________________
178
THE INDIAN ANTIQUARY
(AUGUST, 1915
tion of this period, as far as it is known, would be 171 to 104. Hence these texts are surpassed in the prevalence of the long over the short fifth syllable by the MNU., which does not come up, however, to the Dhammapada the figures of which are 131 : 18. This is just what we expect in the case of the two Brâhmaņas. For, as is well known, the oldest Recension of the MNU., bearing the name of the Dravidas, forms a kind of supplement to the Taittiriya Aranyaka. Hence it is only too natural that the Brahmaņas in general should be of earlier date than our text. Thus testimony is borne to the soundness of the method, whereas the chronological question is furthered by the fact that the MNU, shows a younger type of verse than the Katha Upanishad. The priority of the latter has already been stated tentatively10 on other grounds, here we have got a metrical proof for the same.
It is strange that in our book the caesura does not occur as commonly after the fourth syllable as it does after the fifth, the figures being 17 to 25. This is the inverse proportion of the texts compared above where the figures, representing the average, are 262 to 123. There is a point, then, where the modernising tendency of metre has not influenced the writer of the MNU. to an equal extent as it has the authors of other books of the same period. "
A date far remote from the time when the correct verse of the Rigveda had been composed is pointed to by the careless handling of metre in many of the pádas. Catalexis or hypercatalexis or even faulty prosody occurs in 1, lc, 3b, 6a, c (only in MS. A correct); in 13.2 : fani ya UT #rao; in some of the pâdas of 16. 4; in 22. 1 TTA ET!'; la ufag. Most of these verses could be mended by means of but slight changes. The metrical defects of 1. 6a ( instead of wo) and 22. 1 (an F : instead of agen : ) may be due to unphonetic spelling.20 The part that appears least injured is the beginning of the line, the iambic-spondaic cadence there prevailing being kept throughout, except in 13. 2 a 8;° 22. 1 aai To, a um: 23.1 Reao. It has been pointed out above that in six cases original in the MNU. the caesura is after the sixth syllable. Hence it cuts the Vedio anapaest or the classic dacty! which is or at least ought to be, formed by the 5th, 6th, and 7th syllables. Here are the instances : 1. lc, 2d, 3c, (40 only in MS. E wrong) ; 40, 50 (?), 22, 1 agarrara. But worst of all, there are currupt lines in our text which must have been spoiled by the compiler of the MNU. himself, since they appear in the souroes from which they are taken in their correct form. This certainly proves " the great and universal confusion by which the prosody of this period is characterised." Thus in 2. lo, 10. 7a the caesura is found after the sixth syllable being removed from its proper place in the original; 2. la has obliterated the iambic cadence at the beginning; 2. 3a, 4cd, 5a, 6b are also prosodically corrupt. For these blunders we can, indeed, make only the clumsiness of the author of the MNU. responsible, as the text is otherwise in comparatively good condition and, on the part of the author, no definite plan of these changes is discernible.
11. The lines of the Anushubh-Gayatrt family. There are about 110 anushtubh and gayatri lines to be considered.21 Neither the repetitions of former passages, nor, on the whole, stray verses and padas, nor borrowings from other texts have been included. Thus we omitted 17. 4 being a repetition of 3.2;
19 Cf. Die Quellen der Mahanarayana Up. p. 40. 20 cl, Aluindische Grammatik von Jakob Wackernagel, I § 28 78a and c. with note.
21 The affinity of the anushtubh and gayatri line will justify their joint treatment. The first two nadar of the glyatri are prosodically identical with those of the anushţubh. Moreover, oddas one and two are, wait were, enjambed, whereas pddae two and three are, so to say, end-stopped. Finally in neither type of verse is the second part of the line of such a nature as to necessitate a definite shape of the first hall line. Hence we may safely consider the first half line (two poddas) independently of the rest of the line (pdda three or pddas three and four) of both verso types, though these be of different length. In fact there are gayatris to be found only in the third khanda; the ronson why this line is used there see below, note 24.