been made by a Tibetan . . . .".(5) No doubt the compilers did rely on Dge-'dun-chos-'phel's catalogue, at least with regard to authors, titles and editions. The identifications of a number of fragmentary works such as no. 951-1 as a treatise by 'Jam-dbyangs-bzhad-pa or the proposal of Sog-zlog-pa as the author of no. 1038 was no doubt made by someone who was quite well versed in Tibetan literature, i.e. probably by Dge-'dun-chos-'phel himself, though conceivably the later lama scholars who worked at the society may also have contributed something.
The catalogue is sufficient to give a rough idea of the scope of the collection. But it suffers precisely from some of the defects of Dge-'dun-chos-'phel's catalogue enumerated by Sarkar. Since it does not list titles or author names in their full forms and also does not record the complete contents of each bundle, it allows one to identify positively only a part of the rich holdings of the collection - this recognizable portion consisting mainly of works that are well-known anyway. The present investigation is an attempt to make the contents of the miscellaneous series still better known and thus to encourage the study and publication of the most important texts found within it. This report is basically a slightly expanded shelf list - a series of descriptive notes I took while going through the collection bundle-by-bundle during a visit to Patna in March and April of 1987. At that time I was using this collection in connection with my research on the Kashmiri Pandita Sakyasribhadra (d. 1225). I was of course curious to find out what other sources on related subjects I could find in this unexplored but obviously rich collection. So, with the consent of the society, I went through the central part of the non-canonical or miscellaneous collection, taking notes as I went. What mainly interested me were the truly "miscellanous" works - i.e. the early manuscripts and old blockprints, and any works published separately from the standard oeuvres and other collections. Therefore I worked through the books from bundle numbers 167 to 595, and did not investigate in detail the better-known (and for the most part already-published) collections at the beginning and end. After I finished my survey of the middle of the collection, I did go back and verify the identities of the initial and final sections too, opening the first and last bundles of each gsung 'bum or compilation and counting the intervening bundles to ascertain (albeit in a cursory way) their completeness.
Because of the onset of the hot season and the time constraints imposed by my other research, I did not have time carefully to study each work and fascicle. I went through the collection as quickly as possible, averaging two to three bundles per hour. Of course even at a slower speed I could not have immediately assessed the importance of all I was seeing, but it soon became clear to me that this collection did contain a number of rare and significant