1G5 
1872.] Hoernle — Essays on the Gaurian Languages. 
small, in Hindi it is and stands, therefore for an original Prakrit wreXT 
or WPMt. The equivalent for the latter in Hang.il i would be i~rzx (or xrxx 
i- e. the crude base) and the gen. of it XT^X t ft or WTXX + XX or (elid- 
ing the final X in Sandhi with VX) XIXXT or XT«vy; and now the initial 
T of the enclitic yy after the final X inherent in WTX is elided just as U 1 XT -f- 
X instead of XXT + V, hence we have XTXX chhotar (or cliliotor). There 
is one exceptional gen form in Bang&li, which proves and illustrates well the 
process by which xtxx and such genitives w T ere formed. I refer to the gen. 
of the pronominal adjectives VX or XX or XX so many, XX as many, XX how 
many, which have a double form (see Beng. grammar of Samaehum Sircar 
P- 85), either V 4 V, XWX, XXX, *JHX, XdV 1 or VXXX, ’XXXX, XXXV, XXXX, 
XH^r. From the way Samachurn spells the latter forms {viz., appending a 
virama to y?(, XX. etc , and thus making them terminate with a consonant), 
it would appear that he considers the whole of XX to be the sign of the 
gen., and the base to be only vx, XX. XX, XX, XX. If this be correct the 
form %y confirms my theory that the gen. - sign XX or X is but a curtailment 
of an original XX (the remnant of the Prakrit x r X). But I am inclined to 
think that the pronunciation of fXXX XHXX, etc. as et-kcr, at-ker, etc., instead 
of etaker, ataker, etc., is only a vulgar corruption, and that the words XXXX 
XXXX, etc. ought really to be divided into VXX 4- XX. XXX + IX, XXX + 
Xr, jjrTEfr 4 yy, XToX + XX, so that the base is really XXX, Xnx, etc., and 
the gen. sign XX My reason is this : the corresponding forms in Hindi are 
?XXT or TXT so many.fsfXXT or fxXT as many, fxXWT orfxXT how many, with 
the oblique forms resp. TXX or TX, fxxx or fxx, fxxx or fxx. These words, 
having a,. direct form in XT and an oblique form, in x, belong, according to 
the ordinary principles of Hindi, to the Pr&kritic element and are derived 
form the particular Prakrit base in X. They presuppose, therefore, a Prakrit 
original yxx or yfxx, fxxx or fsrfxx, fxxx or fxfxx. As a matter of 
fact, these forms or, at least, forms almost identical (see Pr. Pralc. IV, 
25.) occur in Prakrit ; viz., either xfxx, xfxx, xfxx, xfxx, (for xfxx, 
Xfxx, xfRX, Xfxx), or VXX XTX. X^X (in which T stands for 
X and x for X, see Pr. Prakrit II, 4, xf%iT for Sanskrit Wifxx: ; hence == 
Vtjx, XXX, XXX, XXX). Here, on the one hand, the Prakrit X lias been 
reduced in Hindi to X (thus xfxx or xfxx for xfxx or xfflX) ;_on the 
other hand, in Bangali, it has been reduced to X (thus XXX for XXX or 
XXX), and besides the double consonant X is reduced to one X.* Thus we 
* Theso Bangali forms yj,yr, XXX, XXX, throw light on the origin of another 
Gaurian form ; via., that of the Hindi participle present in x' or 47 (as =^^rx or 
XXXT going). The Sanskrit affixes ■yjx , XX. XX become in Prakrit regularly 
XX, XX, XX ; of. Pr. Prak. IV, 25. VII. 10 ; thus Skr. ^HXX rich is Pr. XHXX, 
Skr. tfg ff reading is Pr. xr^vfl. Similarly Skr. fqr«(X, XXX, XTXX oll g ht t0 be iu 
