The arguments generally advanced in favor of the system are these:
1. The instructor needs to be able to refer, at a moment's notice, to any book relating to his subject.
2. The system enables the instructor to keep a more careful watch over the reading of his students.
3. The best interests of the library demand that each division of the library shall be directly under the eye of the men most interested in it, that is to say, the instructors in the various departments; that they should direct its growth and watch over its interests.
That the first and second of these arguments have great weight cannot be denied, but with a properly constructed library building and most careful administration the requirements of both instructor and student can be met quite as well by a central system.
It is, of course, quite impossible for each instructor to have in his office all the books necessary for his work. The duplication necessary for this purpose would be impracticable even for the most wealthy university. He must, therefore, go from his office or class room to the department library and search for the book himself. With the confusion which generally reigns in a library of this sort, and with the lack of effective registration of loans, this is quite often a matter of some difficulty.
At Columbia University the office of each professor is in telephonic communication with the central library. When a book is wanted the library is notified by telephone, the book is found and sent out at once. Within ten minutes from the time that the request reaches the library the book is generally in the instructor's hands. He may lose two or three minutes' time, but the amount lost is more than compensated by the readiness with which others can use the books of the department, and by other advantages to be considered later. At Columbia, too, the system of stack study rooms provides in a very satisfactory way for the second objection. There, as many of you have seen, the stacks are distributed through a series of small rooms, the light side of which is supplied with tables and used for study rooms and for seminar purposes. If the instructor can use the departmental library for his work room, he can certainly use this room to as good advantage, for here he has the entire collection and not a selected few of his books. I believe fully that an instructor who is sufficiently interested in the reading of his students to watch over it carefully in his departmental library, will find that he is able to keep just as close a relation to it, if his students are working in a central library. He may be obliged to make slight changes in his methods, but the result ought to be the same.
The third argument in favor of the departmental library system is of a different nature. Is the librarian or the professor best qualified to direct the growth and watch over the interests of the different departments of the library? So far as I know, this argument is given more consideration at Chicago than anywhere else. It may be true, in certain cases, that the professor has the greater qualification for this work, but when this is the case it argues that the professor is an exceptional one or that the university has been unfortunate in the selection of its librarian.
It is quite needless to say that the librarian should be in constant conference with the teaching force regarding purchases, but that he should delegate all of his powers of purchase in any given field, admits of the gravest doubt. Laude, in his recent work on the university library system of Germany, attributes a great deal of the success of those libraries to the fact that they are independent and [autonomous] institutions, enjoying a much greater measure of freedom than is accorded to any similar American institution. Too many professors are apt to buy books in their special field and slight other lines of research in their own subject. For example, a zoologist, who is doing research work along the lines of embryology, is very apt to overload the collection at that point and neglect other equally important lines.
Again, very few instructors, even granting them the qualifications necessary for the work, have the time or patience for it. If the amount appropriated to the department is at all large, a considerable portion of the sum is quite frequently unexpended at the end of the year. Some interesting tables, prepared by Mr. Winsor for his report for the year 1894-95, show that in seven selected departments the amount of books ordered, including continuations, was only about 50 per cent. of the appropriation, plus one quarter, the allowance for orders not filled. While this proportion would probably not hold good in all departments or in all places, it exhibits an almost uniform tendency and a tendency which must be corrected if a well-rounded out library is to be secured.