________________
NOVEMBER, 1893.) ASOKA'S SAHASRAM, RUPNATH AND BAIRAT EDICTS.
299
ASOKA'S SAHASRAM, RUPNATH AND BAIRAT EDICTS.
BY G. BÜHLER, PA.D., LL.D., C.I.E. THE subjoined new edition of the Sahasråm and Rupnáth Edicts has been made according
1 to most excellent materials, rubbings (A) and paper-casts (B) made over to me by Dr. J. F. Fleet. The casts show the letters reversed in high relievo and indicate even the smallest flaws, abrasions and exfoliations in the rocks. It is in fact chiefly owing to them that a really trustworthy edition has become possible. Though, thanks to Sir A. Cunninham's kindness, a direct photograph of the Sahasram rock and a very fine rubbing of the Rūpnåth inscription were available for the first edition, they could not render the same services. For, the nature of such reproductious makes it impossible to answer a good many questions, which the decipherer must put to himself. They give merely surface-views, and necessarily leave one in doubt regarding the depth of the strokes and the minor details of the state of the stones. Nevertheless, one portion of the old materials, the photograph of the Sahasrâm rock, still retains a considerable value. For, since it was taken, the rock has suffered a good deal. Pieces have peeled off at the edges of the old exfoliations, and a new one has formed. Thus, to the left of the old exfoliation the letters vari á have disappeared in line 1, and on its right side the signs - iyani savachhal.. Similarly line 2 has lost, after sadhike, a stop and the syllable am, and to the right of the exfoliation the letters t.-éna cha anta. The new exfoliation has destroyed some letters in the middle of lines 6-8,2
The most important changes in the text of the Sahasram Edict, which the new edition exhibits, are l. 2, sadvachhalé for savinchhalé, suint[] for the conjectural husam te and 1. 8, -, i. e., ti, for yi. With respect to the first word it must be noted that the paper cast proves distinctly (1) that there is no Anusvira after the second sign, (2) that the shape of this second sign slightly differs from that used for vi. The corresponding passage of the Râpnath Edict las according to B quite distinctly chhavachhare, which represents exactly the Sanskrit
shadvatsaram, "& period of six years."
There is not the slightest doubt that the sign
may
be equivalent to and I, and that it is possible to read sadvachhalé. The form sad for Sanskrit shad occurs in the dates of the Pillar Edicta I.-VI., where we have sad-w-visati "twenty-six," and it must be noted that the dialect of the Pillar Edicts and of the Sahasrâm inscription is the same. The forms tadatva (Kalsi, Dhauli, Jaugada X.), dvo (Girnår I.), dve (Girnår II.), and dvadasa (Girnår III., IV.) prove that groups with va are admissible in the ancient Pali of the inscriptions just as in that of the Buddhist scriptures. Hence the word sadvachhalé is also grammatically unobjectionable.
These reasons appear to me sufficiently strong to warrant the assertion that the reading savichhalé can only be upheld in defiance of the fundamental principles of philology. He who still adopts it, has first to select an interpretation of the second sign which yields & word without any meaning, and next has to emend it as well as the perfectly intelligible form of the Rupnáth version. I, of course, have to plead guilty to having committed both these mistakes. My excuse must be that in 1876 I was still under the erroneous impression that the Asoka
1 Ante, Vol. VI. pp. 149ff. The facsimile of the Råpnáth version is an exact reproduction of the rubbing, which has not been touched up or corrected in any way.
. For further details see the notes to the transcripts. 3 See E. Müller, Simplified Grammar of Pali, p. 54.
* It is quite possible that the lovers of emendations will point to the readings savachhall or samhrachhall in the Myaore versions, as to proofs for the necessity of correcting those of Sabaarim und Rupoath. I have shown in my paper vu the new inscriptions, to be published in Dr. Haltzsch's continuation of the Epigraphia Indica, thalsa and sati may likewise be equivalents of Sanskrit shad.
