We are therefore driven to the conclusion that the Dravidian which exercised the main influence on Sanskrit was a Northern Dravidian. There is no difficulty in this assumption, since certain islands of Dravidian still exist in the North. These are Kurukh and Malto, and in the far west, the Brahui of Baluchistan. Of course Dravidian on such a small scale, as it exists now in the North, could not have exercised any serious influence. So we must conclude that Dravidian was at one time extensively current in the North and very likely also (in order to account for its influence on Sanskrit) that it had been current in some of the main areas of Aryan settlement.

A consideration of the type of words adopted by Indo-Aryan leads to certain conclusions. We have on the one hand, for instance, such words as the names of Indian trees for which the Indo-Aryans cannot have had any words to begin with, and some of which one would therefore expect to find borrowed from Dravidian. Such are ketaka (pandanus odoratissimus), tāla (palmyra palm), and so forth. But in addition to these, we find a considerable number which are 'unnecessary', inasmuch as Indo-Aryan was perfectly well supplied with words having these meanings already. Such are the above mentioned nīra (Aryan : udaka, āpas) and mīna (Aryan : matsya). Similar cases are kānana, forest (Aryan : vana): kāca, carrying yoke (Aryan : vīvadha); kuntala, hair (Aryan : keśa); and so forth. It is difficult to see how such words could have become current in Sanskrit, unless at one period there was a considerable amount of bilingualism in North India. The unnecessary adoption of, for example, Dravidian nīra is reminiscent of the Anglo-Indian slang in which, for instance, 'pawny' would commonly be substituted for the English word 'water'. Such usage was particularly common among the lower ranks of the army and corresponding grades elsewhere, but largely avoided in official English speaking and writing in India. Much the same was obviously the case in the later Vedic period in India. The official language consciously avoided the use of mleccha terms, though not with complete success, since they were not always detected. On the other hand, they must have had considerable currency in the language of the lower classes during the bilingual period. In course of time, a considerable number were elevated to the status of literary words, and as such we meet them in the epics.

Another point that would be worth special investigation is these loan-words in modern Indo-Aryan. It is interesting to note that such words as kānana, kuntala, mīna, nīra are not preserved in modern Indo-Aryan, which preserved only the words of Indo-Aryan origin (Hindi : ban, keś, machlī) or in the last instance, where there is excessive reduction in Prakrit (uaa), by a new Indo-Aryan formation.