than C. Candrasekhara's text is the basis of the present edition. In the Appendix I have added his version of the Prakrit passages, and his various readings as far as they could be ascertained from his com mentary. To make the meaning more clear, I have enclosed in brackets his explanatory motes which occur very often in the middle of the version, and have not corrected the numerous passages in which the laws of sandhi are neglected in the mss S=Samkara's commentary (Wilson, no. 4०; Aufrecht, Cat. Cod . no. 254 B). From folio 1०8b the ms. contains Candrasekhara's loss; the last folios, however, again belong to Samkara. Samkara's readings have for the most part been incorporated into Prof. B6htlingk's edition, and I have therefore abstained from पृuoting them. Besides, Samkara can hardly be called a commentator of the Bengali recension His readings are generally quite modern and not supported by the best Bengali mss All these manuscripts-2SNRI and CCa5-are written in the Bengali character D = a ms. of the Royal Library at Berlin (Chambers, 3०8) written in the Devanagari character. Of this ms. the various readings are given only to the beginning of the second act. I have elsewhere [[De Kalidasae 9akuntali recensionibus, p. 14 39.] described the ms. at full length y = Ch62y's edition, Paris, 183०. I have had no opportunity of collating the Paris ms., but it may easily be seen where Chey has misread it. It is true that Che2y's edition is ,a very imperfect work ' and that ,it abounds in typographical and other more serious errors' but this is not so much the fault of its editor as of the time and cir cumstances under which it was accomplished. Ch62y is indeed worthy of much praise for his energy and 2eal, and I look on this book as a second issue of his edition which I have been fortunately able to bring out in a more correct shape. After all, Che2y's edition has been to the present day the only one from which a true idea of Kalidasa's Sakuntala could be obtained. The editions prepared by Indian Pandits which have come under my notice and which profess to be editions of the Bengali recension, are all based on muss. of doubtful value and are alto gether uncritical. They are as follows: 3=ed. Calc., sak. 1786, edited by Ramamayasarmatarkaratna, repu blished from the edition prepared by Premacandra Tarkavāgsa. 8 = ed. Calc , sarvat 1926, edited by Jagammohamasarman and ००2००mder's Series) [Digitized by (Google
