Preface by Cappeller to this, the second, edition We may therefore assume that, beginning at iii.28.3, B reads porava anicchapurao vi sambhasapametaepa paricido[] viramati. This senseless colocation does indeed convict the scribe of gross negligence; and yet it seems to me quite possible that he had a vague feeling that the second part of the dialogue in question was spurious or at least super fluous. Taken by and large, the new text has not gained much by the collation of ms. B The advantages of the new text over the old are due to a more careful working up of the old material, and, above all, to Pischel's Prakrit studies. The results of those studies are embodied in his Prakrit Grammar (Grammatik der Pralkritsprachen, Strassburg, Truebner, 19००) and they appear in this text at every step. It is a matter of course that the new text is printed exactly as Pischel left it in his interleaved copy. The divergences of the new text from the old are given in summary form together near the end of this volume. Candrasekhara's Sanskrit version of Prakrit words or phrases, and his Wariants, together with his explanations of words or phrases, are put by Pischel in his first edition immediately after the text (page 171: atha Sricandrasekharasya praktachāya pathantarapi ca). It seemed to us that we could better meet the needs of the beginner by giving in this volume, at pages I I 1 to 147, the approximate Sanskrit equivalents of the words of the Prakrit passages in the order of those words. This table of equivalents follows exactly the Pralkrit text of this edition. It disregards entirely the rules of phonetic combination, and treats each word as an isolated unit. Thus, talking Sakuntala's request at i. 17.3 as an example: for ta sidhilehi dāva map, we give , at page 1 12, as equivalent, tat sithilaya tāvat emat, and not tasmacchithilaya tavadermat. The prior eguivalent, in its external look, corresponds better with the Prakrit; it is clearer for the eye; is on the whole more easily and quickly compre hended. Our procedure saves the trouble of first making the combi mation, and then saves the student the trouble of unmaking it. Both processes are here gratuitous and useless . A knowledge of only the most important phonetic correspondences between the Prakrit and the Sanskrit is here presupposed; and accord ingly, for difficult cases, references to Pischel's Prakrit Grammar h been given in the foot-notes of pages I 11 to 147. In some cases the Sanskrit ,euivalent* given by us is not the phonetic eguivalent of the 17 [Digitized by (Google
