Note 2. I am doubtful about the three lakes of Madhu, here of rain, poured from their udders by the clouds. The number three is common enough, and Ludwig has pointed out a parallel passage from the AV. X, 10, 10-12, where we read of three pâtras, filled with milk and Soma. Many similar passages have been collected by Bergaigne, I, 177, but again without a definite result. The question is whether the three words utsa, kavandha, and udrin are meant as names of the three pâtras, in our passage, of the three lakes, or whether they should be taken as an apposition,

the three lakes, namely, the well (of the sky), the skin full of water, and udrin, the watering-pot. Udrin is elsewhere an adjective only, but I think we must here translate, 'the well, the water-skin, the watering-pot.'

Verse 12.

Note 1. On sudânavah as vocative, see Delbrück, Syntax, p. 106.

Verse 14.

Note 1. For adhi with genitive, one expects ati. But Delbrück doubts whether ati can govern the genitive. See Altind. Syntax, p. 440.

Verse 15.

Note 1. As ádâbhyasya can only refer to etâ´vatah, I have taken etâvat in the sense of gana, followed by eshâm. But I am not certain that the rendering is right.

Verse 16.

Note 1. I have ventured to translate drapsâh by torrents. Neither drops nor sparks nor banners seem to yield an appropriate simile, but I feel very doubtful. See VIII, 96, 13; IX, 73, 1.

Verse 22.

Note 1. I thought at first that by sám parvasáh dadhuh was meant the mixing or confounding together of heaven and earth; it being impossible, during a storm, to distinguish the two. But there is clearly, as Ludwig points out, an opposition between sám dadhuh and ví yayuh. I therefore take parvasáh in verse 22 in the sense of piece by piece, as in AV. IV, 12, 7. sám dadhat párushâ páruh, while in verse 23 it means in pieces.

Verse 25.

Note 1. On siprâh, see note to II, 34, 3.

Verse 26.

Note 1. Ukshnáh rándhram, 'the hollow of the bull,' whatever that may be, is not mentioned again. If it is meant for

the dark cloud which hides the rain, then the roar of the bull would be the thunder of the cloud, stirred by the Maruts. Aukshnorandhra, however, is the technical name of certain Sâmans, so that Ukshnorandhra may have been, like Usanâ (later Usanas), a proper name. See Tândya Br. XIII, 9, 18; 19.

Note 2. If usánâ stands for usanayâ it might mean, 'with desire,' but it seems more likely that it refers to the Rishi, who is called Usanâ in the Rig-veda, and Usanas in later writings. See Lanman, p. 562, l. 21; Bergaigne, II, 338, n. 3; Schmidt, K. Z. XXVI, 402, n. 1.

Verse 27.