________________
DECEMBER, 1903.]
NOTES ON DRAVIDIAN PHILOLOGY.
458
On the other hand, there are several compound tenses, as is also the case in other connected languages. Thus, the imperfect kindán, I was doing, is formed from the participle kié, by adding andán, I was, Another form of the same tense is kid mattond, I was doing. The pluperfect kioi mdttond, I had done, literally means having done I was,' and so forth.
We have thus seen that the formation of tenses in Gondi is essentially the same as in other Dravidian languages, and that there cannot, therefore, in that respect be any question about an influence exercised by Santali.
It has often been stated that the negative verb in Göpdi is formed by inserting hille or halle between the pronoun and the verb. This use of hille or halle does not, however, appear to be more than a tendency, and I have over and over again found forms such as sóvór, he did not gire, without the addition of any separate negative particle. It will thus be seen that Gôndi in all such essential points agrees with other Dravidian languages, and there is no philological reason for separating it as . Borthern group of Dravidian languages, as has sometimes been done.
It has already been pointed out in the preceding pages that the third person singular of most Dravidian tenses in form does not differ from a composite noun or noon of agency. Thas Gôndi kitur, he did, seems to be formed from kitu, corresponding to Tamil seydu, having done, by adding the suffix of the demonstrative prononn. Compare Tamil seydan, he did. The other persons of ordinary tenses are not, however, formed in the same way as in the case of the Gondi present by adding the personal suffixes to the base of the third person, but by sabstitating the suffixes of the first and second persons for that of the third. Thus Godt kitan, I did ; kiti, thou didst. The forms of those suffixes vary in the different Dravidian languages. The reason for this state of affairs seems to be that the fall forms of the personal pronouns have been changed in varions ways, and the snflizes have not always undergone the same changes. Thus the pronoun thon' in Telugu is nivu, but the pronominal suffix of the same person is simply du or pi, where all traces of the original pronominal base have disappeared. On the other hand in Gôndithou' is immi; but the corresponding saffix of the second person is i, probably the oldest form of the Dravidian pronoun for thou.' It is quite natural that the same suffix can, under such circumstances, come to be used for more than one person. Compare Telugu chésiná-nu, I did; chése-nu, he, sbe, it, or they, did, where the same suffix nu is apparently used for the first as well as for the third persons. Prof. A. Ludwig has mentioned several similar instances from Telugu, Tamil, and Kanarese, and has drawn the conclusion that the personal terminations of the Dravidian verb are not originally personal pronouns, but that there is only, at the utmost, an intended assimilation of the sound of the termination of the verbal tenses to the sound of the personal pronouns. See bis paper Uber die Verbalflerion der Dravidasprachen. Sitzungsberichte der königl. böhmischen Gesellschaft der Wissenschaften, Olasse für Philosophie, Geschichte und Philologie, 1900, No. VI.
Professor Ludwig is certainly right in assuming an assimilation in sound between verb and pronoun. Compare Telugu nenu châsinánu, I did ; niru chesindvu, thou didst; vádu chêsinddu, be did; Tamil nán beydán, I did; ni seyda-y, thou didst; avan beydán, he did; Kui éanju gitenju, he did. If we compare Telugu rádku chásinadu, Tamil apan seydán, Kui éanju gitenju, it is evident that the third person singular masculine of the verb has in all cases undergone the same phonetical changes as the corresponding pronoun. In such cases as Telugu niru chê sindru, thou didst, where the base of the pronoun is ni, and the corresponding verbal suffix only is a secondary termination, it is evident that the assimilation in sound has been intended,
There are, on the other hand, many cases in which the verbal forms have not been changed in the same way as the pronouns. A good instance is furnished by Gôndi. Compare nannd kidton-á, I do; nannd kitá (-n), I did ; immá kití, thou didat; mammd! kitom, we did ; immdi kitir, you did. It will be seen that the Gôndt pronouns have undergone great changes, while the corresponding suffixes have retained an older form.
