________________
174
THE INDIAN ANTIQUARY
(AUGUST, 1917
letters as gômútrikábandha, sarvatóbhadra, &c. Chapter 343 dilates upon the figures of xense and 344 on tho figuros of both word and sense. In chapter 345, seven gunas of Poetry are spoken of, and in chapter 346 the blemishes of Poetry are dealt with.
The evidence for arriving at the conclusion that the Agnipurâna is not the most original work on Alankâra literature is both internal and external.
The internal evidence may be stated as follows:
I. We have some indications in the Agnipurina itself showing that it was not Bharata who copied from it, but rather the reverse. The Agnipurana says that the riti styled Bharati was so called because it was first promulgated by Bharata. In the Na!yasastra of Bharata we are told that the four vrittis Bharati, Sattvati, Kaibiki and Árabhati were received by Bharata from Brahmå and that Bharati Vritti was named after the Bharatas. From the above it is clear that the Agnipurâna knew that Bharata was the originator of the Natyasdstra (or at least of the Vittis that form a very integral part of it) and that perhaps it had before it the very words of Bharata quoted by us above. Another noteworthy fact in this connection is that Bharata nowhere ailudes to the Agnipurâna in the extant Natyakâstra, although he shows an acquaintance with works of the Purana class. 4
II. The very nature of the contents of the Agnipurana precludes the idea that it is an ancient and original work. Even a cursory examination reveals the fact that the Agnipurana is a professed conglomeration of heterogeneous material borrowed from many sources, especially in that part of it which deals with the various branches of Sanskrit literature. On the other hand the Natyasastra appears to be a very original work. Bharata speaks of only four figures of speech, while the Agnipurára mentions a large number. If Bharata had the Agnipuraya before him or if he had known more than four well-recognized figures of speech, he would have given a full exposition of them and would not have been held back by considerations of irrelevancy and prolixity. He defines and illustrates about a hundred different metres, which have as much connection with the dramatic art as figures of speech.
III. We shall later on adduce evidence to show that the Natyasastra of Bharata must be at all events earlier than A.D. 500. From an examination of the contents of the Agnipurana, it follows that it was put together later than A.D. 700 or even A.D. 1000. Our reasons are
(a) The Agnipurâņa refers to the seven Kandas of the Ramayana, the Harivariáa, to Pingala, Palakäpya, Salihotra, Dhanvantari and Susruta. It gives a short summary of the Bhagvaldgita in chapter 380, in which halves of verses occurring in different chapters of the Gita have been combined in one verse. One of the most significant facts for our purposes is that the Agnipurana borrows from the Amarakósa in chapters 359-366. Almost all the verses are directly taken from the Amarakófa or are formed by taking half verses from the Kosa and then piecing them together. If Amara borrowed at all he would do so
2 Bharatena pranftatvad-Bharati rtiruchyate.-Agnipurdna 339.6.
3 Maya Iduya-kriya heton prakshipid drwin-djfayd.-Natyasdatra 20-23 ; again at 20-26 we read Sva-námadheyair-bharatai prayukta ad Bharati ndma bhavel.tu vrittih it
Anyepi des& ebhyo ye Purdạe samprakirtitah teshus prayujyate hy-esh praprittis-to-Audr Magadh.--Na!ya. 13-35.
Kavyasy-aite hy-alankdrdf-chutedraḥ pari-krtitah-Natya. 16-4,
6 Agni 380.12 is the same as Ou VI. 40 and VII. 14. Na af kalydnabrit kafohid durgatim lata gachchhati I Dair hy-esha gunamayf mama mdyd duratyayd 11