________________
48
EPIGRAPHIA INDICA.
[VOL. VII.
authority of parallel passages or by the evidence of the context in which itoccars. Notably, in a vast number of instances it is perfectly impossible to decide whether we have to deal with intentional strokes or with accidental cracks of the stone. This observation refers to the vowelsigns, and especially to the an usvára. The latter, we know, is frequently omitted; hence it is all the more uncertain in doubtful cases. Fortunately these are details of secondary im. portance, at least for the interpretation. Is there any need to add that I have scrupulously recorded all the various readings of former editors whenever they are of any interest, i.e. unless they are due to palpable mistakes or to misprints ?
The chief value of these documents consists in their throwing light-- though dimly- on a number of historic, paleographic and linguistic problems. But they are not the only documents of their kind, nor are such documents our only source of information. Though interesting, it is unwise to take up general questions in connection with a partial publication. This manner of proceeding is inevitable in the first period of researches and of discoveries, but later on it is apt to scatter the information inordinately and to hamper its discussion. It is at any rate natural to rescind the explanatory remarks and the historical conclusions. To my re-edition of the texts and to the explanations which it will suggest to me, I propose to add as a postscript the more comprehensive remarks which appear to be called for.
The really useful bibliography of the Karlé inscriptions comprises only two works - (1) Inscriptions from the Cave-temples of Western India, 1881 (CTI), and (2) Reports of the Archeological Survey of Western India, Vol. IV., 1883 (AS). In the first publication, Dr. Burgess has transcribed and translated the inscriptions in accordance with the readings and remarks furnished by Pandit Bhagwanlal Indraji; and in the second the texts and explanations have been subjected to a fresh revision by Bühler. Though this edition is more recent and marks in the majority of cases a progress on the first, it has the drawback of assigning texts of the same locality to different chapters according to their official or private character. The new Plates follow Bhagwanlal's numbering. I have added in brackets the number given in AS.; "K." refers to the chapter entitled " Earliest Kárld inscriptions" (p. 90 ff.) and "Ksh." to " Kshatrapa and Andhra inscriptions" (p. 98 ff.). The remarks regarding the position of each inscription are copied from the labels accompanying the estampages.
No. 1, Plate iil. (K. 1). Chaitya cave. Left end of the verandah.
Техт. Vejayantito sethiņa Bhuta palen[8] (1) selaghara (2) pariniţhapita[m] (3) Ja[m]budipamhi (4) Atama[ñ] (5).
REMARKS. (1) CTI. Bhata'. The mark of & seems to exist clearly in nd; but it is rather short and, as there are traces of a crack on its right, it may be the result of the wearing of the stone.-(2) AS. and CTI.gharar.-(3) Probably the antevúra has disappeared in the crack; it cannot be said to exist with certainty.-(4) I feel inclined to explain the development of the central loop of ja by the presence of an anusvára.-(5) AS. utama; CTI. uttama. The long appears to
The partial attempts at interpretation of Bird (Historical Remarches on the Origin and Principles of the Buddha and Jaina Religion, 1847) and Stephenson (J. Bo. Br. R. As. 800. Vol. V. p. 162 , and p. 426 4.) DO SON now only historical interest. As regards the general description of the caves, the first exact reports are those of Ferguson (J. R. A. Soc. Vol. VIII. p. 80 11.) and of Wilson (J. Bo. Br. R. As, Soc. Vol. III. p. 48 ff.). At prenent exact and methodical information may be found in the Rock-out Temples of India of Fergusson and Burgem, and in the Reports of the Arola logical Surney of Western India, Vol. IV.
