________________
- FEBRUARY, 1918)
SANTIDEVA
49
take due to the i suorance of the Ganda pronunciation of Sanskrit, I may mention that such a scholar as Bloomfield in his Religion of the l'eda, p. 54, speaking of the Persian translation of the Upanishads ma le for Dara, says that the Persian pronunciation of the word "panishad is oupanekat", whereas it is the Gauda pronunciation. Idg. sweks became Skt. 2, which Gaudas pronounce khash; Idg. skeut bicame , which Gaudas make khubh. In this connection it must be remembered that Iug. sw in some cases become a in Sanskrit and kh(w) in Persian ; thus the Persian analogue for sreilas is kh(wjay, for srasar is khwa har, and for s11-karas is khal. Curionsly enough Idg. kw when fronted by the influence of front vowels becomes s in Persian, corresponding to Skt. &; thus Idg. kweit, Skt. śrētas, Pers. safid. Hence the history of Skt. C ought to be rediscussed in the light of these facts.'
Scientific conclusions on the gradual changes of Sanskrit sounds are vitiated by four facts, (1) Maharashtras have been the main teachers of Sanskrit Grammar for the past two centuries or more and have imposed their Dravida pronunciation on Sanskrit ; and European Scholars have on that account not given the Gauda pronunciation its dues. (2) The Gaudas of Benares have for a long time been under the influence of these Maharastras and their own pronunciation to-day is a very mixed one. (3) Sanskrit was never the spoken langungo of the people; it was the Sanskrita, the literary, conventionalized form of the language of the people, first of the Indns valley, then of the Madhyadeśa, and lastly of Magadha and perhaps also of the Marntha country, before it became finally fixed in its present highly artificial form, dennded of syntax, divested of idioms, eminently suited to be tbe language of scholars, but nn fitted to act as a means of registering the changing sounds of a living language. (4) The linguistic survey of Northern India has been conducted by gentlemen without a training in phoneties, and their enquiry has been to some extent vitiated by a belief that Sanskrit is the norm and the languages as spoken are corruptions of the Sanskrita bhashd.
. My object is not to solve these problems, bnt merely to prove that the Sanskrit alpbabet is not deroid of perplexing difficulties, nor is Sanskrit pronunciation an invariable fixed thing ns people usually supposc. To one who knows the facts of the case and is not blinded by prejudice, it is as full of difficulties, as full of variations, as any other language.
SANTIDEVA, BY MAHAMAHOPADHYAYA HARAPRASAD SASTRI, M.A., C.L.E.; CALCUTTA. SÅNTIDEVA is a great name in the later Malâyâna literature. He is credited with the authorship of three works: (1) Bodhicharyavatara, (2) Siksha-Samachcbaya and (3) Sútra-Samuchchaya (See Silishd-sa muchchaya of Bendall, Introduction, page IV., on the authority of Târâpâtha). Sútra-Samuchchaya has not yet been found. But there is ample evidence that this was also written by Santideva, ns will be found in the sequel.
Bolhicharyavatara has been several times published and even translated into English. It was first published by Professor Minnef in the eighties. Then it was published in the Journal of the Buddhist Text Society by me. I had the advantage of collating a beautiful palm-leaf manuscript belonging to the Hodgeson Collection; in the Library of the Asiatic Society of Bengal. In 1893 I acquired a copy of the Panjiká commentary of the work by Prajnakaramati.. The manuscript was copied in the year 1078 A.D. in Newâri character. The copyist's name is not given. But he describes the commentator Prajia karamati as his tatapada, from which it may be in'erred that he was a disciple of the monk Prajñákaramati who was a well-known scholar of the Vikramasili-vihara (See M. M. Satis Chandra Vidyabhusbaņa's Indian Logic, Mediaval School, page 151) and flourished about the beginning of the 11th century. Another