example, the synonym বা or বার for আপ, to denote streaming or flowing, can be very naturally coined, but such synonyms as নীর, তোয়, or জল, may be suspected to be of foreign origin. If the word জল has come out of গল (trickle out), the existence of it may be justified, but the two other words নীর and তোয় which cannot be connected with Sanskrit roots naturally arouse our suspicion regarding their origin. The Dravidian root রু which we get in অ-রু (tear-drop), আ-রু (river) is in the Tamil word নীর of which নীরু or নীলু is a Telegu variant; we have to notice along with it that, the word নীর does not occur in the Vedic language; we are therefore justified to hold that the word was introduced in Sanskrit from the Dravidian source. As to তোয় unknown in early times, we notice that an aboriginal tribe of Tippera use the word for water, and tui to signify water occurs in many dialects spoken in and near Manipur. T. C. Hodson shows (J. R. A. S., 1914, pp. 148-50) that this 'tui' is connected with Chinese 'sui.' The word in question may therefore be presumed to have come from the Kirāta source. The word kuri for twenty occurs in some Mongolian speeches in the Himalayan region; this word may be presumed to be identical with our কুড়ি.
We notice in this connection another phenomenon of equal importance. Some Sanskrit words naturalized by the Dravidians, in their Dravidian method, are found retaken in Sanskrit as new words, unconnected with their original forms; for instance, দম্ভ, reduced to i-ṭampau, has come again as āṭopa (cf. সাটোপম্ পরিক্রমতি) in Sanskrit, and the Dravidian চেত্ত a derivative of শব্দ, appears in Sanskrit in the form চিৎ as in চিৎকার . It is very interesting to note that some Sanskrit derivatives, in Bengali, disclose this fact that we have reduced some Sanskrit forms to Bengali, exactly in the manner in which the Dravidians do;
