Atharva-Veda Samhita/Book XIX/Hymn 48

48. To night: for protection.

[As 47.—ṣaṭ. ānuṣṭubham: 1. 3-p. ārṣī gāyatrī; 2. 3-p. virāḍ anuṣṭubh; 3. bṛhatīgarbhā; 5. pathyāpan̄kti.]

Found also in Pāipp. vi., in connection with hymn 47, with which it also shares its liturgical application.

Translated: Griffith, ii. 305.

1. Now then what things we note, or what things are within the box, those things we commit to thee.

The first pāda here differs widely from the traditional text, which instead of cáyāmahe has ca yásmā (p. yásmāi) āha, or (a minority) yásmāha; the comm's text (according to SPP.) gives ca yasmā ha, but his explanation implies instead yasmāi (explained as = yasya); and SPP. accordingly reads in saṁhitā-text. ca yásmā ha, and in pada-text ca: yásmāi: ha, which is altogether to be condemned, since the two texts must correspond, and yásmā is also no word. SPP. in a note proposes further emendation of our cáyāmahe to ca yā́mahe (= īmahe or yācāmahe), which seems entirely unacceptable, as regards both form and sense. The whole verse is so obscure in meaning that we get little help from this element in reconstruction of the text; yet it is plain that we do not commit to another that which we do not possess, but are only wishing for. Ppp. also fails us; its reading of a, b is atho yāni tamassahe yāni cā ’ntaṣ pareṇihi. The yā́nivāntáḥ of b is by the pada-mss. strangely resolved into yā́n॰iva: antáḥ (but one of SPP's has p.m. yā́ni: vā: antáḥ). It is also strange that SPP. emends to cā ’ntáḥ, on the sole authority of the comm., though as regards the sense there is nothing to choose between ca and vā. All the mss. accent párīṇahi (except our pada-mss., which have pariṇáhi; and one of SPP's has s.m. pári: ṇahi); but SPP. follows our emendation parīṇáhi. Nearly half the mss. have at the end dadhmasi, and Ppp. agrees with them. The comm. in his explanation connects the verse closely with 47. 9: there one's ⌊domestic⌋ animals were spoken of, here one's very numerous house-articles (bahiṣṭhāni gṛhavartīni...vastūni); and in two classes: those out in open sight (anāvṛtadeçe), and those inside an enclosed house or the like (parito naddhe pariçrite gṛhādāu).

2. O night! mother! commit thou us to the dawn; let the dawn commit us to the day, the day to thee, O shining one (vibhādvarī).

Compare 50. 7, below*; also MB. i. 5. 15, where more such commitments are given. The comm. reckons the first division of the verse to vs. 1. The metrical definition of the Anukr. is mere arithmetic (12 + 10 + 8 = 30). The pada-text, both here and below, leaves vibhāvari undivided (RV. vibhā॰vari). *⌊Also vi. 107. 1-4.⌋

3. Whatsoever flies here, whatsoever that is crawling (sarīsṛpá) is here, whatsoever creature is on the mountain—from that do thou, O night, protect us.

The third pāda is wholly corrupt. All the mss. read parvatāyāsatvam (p. párvatāya: sáḥ: tvám), with some differences of accent (-yāsátvam, or -yāsátvám, or -yā́sátvám); and the comm. and SPP. (in saṁhitā) follow them (SPP. -yāsátvam; but in p., by emendation, -tāya: asátvam, since the comm. so understands). Our text emends to párvaṇy ā́saktam 'what has fastened on the joint,' which seems extremely unsatisfactory. The translation above is perhaps hardly better, but it implies a text much closer to the mss.—párvata (i.e. -te, hence -tay) ā́ sattvám; and so it may pass for what it is worth; it is by no means proposed as a definite solution of the difficulty. Ppp. has padvad āsunvan, which gives no help. Ppp. also reads yadi kim three times.

4. Do thou protect behind, thou in front, thou from above and from below; do thou guard us, O shining one; here we are, thy praisers.

5. They who follow (anu-sthā) the night, and who watch over beings, who defend all cattle—they watch over our selves (ātmán), they watch over our cattle.

The mss. accent jā́grati* in d and e, and in d the accent might well enough be retained, on the principle of antithesis; SPP. gives jāgrati, like us. Emendation to jāgratu would be decidedly welcome in both pādas. Ppp. has considerable variations: for b, yeṣu bhūteṣu jāgrabhi,† and, for d, e, tenā tvam asi jāgratu te naṣ paçubhir jāgratu: corrupt, but supporting our proposed jāgratu. *⌊It is very reasonable to suppose that in jā́grati at end of d and e we have two cases of assimilation of an original jāgratu to the jā́grati which stands correctly at the end of b (cf. note to xix. 43. 7), each case being doubly faulty, in respect, namely, of accent and of ending.⌋ †⌊In Roth's Collation there is a note which perhaps means that Ppp. reads jāgrati for rákṣanti at end of c; but I am not at all sure.⌋

6. Verily I know thy name, O night; thou art "ghee-dripping" (ghṛtā́cī) by name; as such Bharadvāja knows thee; do thou watch over our property.

In b, Ppp. combines vā ’si; in c and d it reads, with our edition, tvā and jāgṛhi, while SPP. gives tvā́m and jāgrati, the latter for jā́grati, as all the mss. ⌊save one⌋ read; our emendation was a perfectly obvious one, and should have been followed by SPP. The comm., to be sure, reads -rati, but, by the simple application of his general rule, that any verb-form can be used for any other, he is able to declare it = jāgartu—which SPP. has too much knowledge and conscience to do. The pada-mss. have tvā́m, tvām, and tvám; our tvā was an emendation, called for after tā́m. More than half of the mss. accent bharadvājás. In our text, the accent-sign printed over jā in d should be shifted to over dhi: it is a misprint.