Page:Atharva-Veda samhita volume 2.djvu/514

xix. 45- of us, the evil-hearted, the unfriendly, put upon himself.' The tád is supported by the comm., and in a measure by SPP's D.V. and C$3$.; the ápriyas, by the comm. and V. (as above) and the meter; and anāsmākás comes near to the ms.-readings, and, indeed, considering the ánāmakas of the living reciter V. and of the comm., is not ill supported.⌋

⌊It remains to note that no valid objection can be taken against making a nom. sing, of durhā́rda-s: its use as such is a natural way of avoiding the form from the consonantal stem (which was as much of a stumbling-block to the ancient Hindu as it is to the modern tyro in Sanskrit), and is entirely analogous to the use of hṛ́dayam rather than hṛ́d (cf. my Noun-Inflection, p. 471). The comm. is accordingly right in saying here durhārdo duṣṭacittaḥ, as he was also in glossing the suhā́rt of ii. 7. 5 by çobhanahārdaḥ sumanaskaḥ. The nom. durhā́rda-s is a form of transition to the a-declension, with durhā́rd-am (so viii. 3. 25) as its point of departure (cf. durhā́rdān of the mss. at xix. 28. 8); just so the later pā́da-s (from pā́d-am) replaces the older pā́t (Noun-Inflection, p. 471).⌋

⌊Of the older nom. sing. masc. or fern., however, the true form is suhā́r, p. su॰hā́ḥ, of which traces, albeit scanty, are found in the Veda: one is at MS. iv. 2. 5, p. 26$16$, priyā́ naḥ suhā́r ṇaḥ; and another is at AV. ii. 7. 5, where both ed's read yáḥ suhā́rt téna naḥ sahá. Here the saṁhitā-authorities taken together are divided between suhā́rt téna (so 11) and suhā́t téna (so 6); but the pada-authorities (7 out of 8) give su॰hā́t, the notable exception being the çrotriya K., who recited the true form su॰hā́ḥ. The saṁhitā-form for this ought to be suhā́s (téna), and possibly this form is concealed in the reading suhā́tténa of S$m$. etc. If not, then (since rtt = rt: Gram. §232) we may regard the combination suhā́rt t- as representing suhā́r t-, nom. suhā́r, with breach of the rule of sandhi requiring the change of suhā́r to suhā́s before t-. The motive for this breach was perhaps to avoid disguising still further the form suhā́r, itself extremely rare and none too easily recognized; and the motive is perhaps as clear as it is in the case of aves avet, ajāis ajāit, etc., cited by W., Gram. §555 a.—The nom. suhā́rt seems to be grammatical and not intolerable in the texts, and to be unparalleled (cf. Noun-Inflection, p. 472).⌋

3. Increasing from the force of the refreshment of the waters, born out of Agni Jātavedas—may the ointment that is four-heroed, that is of the mountains, make the quarters, the directions, propitious to thee.

4. The four-heroed ointment is bound to thee; be all the quarters free from fear for thee; firm shalt thou stand, like Savitar desirable; let these people (víças) render thee tribute.

In a, b Ppp. combines badhyatā ”ñj-, diço ‘bhayās. In c, the mss. ⌊with one or two exceptions⌋ read cāryà i- (p. ca: āryàḥ); SPP. alters the accent to cā́rya (p. ca: ā́ryaḥ); our emendation to vā́ryas ⌊W's B. has vāryà i-⌋ is not absolutely necessary, yet certainly a plausible improvement; and it is in a certain measure supported by Ppp., which gives vāri imā; the translation above implies it. The comm. understands arya, vocative, rendering it by svāmin. In d all the mss. have víças ⌊but Ws E. seems to have díças⌋, and our substitution of díças was hardly called for; but Ppp. favors it, reading