114 § 152-153. sitions. Prepo- positions," as they are generally put behind the nouns, they are construed with, I being the one, that is always put before. As a rule, they are also allowed. to be compounded with their nouns; in that case, the preposition is generally the former member '). 153. The archaic dialect used more prepositions and used them oftener, than the classic language does. The more we go back in time, the greater the number and the variety of idioms. So for instance, in the days of Panini some prepositions #fù, afù, 34, aft seem to have been in common use, but in classic literature they are, if at all, rarely met with. Rem. The vaidik mantras contain accordingly a still greater number of prepp. and are displaying a still greater variety and manifoldness in employing them. So the old words afg: (without) preceded by an acc., its synonym preceded by an abl., (with) construed with instrum., do not occur but in the mantras, likewise : and 471, The upasargas, see 160. A:, A do not do duty of karmapravacaniya, but for a few passages; Pånini does not mention them in his list of karmaprav., nor are they used so in the liturgical books of the Veda. The once pre- - 1 geneous set of terms styled parts of speech, has no term exactly ans- wering to our »prepositions," but it calls them by different names according to their phonetical, etymological or syntactical properties. When compounded with roots, so as to make up compound verbs and the like, they are styled upasarga. But the same particles will be styled karmapravacaniya, when separate words. For this reason, the karmapra- vacaniya-class does not comprise such prepp., as 34f, g, , but on the other hand it contains some particles, which cannot at any rate be called > prepositions," as #f, . Cp. P. 1, 4, 58; 59; 83–98. 1) It is wrong to say that the noun-case, attending on the prepos., is governed by it, for it is not the preposition, that causes the case, but it is the general hearing of the case, which is qualified and limited by the preposition,
