1872.] Hoernle — JEssays on the Gaurian Languages. 153 
Ganwari and in poetical and old Hindi, the original forms still commonly 
occur, e. g., 
sr? WlfJT I 
?rq -UTK ^nfjt II i. e. 
H. H. 5r?t srm htji qr i 
rr^t % ^tjt ii 
Lanka Kand. 
Or srsiffr ^TfJI Wlf^Nrl II i- e. 
H. H. STtsral ^TJI II 
Prithiraj Rayasa, 1. 18. 
Again faff WrT HTft II i. e. 
H. H. u * ft ^i’n Pfrni % wet ii 
Ayodhya Kand. 
Again ^THT jpv’nsrqe" WKW II *'• e. 
^Tq >4UT;i5T8^ tTH #T »TT*t II 
Ayodhya Kand. 
Again fqrrf% wilST ^nft II i. e. 
H. H. fqrTT €r ^<OTT ^oa|T ^HiW] ^Tlft II 
Lanka Kand. 
But also ftf?j nq spfiqr ^'qrr || i. e. 
H. H. fq<ri qrr h?i vh qit tlqnr % n 
Ayodliya Kand. 
In Marathi, where ^TJT and similar words are mutilated in the same 
Way (see below), the original r appears again in the oblique eases ; e. g., 
the gen. is ^TJll ’qT. just as nom. qfe. gen. qrit "qT. 
An interesting question here arises : why is it that the Prakrit termin- 
ation has not always become ^ in Gaurian, hut has remained unchanged 
(or become WT as in Hindi, Panjabi and Marathi) in many instances ? e. g., 
while the Braj BhasM forms the Pres. Partic. %T3 being (= Prakrit %m) 
the High Hindi has qr<TT, or while the Prakrit becomes in old Hindi 
the Prakrit UT^yT, horse, becomes in modern Hindi ii^q (also in 
Panjabi, Marathi, or f?T%T in Sindhi, Gujarati). If my previous remarks be 
remembered and also that % 7 J, U*i3, etc., admit of no oblique form , while 
%T<TT, iu%T, etc., do admit of one (viz. 3i%, etc.), it will be seen that the 
forms %Trj. qws, etc., belong to the proper Gaurian element, while qirTT, qlfT, 
etc., belong to the Priihritic element. This, however, is not yet an answer 
to the question. The question still remains why did some Prakrit words 
ending in qjy not submit to the Gaurian principles, but retained their 
In modem Hindi the form (Rao) is limited to being a certain title of 
nobility less than Raja, exactly as it is also the case with the form (mentioned 
above in the text) , a perfectly parallel formation to 
