⌊xi. 2—so comm. and Keçava⌋ with offering thirteen different articles; and again (59. 25), in a rite for universal dominion (comm.), worshiping Atharvan; further (note to 42. 11), a schol. adds it to vii. 20. 6, to win wealth by Vedic knowledge. Vāit. has it in the agniṣṭoma (17. 2), repeated by the adhvaryu, as he looks at the udgātar. None of the Kāuç. uses seems at all characteristic.
Translated: Florenz, 249 or 1; Griffith, i. 245.
1. Sing at evening; sing greatly; put clearly, O son of Atharvan; praise god Savitar.
All the mss.—and SPP., following them—put the avasāna-mark after dhehi, thus falsely dividing the irregular gāyatrī into two pādas ⌊12: 12⌋; hence, of course, they accent ā́tharvaṇa; and most of the pada-mss. (all save our Bp.) read -ṇaḥ (as if the combination -ṇa st- were made by the common and allowable loss of the final ḥ before st: but many of our saṁhitā-mss. also have -ṇaḥ st-; SPP. makes no such report as to his). Both the other texts make the proper division, after ātharvaṇa; and so does Ppp., reading also gāya for the obscure dhehi. Both SV. and AÇS. have ā́ ’gād (which is better) for the first gāya; and SV. gives dyúmadgāman for dyumád dhehi. The comm. explains doṣo ⌊i.e. doṣā́ (instr.) u⌋ by rātrāv api, understands bṛhat as the sāman of that name, and supplies dhanam to dyumat in b. In our edition, the accent-mark over the sa of savitāram in c is lost.
2. Praise thou him who is within the river, son of truth (satyá), [him] the young, of unhateful speech, very propitious.
Again all the mss. spoil the structure of the verse by putting the division-mark after sūnúḥ. In both verses, SPP's text follows the mss., while ours emends in accordance with the true sense—which is now further supported by the other text, and by Ppp. An easier reading is offered by AÇS., namely tam u ṣṭuhy antaḥsindhuṁ sūnuṁ satyasya yuvānam: adro-. Indra is called "son of truth" in RV. viii. 58 (69). 4; the descriptions of the verse suit Savitar ill. The comm. understands 'the river (sindhu)' as "the ocean (samudra), in the midst of which the sun is seen rising," and foolishly gives as alternative sense of yuvan "repeller (yu) of darkness." The Anukr. apparently scans 11: 6 + 8 = 25.
3. May he, indeed, god Savitar, impel (sū) for us many amrtas, both the good praises, unto welfare.
The division of the mss. is this time that also of our text; but the meter is pretty hopeless (bhū́ri amṛ́tā would rectify b), and c apparently corrupt. Ppp. has, for b, sāviṣad vasupatir vasūni (making a, b nearly equal RV. vii. 45. 3 a, b), and AÇS. the same with omission of vasūni. In c, Ppp. ends with sugātum (perhaps 'to sing well both good praises'); AÇS. reads ubhe sukṣitī sudhātuḥ. The Pet. Lex. ⌊vii. 1045⌋ suggests, for c, ubhe srutī su gātave: cf. RV. ix. 78. 2; the varieties of reading show that the pāda was virtually unintelligible to the text-makers. The comm. takes suṣṭutī first as the bṛhat and rathantara sāmans, and then, alternatively, as the stuta and çastra. Of course, if the verse is to be taken (as seems necessary) as a spoiled gāyatrī, we ought to read sā́viṣat, with accent. The Anukr. seems to scan 11 + 6: 9 = 26. ⌊See p. lxix, note 2.⌋
