*⌊The Anukr. (the text of which is perhaps in disorder at this point) reads añjate vyañjata (vs. 18) indro mā marutvān iti pañca (vss. 25-29) jagatyas: tatrāi ’kādhikā (?) bhurig antyā (vs. 29) virāṭ. See under vss. 18 and 29.⌋
⌊The prose parts are the first pādas (the ūha-pādas) of vss. 25-28 and 30-35, and the yajurmantras, vss. 36-37: see Part III., below.⌋
⌊In Pāipp. (in xx.) is found of this hymn only verse 56.⌋
⌊Ritual uses.—Only eight verses (2, 5, 18, 25, 38, 39, 44, 45: the last two together twice, and both times in the order 45, 44) are used in Vāit., and, of course, in rites other than funeral rites: see under the verses. In Kāuç., about three quarters of the hymn (all but 21 vss.) are used, and used in the chapter (xi.: as noted by Whitney, page 814) on funeral rites: see under the verses. It is of critical interest that two blocks of verses (Parts II. and IV., as divided below, where see) which find no use in the funeral ritual, form each a nearly corresponding block in RV.⌋
⌊The provenience of the material of this hymn.—In this hymn, as compared with hymn 2 (see p. 830), the proportional part of material recurring in the RV. rises again, and is about 33 verses out of 73, or nearly one half.—The "Parts" into which the hymn is here divided are primarily for the convenience of the discussion, although some of them (as II., III., IV., V., VI.) have also a critical significance.
Part II., verses 21-24.—This block of verses corresponds, without changes of order, to the last varga of the second Vāmadeva hymn, more precisely to RV. iv. 2. 16, 17, 18, and 19 a, b, to which is then appended the last half-verse of RV. ii. 23, with the Gṛtsamada refrain. Neither ritual makes any use whatever of any verse or pāda of this part.
Part III., verses 25-37.—This part consists of two six-membered sequences, a and b (a=the five verses, 25-29: b = the six verses, 30-35), each sequence with one member for each of the "six directions" (E., S., W., N., fixed, and upward); the whole followed by two yajurmantras (vss. 36-37: comm., p. 1584).—All the 11 verses of sequences a and b have the refrain lokakṛtas etc. (a jagatī-triṣṭubh half-verse) in common. Moreover, all those 11 verses (except one, namely vs. 29) have as their second pāda the obscure jagatī-pāda, bāhucyútā pṛthivī́ dyā́m ivo ’pári; and for their first pāda an ūha-pāda applying in turn to one or another of the six directions.—In the excepted verse (vs. 29: see my note below), it would seem as if two directions had been crowded into one verse, the 'fixed' into pāda a and the 'upward' into pāda b: if so, it is this condensation that has reduced sequence a from 6 verses to 5, and effected the displacement of the pāda bāhucyútā etc.—Thus the refrains of this part are all metrical (smooth jagatī or triṣṭubh pādas), as is also the first half of vs. 29; while the ūha-pādas are prose, as are also the yajurmantras.—Parts III. a and III. b look to me like antiphonal sequences (cf. the introduction to ii. 5), the verses of a containing the prayers that are worded as if uttered by the suppliant dead man, and the verses of b containing the responses* of his helper, very likely the dead man's eldest son (see my note to vs. 25: but just how they were used, of course, I cannot say). If I am right, vs. 34, reinforced by the first two clauses of 36, would answer to 29 a; and vs. 35 would answer to 29 b. But against my view is the fact that we have tvā in 29 a where we should expect mā.—* [After writing the above, I find that Weber, Sb. p. 265, had expressed a similar view as to the responsive structure, and had proposed to emend tvā to mā.]