Introduction] PROBABLE AUTHORSHIP [xxiii hardly be placed later than 400 B.C. Though ślokas have undoubtedly been added here and there and some modifications of diction have probably crept in, the authenticity of the text as a whole is better guaranteed than that of perhaps any other ancillary Vedic work, by the fragments of it which are found embedded, from beginning to end, in the Sarva- nukramaṇi ¹, itself a text which has been handed down with peculiar care. The early character of the Brhaddevata is further indicated by its mythological standpoint, which is that of the Vedic triad Agni, Indra, and Sürya; Indra, moreover, is the predominant deity. The com- paratively large proportion (one-fourth) of narrative which it contains, in illustration of the hymns of the Rg-veda, is thus the earliest col- lection of epic matter which we possess, dating as it does from a period when the Mahabharata could only have been in an embryonic state 2. The fact of its containing this additional epic matter, points to its being somewhat later in date than the Devatänukramani, which though not extant is known from ten quotations in Sadgurusisya and was doubtless an index of deities only. There seems no reason to doubt the tradition which attributes this and the other short Anukramanīs to S'aunaka. 3 But is the ascription to him of the Brhaddevată also probable? There can be little doubt that the Devatānukramani, as the simpler of the two, was also the earlier. It is in itself unlikely that the same author should have reproduced this work in another form with exactly the same object, viz. to give a serial account of the deities of the Rg-veda. This assumption is borne out by the internal evidence of the Brhaddevata itself. The writer several times speaks of himself in the first person. On the other hand, S'aunaka is mentioned fifteen times by name, generally with other authorities, especially Yaska, when different views are stated. Indeed, the impression is once or twice conveyed, that S'aunaka's view was not shared by the writer. In any case, the author could not possibly have meant himself when speaking of the 'Acarya Saunaka' (ii. 136; 6 1 I have shown in an article on the Arsa- nukramanī, Festgruss an R. von Roth (Stutt- gart, 1893), p. 112, that the Sarvanukramanī contains metrical fragments from that work also; and it is clear that the Sarvãnukramani borrows from it throughout. In the longer recension, the Brhaddevata has one śloka (v. 144) in common with the present text of the Mahabharata. One half- śloka (viii. 98cd) has also a closely parallel form in the Bhagavadgita. All points in which the narrative matter of our text is akin to the Great Epic deserve careful inves- tigation. Cp. my edition of the Sarvänukramani, pp. vi and 206.
- Brhaddevatăi. 1; ii. 146; iv.32 (B); vii. 84 &c.
5 The verb is in a past tense, aha, uvaca, mene, atravit, excepting in two which his opinion is stated with that of Yaska and Galava (manyante), or that of Yaska and Sakaṭāyana (manyate). ages in See the references in Appendix ii, p. 115, under 'Saunaka.'
