Coordinamento Servizi Informatici Bibliotecari di Ateneo
	Università degli Studi di Lecce

I wish to begin this short article by thanking the organizers of the splendid Lecce conference for inviting my to speak to a selected audience of Library Scientists. What I am going to say comes from my life experience with mathematical journals, books and their editing; this latter experience was actually not very long (the first year or so on the life of the Boletim da Sociedade Brasileira de Matematica, three years as co-editor-in-chief of the Canadian Mathematical Bulletin and three years as a member of the now defunct Grant Selection Committe of the Natural Sciences and Engineering Research Council of Canada). My observations are not those of a specialist on Library Science (which I am not) but rather, those of a person who loves books and wishes to lend his support for what the valiant group of librarians at the University of Lecce has been doing! I also take the opportunity to dedicate this paper to the memory of Michele Sce.

I had my first contact with a big library system at Harvard University; that institution of research and higher learning had in early sixties three main libraries (Widener Memorial Library, Lamont Library and Houghton Library) plus about seventy departmental libraries. I still remember the feeling of reverence while wandering through the stacks of the Widener Library, which at that time already had about six million volumes! Next, I got in touch with the central library at the Swiss Federal Institute of Technology in Zürich (ETH), in the beautiful building designed by Gottfried Semper (built between 1858 and 1864); there, one was not allowed to see the books: everything was done with marvellous swiss efficiency within minutes of one's request typed on a special terminal. All in all, I felt sad for being separated from the books and having to borrow them in a rather authomatized fashion. Finally, I should mention the Queen Elizabeth II Library at Memorial University, in Canada. That library was built, at the cost of some seventeen million dollars, while I was still at Memorial, and was supposed to accommodate about two million volumes and give reading accommodations to over one thousand students; it is a modern building, full of light but alas, already too crowded with books at the dedication of the building (indeed, book overcrowding seems to be the common characteristic of all those libraries; that was the reason why nobody could get to the books at ETH).

Libraries' overcrowding is easy to understand: according to [ 1 ], the number of books published around the world in 1991 (UNESCO figures) was 863.000! Furthermore, the precious available space at university libraries is also contended by huge amounts of bound volumes of all sorts of journals. The same source I quoted before - see [1, page 171] -tells us that there are currently l 8,000 journals worldwide! To complicate matters, books and journals can be very expensive; this is particularly serious in the sciences: I would suggest that the average book on advanced mathematics could cost around one hundred dollars, and there are mathematics journals whose subscription could run into a couple of thousand dollars per year! Of course, libraries should not purchase books just because they belong to special collections (the Killam Memorial Library at Dalhousie University in Canada has now implemented a selective purchase program; for example, it does not buy anymore automatically all volumes of the Springer Lecture Notes in Mathematics), nor should they subscribe to all possible journals. We all know that many journals in circulation, particularly some of those edited by various Mathematics Departments, have little scientific significance and libraries should not waste their space or money with them; I must also say that I do not believe in the argument that such journals are good for interchange: as a general rule the journals with which they are exchanged are also of bad quality....

To solve the overcrowding problem, libraries around the world have recurred to several solutions: storage in special deposits (Harvard used to send part of its collection to be stored at the New England Book Depository; Memorial University sends most of the older journals to a deposit downtown St.John's), microfilming of some collections and successive disposal of the volumes, etc. The solution to the ever increasing price of publications at a time of shrinking budgets is much easier: you just stop buying books and journals, like Dalhousie University is now doing. These are certainly headaches for the Librarian; but what about the user, the students and scientists who should be consulting these publications? For them, and according to the work they actually do, the problems are normally of two kinds: difficulty of access to older volumes and information that does not reach them fast enough (there are journals with a large backlog and with long publication time). Of course, nowadays after the advent of electronic mail, the active research mathematician is always up to date with current research in his or her field; nevertheless, these persons would like to have fast access to results in nearby areas of research.

A partial solution to the questions raised before may lie on electronic journals, provided some ground rules are firmly established. Of course, this idea is not new and several good electronic journals already exist. An example of what I consider a good electronic journal is the New York Journal of Mathematics (http://nyjm.albany.edu:8000/nyjm.html); this is a refereed journal-with referees not in the editor's board—with high quality papers and very fast publication time; last, but not least, it is free!

There is no problem about a certain uniformity of presentation for mathematical journals: one could request that the papers be written in a version of Donald Knuth's TEX (see [2]), but this is just a proposal. On only one point I would not compromise: the high scientific standards of the journal!

Yes, I agree that papers published in this way will have problems in the long run like the eventual demagnetisation of the tapes or CDs or even the possibility that future hardware will be unable to read our present day "storage devices". The fact remains that the really important contributions to science will survive, who knows maybe as books... Indeed, I do hope that libraries will not disappear: in the long run, I would not exchange the dimly lit stacks of the Widener Libram with the sophisticated and totally electronic library in the command deck of Star Treck's USS Enterprise!

The bibliography
[1] Bercuson,D.,Bothwell,R. and Granatstein,J. - Petrified Campus, Random House of Canada, Toronto 1977.
[2] Knuth,D. - The TEXbook, Addison Wesley Pub. Co, Reading 1970.

Prof. Renzo Piccinini
Vicepresidente INDAM (Istituto Nazionale di Alta Matematica)
Direttore Scuola Matematica Interuniversitaria
Professore Ordinario c/o Dipartimento di Matematica e Applicazioni - Università di Milano-Bicocca
Viale Sarca 202, 20126 MILANO
e-mail: renzo@mat.unimi.it