262 DE, J. MUm'S ACCOUNT OF THE 



that early age a learned and sacred language should have existed alongside of 

 the current vernacular dialect. In order to fulfil their intended purpose, as ex- 

 pressions of the devout sentiments of the worshippers who used them, the hymns 

 must have been intelligible not only to the priests and chiefs, but also to the 

 people in general. And, secondly, it is only by supposing that the Sanskrit of 

 the hymns was a spoken idiom that we can account for the gradual modifications 

 which it afterwards underwent. At length, however (at a period which cannot 

 now be settled with any precision, but which must l)e placed several centuries 

 before the beginning of our era), Sanskrit ceased to be spoken, and it continued 

 thenceforward fixed in the form which it had at that period taken, and in which 

 it was exhibited in the standard works of that age. The vernacular dialect by 

 which it was at first succeeded, arose, no doubt, gradually, out of populai- modes 

 of pronunciation, and differed little from the original language, of which it pre- 

 served, with some modifications, the principal forms of declension and conjugation. 

 This we see in the earliest Indian vernacular idiom which has come down to us, 

 the Pali, which, as it was the prevailing form of speech at the time when the 

 founder of the Buddhist religion flourished, and was employed by him for the 

 propagation of his tenets among his countrymen, has been preserved in the sacred 

 books of that religion discovered in Ceylon. This dialect differs from Sanskrit 

 very much in the same way as Italian differs from Latin, and many of the com- 

 plex sounds of the Sanskrit are softened down in Pali in precisely the same 

 manner as similar sounds in Latin are modified in Italian. Thus the Sanskrit 

 words muktas, hhuktas, suptas, guptas, become in Pali mutto, hhutto, siUto, and 

 gutto, just as the Latin jjerfectus, fructus, r-uptus, actus, become in Italian perfetto, 

 frutto, rotto, atto ; and so on in a number of other parallel cases. The Pali was 

 followed by various provincial dialects, called Prakrits, which continued to diverge 

 more and more from the Sanskrit, and from one another, — gradually dropping the 

 modes of inflection proper to the parent tongue, and supplanting them by separate 

 connecting particles, — till at length they issued in the modern vernacular idioms 

 known as the Hindi, Bengali, Gujarati, Marathi, &c. The languages of the south 

 of India — Tamil, Telugu, Canarese, Malayalim — belong to a quite distinct family. 

 They are not, like the north Indian vernacular tongues, derived from Sanskrit, but 

 are (especially the most pure and perfect of them, the Tamil) completely inde- 



course of publication by Messrs Triibner & Co. See also the Quarterly Review, vol. liv. p. 299. 

 On this theory I may shortly remark, \st, That unless the views of all the most competent scholars, 

 such as CoLEBROOKE, WiLsoN, Max Muller, Goldstucker, Aufrecht, &c., in regard to the age of 

 the Vedic hymns, are grossly erroneous, the Sanskrit language must have existed for at least eight 

 hundred or a thousand years before Alexander invaded India ; and this language, as it existed at 

 that early per\od, exhibits, in quite as distinct a manner, all those affinities in roots and structure, 

 with Greek and Latin, by which it was distinguished at a later period. 2dly, That the vernacular 

 language — which, according to the best evidence we can obtain, existed at, or even before, the time of 

 Alexander (I mean the Pali) — in its structure presupposes the prior existence of Sanskrit quite as 

 necessarily as the forms of the Italian presuppose a Latin language out of which it arose. 



