________________
26
NOTES-VIII. iv. 1-259.
VIII. iv.
The fourth qre treats from 1 to 259 of the various socalled substitutes for some specific Sanskrit roots. This section of traicis is not arranged in any systematic way, but gives roots and their substitutes at random. Some of these substitutes are pure dai roots while others could be derived from the original Sk. by applying rules in VIII. i and 1. I have arranged these substitutes alphabetically and given them in appendix II. ir George Grierson in, his monograph on the Prakrit Dhatvadesas ( Memoirs of the Asiatic Society of Bengal, Vol. VIII, No 2, 1924 ) says that Prakrit roots may be looked upon as falling into four classes, viz. :-"1) Those which are identical with the corresponding Sanskrit roots; such a roots is 7f which 18 identical in both languages. (2) Those which are regularly derived, according to the ordinary phonetic rules, from the corresponding Sanskrit forms. Thus, under the phonetic rule that Sanskrit medial 5 becomes Prakrit , the Sanskrit root are becomes fra in Prakrit. Such a root cannot be called an pict, substitute, as there is no substitution, but only development. (3) Those which cannot be connected with any corresponding Sanskrit roots by any admitted phonetic rule. Thus, an strat for the sanskrit root aes is ze with the doubled, so that the anskrit gefa may be represented by as well as by Tag. These are all true ETIS, and a great number of them are borrowed 549768 and cannot be referred to Sanskrit at all. (4) Those which are regularly derived from Sanskrit roots, but which have changed their meaning, and which are therefore, by Prakrit grammarians equated with and substituted for some other Sanskrit root which has a meaning more nearly akin to the acquired meaning of the Prakrit word. Being substitutes they are also zigzTS." In my opinion words change their
