Anarván, then, is the same as ánarus, Sat. P. Brâhmana III, 1, 3, 7; and from meaning originally without a wound or without one who can wound, it takes the more general sense of uninjured, invulnerable, perfect, strong, (cf. integer, intact, and entire.) This meaning is applicable to I, 94, 2; 136, 5; II, 6, 5; V, 49, 4; VII, 20, 3; 97, 5; X, 61, 13; 65, 3. In I, 116, 16, anarván seems to be used as an adverb; in I, 51, 12, as applied to slóka, it may have the more general meaning of irresistible, powerful.
There are two passages in which the nom. sing. árvân, and one in which the acc. sing. árvânam, occur, apparently
meaning horse. But in I, 163, 13, and IX, 97, 25, árvân stands in the Pada text only, the Samhitâ has árvâ̐ ákkha and árvâ̐ iva. In X, 46, 5, the text híri-smasrum ná árvânam dhána-arkam is too doubtful to allow of any safe induction, particularly as the Sâma-veda gives a totally different reading. I do not think, therefore, that árvat, horse, admits in the nom. and acc. sing. of any forms but árvâ and árvantam a. Pânini (VI, 4, 127) allows the forms arvân and arvânam, but in anarvan only, which, as we saw, has nothing in common with árvat, horse. Benfey: 'die rascheste (keinen Renner habend, uneinholbar),' the quickest (having no racer, hence not to be reached). M. Bergaigne (Journ. As. 1884, p. 188) tries to defend anarvan in the sense of anasva, without considering the grammatical objections. In VI, 66, 7 (not I, 6, 7) anasváh does not refer to yâmah.
The masculine anarvâ´nam after the neuter sárdhas is curious; sárdhas means might, but it is here used to express a might or an aggregate of strong men or gods, and the nom. plur. yé, who, in the next verse, shows the same transition of thought, not only from the singular to the plural, but also from the neuter to the masculine, which must be admitted in anarvâ´nam b. It would be possible, if necessary, to explain away the irregularity of anarvâ´nam by admitting a rapid transition from the Maruts to Indra, the eldest among the Maruts (cf. I, 23, 8. índra gyeshthâh márut-ganâh), and it would be easier still to alter sárdhas into sârdham, as an accusative singular of the masculine noun sárdha, which has the same meaning as the neuter sárdhas. There is one passage, V, 56, 9, which would seem to give ample countenance to such a conjecture:
tám vah sárdham rathe-súbham—â´ huve.
I call hither this your host, brilliant on chariots.
Again, II, 30, u, we read:
tám vah sárdham mâ´rutam—girâ´ úpa bruve.
I call with my voice on this your host of Maruts.
VIII, 93, 16. srutám vah vritrahán-tamam prá sárdham karshanînâ´m, â´ sushe.
I pant for the glorious, victorious, host of the quick Maruts.
From this sárdha we have also the genitive sárdhasya, VII, 56, 8 (4):
subhráh vah súshmah krúdhmî mánâmsi dhúnih múnihiva sárdhasya dhrishnóh.