1S73-] A. F. Rudolf Hoernle— Essays on the Gaurian Languages. 93 
to a form in But if this is the case, the Hindi corresponding 
termination ^ must also be a contraction of an original termination ^T, and 
not And further it follows, that both in Marathi and Hindi, the 
Prakrit base from which this oblique form in and y is derived, must have 
ended in 
3. In Marathi there is one exception to the rule that the initial 
consonant of the obi. form termination yq is compounded with the final 
consonant of the base. It is the gerund in According to both the 
Manual (see § iii, III.) andDadoba’s Grammar (see §. 463.) the oblique form 
of these gerunds does not end (as we should expect according to the analogy 
of other neuter nouns in y, as [obi. form ^ywr], ^iF[obl. ^^jrJ) in 
UT3JT, but in e. g., T to do, obi. form ^yR£fT (not cSvCRJT), 
to go, obi. form ^TT^'T (not ^TRIT). Here the alternative form in 3JT does 
not exist at all. Now this exception proves the rule extremely well. It 
has been observed several times already that these Gaurian gerunds or 
infinitives in are derived from the Skr. and Prak. Part. Put. Pass, in 
rT^T, and it has been shown in a previous place, that the Sanskrit termination 
rfeST may become in Prak. ^* 7 ; thus Skr. becomes Prak. (^ify^rNf or) 
or ^FTRERsi. The genitive of the latter form is efiy^^yg - , which 
changes to ^y^Tyr or or s^y^j^Rr. Here the form passes into 
Gaurian which, according to its law, contracts the form, by Sandhi of the 
hiatus-vowel, into ^fyRRT ; and thus we obtain the present Marathi 
oblique form. Now let it be noted that here the semivowel is not a 
euphonic insertion of the Gaurian, but an original, integral part of the word, 
taken over from the Prakrit. The case would be very different with any 
other neuter nouns, as e. g., fhigh. In Prakrit this neuter would be 
which in Gaurian would become ; the genitive of the Prakrit 
would be or or or in which last form it 
passes into Gaurian, and now if we are to obtain the form we must 
assume that the Gaurian inserts a euphonic yp This, as we have seen, is 
not the case ; the Gaurian, on the contrary, makes Sandhi under these 
circumstances ; we should obtain the form We see, therefore, that the 
reason why the oblique form of the Gerund in y differs from the oblique form 
of other neuters in y, is this, that the consonant \ of the former is organic, 
while the ^ of the other neuters would be an inorganic euphonic insertion. 
But, as I have proved by examples from the Gujarati, Naipali and Marwari, 
it is contrary to the habit of Gaurian to insert y[in this particular case ; it 
prefers to make Sandhi. Hence the difference under discussion proves, that 
the oblique form in must be explained in an altogether different way, and 
the' theory advanced by me, that it is the modified genitive form of a 
Prakrit base in fulfils all the requirements of the case. 
4. The oblique form in Rr is not altogether peculiar to Marathi neuter 
nouns in f, but it belongs also to the Marathi neuter nouns in Now 
