________________
xlvi
SATAPATHA-BRAHMANA.
differences may be mentioned the form 'nililye,' which occurs once in the Mâdhyandina text (I, 2, 3, 1), and is otherwise only found in the Mahâbhârata; while the Kanva recension has the periphrastic form (nilayâm kakre), which the Mâdhyandina text also offers in the other two cases (I, 6, 4,1; IV, 1, 3, 1) in which the word occurs. On the other hand, the Kânvas seem to read invariably 'âtmani (dhâ or kri), where the Mâdhyandinas have 'âtman,' which is also (doubtless on metrical grounds) the more usual formation in the Rig-veda1. Of cases of material differences I can only at present adduce the passage I, 1, 4, 12 (M.), where the Mâdhyandina text is guilty of a transposition of the second and third castes, while that of the Kânvas gives them in the proper order. Though most of these points of difference between the two schools would seem to tell in favour of the higher antiquity of the Kânva text, there will always be great difficulty in deciding this question, as it is by no means impossible that these variations are entirely due to different local or family traditions. In favour of the latter alternative one or two other points may be mentioned. The Mâdhyandina text, as has already been remarked, offers not few grammatical and other differences between the first five and the succeeding four kândas, or, as we may say, between the Yâgñavalkya and the Sandilya books of the Shashipatha. Though I cannot speak with confidence on this point, as I have not yet examined the Kânva text of the Sândilya kândas, I may refer here to at least two points in which the Kânvas, in the Yagnavalkya portion, agree with the Sândilya portion of the Madhyandina text, viz. the use of the imperfect (aspardhanta) instead of the perfect (paspridhire) in the opening clause of legends; and the frequent employment of the particle 'vâva' in the place of 'vai.'
As regards the present translation of the first two kândas,
1 Another curious feature of the Kânva text is the frequent insertion of an 'ity uvâka' in the middle of speeches, much like the colloquial 'says he.' As an instance I may adduce K. IV, 2, 3, 3 (M. III, 2, 3, 5):-Sâ hovâlâ 'ham eva vo yagnam amûmuham iti hovâka yad eva mayi tanvânâ iti mâm yagñâd antaragâta tenaiva vo yagñam amûmuham iti te mahyam nu bhagam kalpayatety atha vo yagñah prarokishyata iti tatheti hokus, &c. The Kânvas also insert much more frequently an 'iti' in the middle of speeches.
Digitized by
Google