not accept the suggestion of Caldwell, as he thought that a Dravidian language could not possibly influence the Aryan speeches in that manner. Trumpp suggested that কে of Bengali came from কৃতে and Beams rightly rejected the derivation, as কৃতে could not signify the sense conveyed by কে. Beams himself, however was wrong, when he sought to derive the suffix denoting the dative from old Hindi কহঁ. Sir R. G. Bhandarkar showed that as in no Prākṛta, either কৃতে or কহঁ (derived from কক্ষ according to Mr. Beams) signified any dative sense, the proposed derivation could not be accepted. Sir R. G. Bhandarkar is right that for many case-denoting suffixes we have to look to pronouns and pronominal roots, but his imaginary case that কেহিঁ as well as তেহিঁ might have been in use to signify instrumentality, and কেহিঁ might have been subsequently used to denote a dative case cannot be accepted, or rather may be easily rejected, by using the very argument with which the learned scholar himself has rejected the theory of Mr. Beams. Sir R. G. Bhandarkar's suggestion that কহঁ in a phrase as রাম কহঁ might mean at first "Rama's somewhere," and thence the sense "to give to Rama," might have originated, is very faulty as the old time forms do not warrant such transformation. The derivation would not have been sought in such a roundabout manner, if the cause of such changes as ধম্ম, সঙ্কপ্প and সিলোকো could then be rightly detected.
How the Dravidian people could influence the speakers of the Aryan speeches in dim past, should be a subject of special research. Many ethnological problems, relating to the Dravidians, have not yet been solved. The ethnologists of our time agree in the main, that the Dravidians have been autochthonous in India; even though this proposition is not free from doubts and difficulties, the situation of the Dravidians in India as neighbours of the Aryans, since
