________________
IX, 2. COMMENTARY.
593
b. Prof. Roth in the Pet. Lex., s. v. bhas, regards this passage as corrupt, and conjectures yasmâd bîbhatse yak ka nâ-bhinande, 'which I loathe and which I do not enjoy.' The motive of the correction, in addition to the poor metre, is the usual transitive use of abhí nand, 'take pleasure in,' and the like; this does not seem to me to warrant so complete a transformation of the text.
c. The Pâda is hypermetric, and may be normalised by reading munkâ for mužkâmi. But the imperative first sing. act. without ni is not elsewhere known in the Atharvan.
Stanza 3.
b. asvagátâ occurs but one other time, AV. XII, 5, 40, asvagátâ párihnutâ, a very obscure passage. The Pet. Lexs. translate 'heimatslosigkeit;' Ludwig, 'unfreiheit ' (cf. Der Rigveda, III, 284); Henry, 'dépendance.' The adjective ásvaga occurs in a closely parallel passage, XII, 5, 45, ásvagam ápragasam karoti; I would compare svastha and asvastha, 'well' and 'unwell;' svasthatâ and asvasthatâ,' well-being' and 'diseasedness.' For ávarti Vâg. S. XXX, 12 has ávariti, 'trouble' (Ludwig, 'verarmung'); cf. perhaps Avestan hamvareti of opposite meaning, ' defence, courage.'
Stanza 5.
Vâk Virâg (cf. Khând. Up. I, 13, 2) is the same Vâk who is designated RV. VIII, 100, 11, 'as the milch-cow whom the gods begot;' cf. AV. VIII, 9, 2, and Oldenberg, Zeitschr. d. Deutsch. Morgenl. Gesellsch. XXXIX, 54 ff. In Pâda d the metre is sensibly relieved by dropping pasávo (so also st. 16); nevertheless I should hesitate to correct, because the same RV. stanza states 'that multiform animals (pasával) speak her (vák).' The argument may, of course, be turned the other way, by assuming that pasávo is due to a reminiscence from that very stanza.
Stanza 9.
a, b. For indrâgnî, nominative for vocative, coupled with kama, vocative, cf. Delbrück, Altindische Syntax, p. 105.
[42]
Q q
Digitized by
Google