________________
JANUARY, 1918]
THE WIDE SOUND OF EANDO
27
If we sound the s-37 in all these words, as also in the words T, 171, *T, TE etc., we shall perceive the peculiar fri (open, wide ) nature of the phonal phenomenon which alone can give the Gujarati विकृत, मैं and औ. (C) Now, examine the sound in the following words on the basis of accentuation
(a) 1977-fi TT fra
(6) 1967 rast ( 4) igrant warar | (0) अन्धकार अंधार (अंधवार) अंधवर अंधार अंधेर (d) पर्णपल्ली पण्णवली पण्पाली पानोली (name of a village in Surat District). (e) # 99 7573 (1) चतुर्दश चउहह चऊवह चौद (a) * Tara (6) * ** *** krst
(c) पृथुलक पहुलडं उलउ यहाळु II
(d) NITxar 3T TWO SITIS
(e) 99 T420 G (from Hindi) i o 2199* TË R2 watt (Hin., arty-separate).
19 True, the case of artit does not fall under the principle under consideration because of the long #; but it is taken for that very reason, as the long furnishes a test and shows how the long , which is necessarily accented, comes in the way of prati-sampraedraņa.
Siddha-Hemachrındra VIII. i. 101 shortens this (TFT), but the glossary tells usदेषु क्वचिनित्यं कवचितिकल्पः , thus giving an opening for option, and we may very well regard गहिरं, and it as alternative forms.
20 Dr. Sir R. G. Bhân dârkar regards the T in P as direct change from Here also sit in for direct from 3, and the in terms direct from 9 . (See his Wilson Philological Lectures, pp. 166, 145). But I believe these must pass through the shortening stage shown above..
21 Hemachandra (VIII. iv. 422) gives 7764 TTC. This 70 (7 5 ) may be advanced as the origin of tg, and with apparent reason. But there are some strong points, in favour of Trag . as the origin of the they are:
(a) One of the Manuscripts of Si. He gives the reading 17 (as the adesa of 7). (6) Spare is used in Gujarati, as well as are. (c) The or at tacked on to 77 by Hemachandra appears quite inexplicable and its arbitrary
nature leaves a mystery, while TFT supplies a good explanation of the land . Thus it seems that art , an evoluto of FUTS before Hemachandra's time, was regarded by him as the axlesa of 77 on the strength of the meaning and external similarity, and he was probably oblivious of the other phonetic phases, especially the loss of the unaccented initial syllable in T .
Sir R. G. Bhândarkar (Philological Lectures, r. 168) gives Hindi Sratar in the sense of unexpected', and derives it from i t . The sense in which are (Tr ) is used in Gujarati is 'separate, and would favor the derivation from tra . It would also be enlightening to necertain the Hindi text whore word is used and what sense fits in there.
22 Here, although the leaves aconted at the early stage, the accent gives way under the double influence of
(1) the loss of the conjunct, FC without the compensating lengthening of the preceding vowel, and (2) the tendency to accentuate the second syllable in a word where the first syllable is unaccoutech
and eventually therefore gets dropped. (The change of in T to indicates that the word is become a part of the whole compound, and honoe the is eventually subordinated.)