[paragraph continues] 4-7; 92, 2), must be considered as addressed not to Vâyu alone, but to Vâyu and Indra. It will be seen from Sâyana's introduction to hymn 90, that he, too, wrongly limits the sentence of the Anukramanikâ, aindryas ka yâ dvivaduktâh, to the fifth and following verses of hymn 90, and that he never alludes to this proviso again in his introductory remarks to hymn 91 and 92, though, of course, he explains the verses, in which a dual occurs, as addressed to two deities, viz. Indra and Vâyu. The same omission, whether intentional or unintentional, occurs in Shadgurusishya's commentary. The other commentary, however, assigns the verses of the three hymns rightly. The subject has evidently been one that excited attention in very early days, for in the Aitareya-brâhmana, V, 20, we actually find that the word vâm which occurs in hymn 90, 1, and which might be taken as a dual, though Sâyana explains it as a singular, is changed into te a.

In hymn VII, 104, rakshohanau might certainly be added as an epithet of Indrâ-Somau, and Shadgurusishya clearly takes it in that sense. The Anukramanikâ says: indrâsomâ pañkâdhikaindrâsomam râkshoghnam sâpâbhisâpaprâyam.

In hymn VIII, 67, it has been supposed that the readings Samada and Sâmada instead of Sammada and Sâmmada were due to a misprint. This is not the case. That I was aware of the other spelling of this name, viz. Sammada and Sâmmada, I had shown in my History of Ancient Sanskrit Literature (2nd ed.), p. 39, where I had translated the passage of the Sâṅkhâyana-sûtras in which Matsya Sâmmada occurs, and had also called attention to the Âsvalâyana-sûtras X, 7, and the Satapatha-brâhmana XIII, 3, 1, 1,

where the same passage is found. I there spelt the name Sâmmada, because the majority of the MSS. were in favour of that spelling. In the edition of the Âsvalâyana-sûtras, which has since been published by Râma Nârâyana Vidyâranya, the name is spelt Sâmada. My own opinion is that Sâmmada is the right spelling, but that does not prove that Sâyana thought so; and unless I deviated from the principles which I had adopted for a critical restoration of Sâyana's text, I could not but write Sâmada in our passage. B 1 and B 4 omit sâmada, but both give samadâkhyasya; Ca. gives likewise samadâkhyasya, and A. semadâkhyasya. This, I believe, was meant by the writer for sammadâkhyasya, for in the passage from the Anukramanî both A. and Ca. give sâmmado. I then consulted the commentary of Shadgurusishya, and there again the same MS. gave twice sâmmada, once sâmada, which is explained by samadâkhyamahâmînarâgaputrah. A better MS. of Shadgurusishya, MS. Wilson 379 gives the readings sâmmado, sâmmada, and sammadâkhyasya. The other commentary gives distinctly sâmanda. [I have adopted sammada in the new edition.]