________________ xliv THE INDIAN ANTIQUARY APTER IV (v) Revision. When the Bower Manuscript was exhibited for the first time in Calcutta in November 1890, it was stated (Proceedings, As. Soc. Beng., p. 223, Journal, As. Soc. Beng., 1891, Vol. LX, p. 137) that "the writing was entirely in black ink." So it no doubt appears at first sight; but on closer examination letters and syllables are met with occasionally, which are written in a very light, or faint, ink. The significance of these light-inked letters, namely, that they indicate corrections, is disclosed by such cases as the following. In Part I, fol. 469, the original writing in black ink was me ?u, which is false for me srinu. Here the omitted syllable-sri is inserted below, in the interlinear space, in almost invisible light ink, and the proper place of insertion between me and nu is marked by two minute strokes, also in light ink, above those two syllables. Again, ibid., fol. 368, the original black-ink writing was prokto 8u, and this is, as it should be, corrected into proktal sa, by inserting a visarga and cancelling the top-strokes of the vowel o by two minute strokes, all in light ink. Similarly, ibid., fol. 367, an originally omitted visarga is inserted in ajaran But not infrequently corrections are found made also in black ink. Thus, in Part I, fol. 464, we have the original reading sa-mustam, which is adjectively made to qualify the preceding noun triphalum, corrected into sa-mustam, which, just as the following sa-sarkkarar (derived from sa and sarkkara), now qualifies the succeeding noun aschyotanam. Here both, the original as well as the correction, are in black ink, Again, ibid., fol. 5a?, (p. 7), the original blundered reading muvva is corrected to miiuva, both in black ink, though another error is left uncorrected; for the fully correct reading should be miruva. Ibidem, fol. 469, there is another instructive example. The original reading pralepaik is corrected to pralepah, both again in black ink. As a matter of fact, the noun prale pa refers to both, the preceding instrumental plural ardha-rupail and the succeeding nominative singular samprayojyah, and may grammatically be made to agree with either. This correction, as well as the correction of sa-mustam in black, and of proktah in ligat ink, shows that the revisers, whoover they were, were far iliar with the technicalities of the Sanskrit language. Equally instructive is an example ibid., fol. 566. Here we have the word laua nopetair entirely in black ink with the exception of the syllable no which is in light ink. It would seem that the original writer in black had left a gap for that syllable, which for some reason he had omitted to write, and that a subsequent reader of the treatise supplied the missing syllable no in light ink. The fact that the original writer should have failed to recognize the com. compounded with upeta, seems to suggest that he must have been a rather illiterate person,-- a conclusion which the occurrence of the numerous other errors (see Section iv, p. xlii) in the original writing tends to confirm. A further instructive example occurs in Part II, on fol. 76. Here the last word of the tenth line appears to have been originally dapaye in black ink. To this the reviser added in light ink the terminal t (dapayet),72 and after it, the vowel e, as if to commence a fresh verse. Then noticing his mistake-for as a matter of fact the vowel e which commences the new verse does stand at the beginning of the eleventh line-he cancelled the superfluous e by two minute double-strokes. The foregoing remarks are concerned, in the main, with Parts I-III of the Bower Manuscript. The general conclusion suggested by the observed facts is that those Parts were originally written in the usual black Indian ink by a somewhat illiterate writer, and that some of his numerous errors were afterwards corrected by a more intelligent user of the manuscript at different times, sometimes in black ink, at other times, when for some reason good black ink was no! at hand, in diluted ink. 1 Both forms dapaye and dapayet, are correct; only the former is Prakrit, while the latter is Sanskrit,-another indication that the reviser was a person familiar with Sanskrit.
