am constrained to say that the learned author has built a stupendous structure with very weak materials, on the foundation of a fancy of his own. Sir Herbert Risley has very rightly remarked that without at all resorting to the theory of patois-speaking hordes, the changes in the dialects of Central India can easily be explained, with reference to the people speaking them. Need I make a statement of the well-known truth, that it is "grammar" and not "sound" or "vocabulary" which gives a dialect its character? Merely because some tribes of the Punjab frontier use some words of Aryan origin, Grierson concludes that these tribes are remnants in hilly countries of the oldest Aryan people. It is on the evidence of sound and vocabulary he has thought out different origins for some dialects of Northern India. It is such reckless assumptions that have brought philology into disrepute with all anthropologists, though philology as a branch of knowledge has a useful sphere of its own. Merely from similarity of sounds Grierson has inferred that the "Pakthas" of the Rigveda are the modern Pāthāns without caring even to ascertain if those, who are now called Pāthāns, existed in the Vedic days with such a tribal name. I would not have wondered if the Āfridis, who, to serve the convenience of a theory, may be called Āpridis, were similarly put forward as the authors of the Āpri hymns. I cannot bring myself to imagine that Grierson, who is widely known to be a great oriental scholar, has made his authoritative statements regarding our archaic and classical languages without possessing sufficient knowledge of them. But, on the other hand, it appears so strange that one having even a very common acquaintance with the languages of old India could for a moment think that Sanskrit is the polished form of the Vedic language. The grammarians, who have been given
