It was not until 1925 that any further attempt was made to develop the thesis propounded by Caldwell. This was an article by F. O. Schrader entitled "Dravidisch and Uralisch".¹⁰ In returning to Caldwell's theory Schrader considerably simplified its terms of reference, in as much as he left out of account all question of relationship with the Altaic languages and others included by Caldwell under the term "Scythian", and confined himself to a comparison of the Dravidian languages on the one hand and the Uralian languages, i.e. Finno-ugrian and Samoyede on the other. As to the exact nature of the relationship of the two groups of languages he did not take up a very clear position as is clear from his own words: "Ich will zu zeigen versuchen, das zwischen der dravidischen Familie einerseits und der uralischen, d.h. der finnisch-ugrisch-samojedischen, anderseits ein historischer Zusammenhang besteht, der wenn nicht als Urverwandtschaft so durch vorgeschichtliche Nach-barschaft und sehr intime einseitige oder gegenseitige Beeinflus-sung erklärt werden muss." ¹¹

Schrader bases his argument on phonetic structure, word formation, and correspondences in grammar and vocabulary. In presenting this material he adds considerably to the somewhat slender evidence produced by Caldwell, and, as might be expected from the advances in linguistic science in the intervening period, the material is on the whole more reliable. This does not mean that the percentage of error is negligible; on the contrary it is quite easy to pick out false equations page by page. Nevertheless, even allowing for all this, there remains enough material against which no obvious objections can be raised, to make the subject worthy of renewed serious attention.

In effect little notice was taken of Schrader's article. His line of argument was attacked by E. Lewy,¹² who maintained that all the facts adduced by him could be explained by the normal workings of chance and that an equal number of correspondences could be found between any two unrelated languages or language-groups in the world. In support of this Lewy produced a list of sixty Kechuan words similar in form, and meaning to sixty Finnish words which he placed beside them. The answer to this argument is that the correspondences in vocabulary between Dravidian and Uralian are far more numerous than the list of sixty which Schrader had given and Lewy parodied. It is possible to produce a longer list of words dealing

------------------------------------------------------------------------------------------------------------------

¹⁰Zeitschrift für Indologie und Iranistik, iii, pp. 81-112.

¹¹Op. cit, p. 83.

¹² In the course of a review of W. Schmidt's Sprachfamilien und Sprachenkreise der Erde in Zeitschrift für vergleichende Sprachforschung, lvi (128), pp. 142-159.