________________ 140 APPENDIX. in our manuscripts, even in other cases; thus we have सेय (seya) for सेत (seta) where य (s) is used for त (t). I have not seen the change of (rj) to 32 (yy), but the reverse, as for qzlaut (paryushana) we have QFBAUT (pajjausana) according to the common Prakrit. Again HTET (bharyya) becomes HIITET (bhariya) instead of the common Prakrit HTTEIT (bharia). From this word, by the application of the grammarian's rules, we shall get something nearer the Marathi बायी (bayi) or बाई (bai). The word E TET (hridayasya) has not the peculiar form with us, the grammarian mentions. Nor is T (r) changed to (1), except perhaps in the doubtful instances of a trata (veruliya) for acei (vaidurya). and उराल (urila) for उदार (udara). I do not know how the change of 3 (ksh) to Foi (sk) which takes place in Magadhi, according to Vararuchi, is to be explained. In the Jain manuscripts EUR (kh) is usually written To (rak), like T (r) and F (k). Was this what the rammarian meant, and was it a mere form of writing, or is the peculiar sound intended utterly lost in the modern vernacular tongue? (chh) is often substituted for a (ksh) as in the common Clialect.
