92 A, F. Rudolf Hoernle— Essays on the Cfaurian Languages. [No. 1, 
2. A second reason is this. To the Hindi oblique form in ^ the 
Marathi oblique form in ^|T corresponds and both must have an identical 
derivation. Now though ^ may be explained as a contraction of in 
Hindi, this cannot be done with Marathi ^T. In Marathi the initial 
consonant of the syllable ^TT is always compounded with the final 
consonant of the base. There does not seem to be any trace that it may 
be separated from the final consonant of the base, and pronounced as ; 
e. g., the oblique form of ifTfT horse is ijrfJT, but not ifT^f3?T. In the case of 
the oblique form in TT, the Manual admits an alternative form in ^TT; e. g. 
ship, obi. form or cTTTffi ; but in the case of the oblique form in 
I neither the Manual , nor Hadobas’s Grammar, nor any other grammar that 
I have consulted, admits an alternative form in If it had existed at 
all, it would surely have been mentioned by one or other of the gram¬ 
mars. Even the alternative form is doubtful, seeing that it is only 
mentioned by the Manual ; but the alternative it appears, does not 
exist at all. Now this fact would be very improbable on the supposition 
that the form in is the original one, out of which the other (the present) 
form in "5JT arose by the suppression of the medial Such a suppression 
of a medial % indeed, is not uncommon in Gaurian ; but whenever it occurs* 
both forms remain equally current, the original one without the ^suppression 
and the derived one with the suppression ; and at all events, whatever the 
pronunciation may be, the spelling wherever accuracy is observed, follows the 
origin of the word. Thus in Hindi, though he knows is pronounced jdntd it is 
always by correct Nagari writers spelled jdnatd [i. e., ^TT«Trf7, not ojr^T). Now 
neither of these is the case with the Marathi oblique form in ; it is always 
spelled with the ^ compounded with the preceding consonant, and always so 
pronounced. Even if we should rely on the analogy of the oblique form in 
qq, it would not help us out of the difficulty. For, as I have shown formerly 
when treating of the Marathi neuter nouns in gf, the case is just the reverse 
with the obi. form in TT- There the original form is that in and the 
derived form is that in ^T, i . e., with the insertion of a euphonic ^ to 
prevent the necessity of pronouncing a compound consonant; such insertion 
being also not uncommon in Gaurian. If, therefore, the analogy of the 
oblique form in proves anything, it proves the very thing demanded by 
my theory ; viz. that the form in is the original form ; and if a form in 
should exist, it could only be a vulgar corruption of the form in *rr 
with inserted Further, it should also be noted, that even if two 
alternative forms in and should exist, this fact, though it might 
allow the opposite theory, would in no way contradict my theory; (for the 
form in ^r, as just shown, might be the original one) ; while if only one 
form in exists, this fact is altogether fatal to the opposite theory, but 
accords entirely with my theory. It seems certain, then, that, at all events in 
Marathi, the termination of the obi. form is original, and not reducible 
