to the two havirdhānas, or vehicles or vessels in which the soma-stalks are brought to the place of pressing; ⌊cf. our comm., p. 158, and Sāyaṇa on RV.⌋. The reason of the introduction of the hymn here is altogether obscure (unless it be the occurrence of the word yama in 38 b), and Kāuç. has no use for it. In a, our mss., so far as noted, accent mā́, but SPP. mentions ⌊only one⌋ among his ⌊as reading mā́⌋; and both editions give mā, as is undoubtedly correct. RV., in b, accents āítam, which, of course, is alone grammatically possible; but both AV. editions have āitám, with all the mss. TA. (in vi. 5. 1) also has the verse, and differs from RV. only in having étam: ⌊so, indeed, the Calc. ed., text and comm.! étam (not etám, pron.) can only be an imperative: but the Poona ed., text and comm., has āítam, like RV.⌋. Our text has sīdatam in d, with RV. and a part of our mss. (not O.Op.R.D.K. ⌊which read badly -tām⌋); but SPP. admits -tām, in spite of its inappropriateness, because ⌊-tam is supported by⌋ only one of his authorities and the comm. ⌊Is the consentaneousness of the mss. in the blundering -tām possibly due to a reminiscence of the correct āsīdatām of the immediately preceding context in TA.? cf. the case at x. 6. 17, and note.⌋ Vídāne might be from vid 'find'; the comm. glosses it with jānatī. One might conjecture that āítam in b is for ā॰āítam 'came,' but neither pada-text views it in that way. The verse cannot be made a full triṣṭubh without violent resolutions in the first pāda—which is, of course, properly prose. ⌊Considering the textual inaccuracies in the tradition of this passage, perhaps it is not too bold to suggest the query whether a mā has been lost: itáç ca mā amútaç cāvatām mā would be a perfect triṣṭubh pāda.⌋ Vāit. (15. 11) makes vss. 38 and 39 accompany in the agniṣṭoma ceremony the driving up of the two havirdhānas.
39. Be ye comfortable (? svā́sastha) for our soma.
I yoke for you ancient worship (bráhman) with obeisances; the song (çlóka) goes forth like a patron (sūrí) on his road; let all the immortals hear that.
⌊See my added note just preceding the translation of verse 38.⌋
The verse is pāda d of the RV. vs. x. 13. 2 ⌊of which pādas a, b, c immediately precede in our AV. text⌋, followed by pādas a, b, c of the RV. vs. 1. RV. accents in asvāsasthé; ⌊the AV. accent seems wrong;⌋ both pada-texts divide su॰ās-. RV. further reads in ⌊its b, our⌋ c, etu and sūrés, and at the end amṛ́tasya putrā́ḥ. The RV. verse is found also in VS. (xi. 5) and MS. (in ii. 7. 1) with the same readings throughout as in RV.; and in TS. (iv. 1. 12), which reads for our cví çlókā yanti pathyè ’va sū́rāḥ, and in d varies from RV. etc. only by having çṛṇvánti. The comm. glosses svā́sasthe with sukhāsanasthe; he takes yujé as 1st sing., as it is translated above; the form might, of course, be 3d sing., like duhé, çáye, etc. ⌊In d, çṛṇváttu is a misprint for çṛṇvántu.⌋
40. Three steps the form (?) ascended, it went (?) after the four-footed one (f.) with its course (vratá); it matches the song (arká) with the syllable; in the navel of right it purifies.
The translation is purely mechanical, the verse being highly obscure, and its AV. version evidently corrupt. RV. (x. 13. 3) reads in apáñca (for trī́ṇi) and aroham, in bemi for the absurd āitat (apparently a blundering extension* of āit), at end of cmima etā́m, and in dádhi (our abhí has to be omitted in translation) sám punāmi. It also has in arupás, which SPP. admits in his text on the authority of the majority of his mss. and of the comm. (the latter takes it from root rup, and makes it mean mṛtaḥ
