1872.] Hoernle — Essays on the Oaurian Languages. 125 
The form of the word to which the post-positions are added, is not 
always the pure hase, as the addition of the Sanskrit affix would require, but 
already inflected (e. //., ilT% hi of a horse, is not the simple base of the 
word ifFfr, but an infleeted form of it). Again, it is most unlikely that 
elements like the affixes 3f, cjtol, etc., which occur in Sanskrit only as integral 
parts of a word, but never by themselves as independent words, should have, 
in a comparatively modem language, separated themselves from the body of 
the word and assumed independent life (as post-positions) similar to that of 
prepositions. It would be a phenomenon contrary to those that have been 
observed in all other cases of (what Max Muller calls) dialectic regeneration. 
It is clear also that by this theory the other post-positions (as H, *j, %) 
cannot be explained. But there can be no doubt that, whatever the true 
explanation be, it must be the same for all post-positions. For these reasons 
among others, any theory which traces the post-positions to Sanskrit affixes 
cannot be the true one. Their explanation must be sought for in a different 
direction. 
In the first place, it may be remarked that the term “ post-position” is 
misleading. It gives the idea as if the words, to which it is applied, belonged 
to that class of words which includes the prepositions, conjunctions, etc., i. e., 
elements of language which are incapable of either derivation or inflexion. 
Now most of the so-called post-positions of the genitive are capable of both. 
They have clearly a nominal or more accurately an adjective character. For 
the Hindi ^rr, #7, agree with the noun which they qualify, in case, 
number, and gender, exactly as for instance, ^ good. If 
the qualified noun is a masculine singular nominative, then «RT is used ; if a 
feminine singular or plural, then cRl ; if a masculine plural nominative, then 
% ; if a masculine in any oblique case, then The same is the case with the 
Panjabi 27 , the Braj Bhaslia 3RT, %, etc. In the Sindhi, 5H and oil 
are used like 3fT and in High Hindi ; 5TT is used, if the qualified noun is 
in the masculine plural nominative, and si is used, if it is a masculine in any 
oblique case singular or plural. With this agrees the use of the post- 
positions ir, eft, of the Alwar and Jaipur dialect. 
After these explanations the following scheme of the agreement* of the 
* This agreement is not altogether perfect ; bnt neither is the agreement of 
the real adjective with its substantive more so; e. g., in Hindi, qyj and % ought to 
have a different form in the plural ; probably these plural forms are irregularly adopted 
from the singular ; but then g° 0< h has also both in the singular and plural 
. Originally, the agreement was much more perfect. This is clearly proved by the 
Marathi, where “ sometimes, in poetry, the adjective takes a case-form corresponding 
to that of the noun it qualifies ; thus : 
HTW «T HT5PJ ” 
Student’s Manual of Marathi Grammar, p. 39. 
