiçāco bicchi dyātayamānakāṁ: gandharvān sarvān oṣadhe kṛṇu tasvaparāyaṇaḥ; this supports the proposed reading in its most essential feature, -māna- for -māma-, and further favors the version of the comm., dyot- for jyot-. R., in the Festgruss an Böhtlingk (p. 97), had ingeniously conjectured the word as a name for the will-o'-the-wisp, deriving it from jyotaya mām 'give me light,' by an added suffix -aka. The comm. paraphrases by matsambandhino gandharvān udakeṣu prakāçaya. ⌊Cf. Whitney, Festgruss an Roth, p. 91; also note to ii. 3. 1.⌋

11. One as it were a dog, one as it were an ape, a boy all hairy—having become as it were dear to see, the Gandharva fastens upon (sac) women; him we make disappear from here by [our] mighty (vīryàvant) incantation (bráhman).

In our edition, stríyam at the end of d is a misprint for stríyas, which all the mss. have, with no avasāna-mark following, though distinctly called for by the sense, and therefore supplied by us; Ppp., however, reads striyam, with sajate before it; and it omits the last pāda, f: which omission would furnish an excuse for the absence of interpunction after stríyas.

12. Your wives, verily, are the Apsarases; O Gandharvas, ye are [their] husbands; run away, O immortal ones; fasten not on mortals.

All the pada-mss. commit in c the palpable error of dividing dhāvatāmartyā(ḥ) into dhāvata: martyāḥ, as if the ā which follows dhāvat- were one of the common prolongations of a final vowel in saṁhitā; the comm., however, understands amartyāḥ, and SPP. admits this by emendation into his pada-text. Ppp. has for c apakrāmat puruṣād amartyā, which supports amartyās in our text.