of gender in Sanskrit; soma, the 'moon' is masculine, ruchi, 'taste' is feminine, and putran,' son' is neuter. It will thus be seen that gender in Sanskrit depends on the peculiar structure of words, but not on the sex or the intelligence of the objects expressed by them.
(3) Tamil nouns are infected not by means of case terminations, but by means of suffixed postpositions and separate particles. The inflectional base in the oblique cases is the root in Sanskrit, while in Tamil it is the nominative, except the first and second personal pronouns நான், நீ, நீர் and நீவிர் which alone change their forms. For example, in Sanskrit the roots vach (speech) and raj (king) become vak and rat in the first or nominative case, while in Tamil the roots சொல் (word) and கல் (stone) remain the same. In declining nouns the same case terminations are added to the root for the singular and to the plural terminations for the plural (e.g., கல்லை, கற்களை). But in Sanskrit and other Indo-Germanic languages, the case endings of the plural differ from those of the singular. As Dr. Caldwell rightly observes.-‘the imitation of Sanskrit was certainly an error, for whilst in Sanskrit there are eight cases only, the number of cases in Tamil, Telugu, &c., is almost indefinite,' being limited only by the number of postpositions that may be attached to the noun. And it is this indefiniteness that has given an endless trouble to the Tamil grammarian Tolkapyar, who has devoted three complete chapters for cases only ; and these have been supplemented by another by the author of Ilakkanakkottu.
