________________
XII, 4. COMMENTARY.
657
Stanza 3.
a. For kútá, cf. Geldner, Vedische Studien, I, 138. asya seems to refer to gavah, understood from the context.
b. kâtám ardati is difficult: kâtám is unquestionably a Prâkritic form for kartám, AV. IV, 12, 7 (cf. similarly kévate, RV. VI, 54, 7). For ardati one would expect some word for 'fall,' or 'push,' but the word means 'burst, go to pieces.' Perhaps kâtám is the subject, 'the deep ground bursts' (cf. RV. IV, 17, 2; AV. XIX, 9, 8, descriptive of earthquakes), but this does not quite do justice to kâtám.
d. The Paippalâda has for dîyate the preferable reading giyate, 'his property is wrung (from him).' For the interchange of the sound-groups di and gî (dy and gy), see the writer in Amer. Journ. Phil. VII, 482.
Stanza 4.
a. vilohitá, designation of some disease, also IX, 8, 1; perhaps, 'flow of blood from the nose.' Henry, Les livres VIII et IX de l'Atharva-véda, pp. 105, 142, 'décomposition du sang.' Both translations are purely etymological, but we may note that the word occurs in connection with other ailments of the head at IX, 8, 1. Cf. also lóhita VI, 127, 1, note.
c, d. The passage is not quite clear. sámvidyam, &π. λey., may mean 'possession.' Ludwig, 'name.' At any rate there seems to be an attempt to etymologise upon, or explain, vasa as a derivative from the root vas, 'control :' the character of the vasa, quasi 'controller,' is such that she cannot be deceived (duradabhna). We should expect ukyate for ukyase: the Paippalâda also reads ukyase. The Pet. Lexs. translate duradabhna (also st. 19), by 'getting the better of gates,' i. e. 'not to be confined.' This is ingeniously improbable, and contrary to the more usual rendering of the perfect passive participle. The prefixes dur- and a- represent a double negative for emphasis; the word is a stronger version of dû-dábha.
Stanza 4 may be suspected of having stood originally
[42]
Uu
Digitized by Google
