be the ten fingers which generate Agni by attrition (III, 29, 13; IV, 6, 8), or the streams of water among which Agni grows up, or streams of Ghrita or the like (comp. II, 5, 5 see below). Why these sisters are said to delight in the dark and in the bright goddess, the Night and the Dawn, remains doubtful.

But I think there are reasons which strongly recommend a correction of the text. In III, 55, II we read sya'vi ka ... arushi ka svasarau, 'the two sisters, the dark one and the red one.' Is it not probable that in our passage also it is the sisters who are described as dark and red? The dark goddess and the red goddess of course are Night and Dawn, and Night and Dawn, as is well known, are sisters in Vedic poetry. And furthermore the 'sisters' are described in our verse as amorously exciting the god Agni: for it cannot well be doubted that the svasarah of the third Pada are identical with the usati'h of the first: similarly it is said in 70, 7--in a hymn belonging to the same collection with out Sukta--that the Nights and Dawns augment Agni's greatness; in other passages Agni is represented as beloved by the Dawn, or as suckled by Night and Dawn (Bergaigne, Religion Vedique, II, 14. 15). The 'sisters' then are stated in our verse to delight (agushran), probably in Agni: now we read in II, 2, 2. abhi tva naktih ushasah vavasire agne vatsam na svasareshu dhenavah, 'The Nights and Dawns, Agni, have lowed at thee as the milch-cows in their stalls at their calves;' comp. Bergaigne, II, 15. Thus everything is clear, if we take the usati'h and the svasarah for the bright and dark goddesses, i. e. for the Dawns and Nights. The correction of the text to which this interpretation leads, is svasarah sya'vih arushih agushran, 'the dark and the red sisters have delighted (in Agni).' It is easy to understand that the corruption of the text was occasioned by the simile of the fourth Pada. The words ushasam na ga'vah seemed to demand a parallel nominative and a parallel accusative in the third Pada. The nominative was svasarah, but there was no accusative. Thus probably arose the reading sya'vim arushim.

Verse 2.

Note [1:71:2:1]1. On the Angiras as the fathers of the priestly tribes, see H. O., Religion des Veda, 278.

Note [1:71:2:2]2. The phrase ushasah ketuh occurs several times in the Rig-veda. I think that ketum usra'h means exactly the same; it has been shown by Kaegi, Festgruss an Boehtlingk, p. 49, and by Bartholomae, Bezzenberger's Beitrage, XV, 185, that a genitive sing. usra'h existed.

Verse 3.