to Dr. D. A. Robertson, " The circulation statistics ought to be available, if they are not; the fact is significant." 22 Twenty-two librarians did not know how many volumes were loaned to other institutions. Twenty institutions did not know the number of volumes they borrowed from other institutions, although such borrowings are a direct guide to the book needs of their own libraries. One librarian replied, "Staff too small to keep statistics." Only 14 librarians could report the length of time periodicals were away from the shelves for binding, although delays at the bindery were noted by 180 (out of 500) faculty members as being a hindrance to their use of the library. The University of California and a few other institutions keep a close check on the length of time publications are unavailable on account of binding, but many librarians, in some cases where the complaints of the faculty were the most numerous, did not know how long binding required, and did not regard statistics on this subject as of any importance. The attitude of some librarians was expressed by one who stated in his reply to the questionnaire that such studies were "too trivial," and their object only "to satisfy curiosity."

Few librarians are able to give, for any appreciable length of time, the percentage of requests for books and information which are not met, with an analysis of the reasons for failures. No case studies of individuals who leave the library without obtaining the material they desire have been found in professional literature or elsewhere A lack of agreement as to terms is also noted. There is no agree ment as to what is meant by " volume," " periodical," "serial." Wha does "number of volumes " mean? Does it include pamphlets placed in pasteboard covers? The lack of agreement as to terms should have had the attention of the American Library Association. Ther is need of careful consideration by some body of university librarian of the subject of the statistics which are significant and should b carefully kept.

Actual figures of delays in binding and ordering of books may wel show the need of more assistants or changes in methods or ever changes in State laws. Statistics as to the number of interlibrary loans, both incoming and outgoing, the proportion of books not sup plied at the loan desks, the growth in number of books loaned, th number of volumes discarded and the reasons, should have a marke influence on the administration of the library and be important fac tors in presenting data for a library budget. Yet comparatively fe libraries kept such records, although all report carefully the numbe of volumes in the library. In many cases these statistics of "num ber of volumes in library" can have no significance, since all vol umes discarded in past years are counted as still in the library.

Robertson, D. A. The college library. Educational Record, 10: 18-19, Januar

The great discrepancy among colleges in presenting data for this survey is noteworthy. The larger institutions where the data were more complicated were able to present much more satisfactory figures than the small colleges.

Possibly the great emphasis laid in the past on the "internal " work of libraries, the cataloguing, accessioning, and ordering, has overshadowed the study of the use of libraries. Possibly, the fact that most library school courses deal with methods of library routine may account to some extent for a lack of knowledge, and even occasionally a lack of interest in studies of a library from a user's standpoint.

Highway experts in progressive States know the traffic load on every main highway. They are studying the failures of pavements to withstand use. They count the cars and from their count determine the width of road, type of construction, and necessity of improvement. They are carefully estimating the future load and are planning construction to correspond.

Similar studies are needed by college and university libraries. What difficulties are found by students in the use of books? What proportion of publications requested by faculty members are not available at the time needed? How many requests for information from books are fully met, partially met, or not met at all? It is the duty of the library to know what service is not rendered. In case of failure to meet any request for books and information the library should be able to give a satisfactory reason. The burden of proof that requests are unreasonable is upon the library, not upon the inquirer.

In addition to these studies three definite means are afforded to librarians to see library service from the standpoint of the reader. They, as readers, can use other libraries where they are not known. and note the difficulties. This method used by two librarians resulted in the rather searching question, "Is the service to readers in my library like this?"

A second method afforded to chief librarians is to schedule themselves for certain hours a day at the loan and reference desks. Even an hour a week at each desk will prove a revelation of the difficulties. some individuals have in the use of libraries. Deficiencies in service actually seen are remedied more quickly than are the same difficulties when reported by assistants. Many an assistant at a loan desk has remarked, "I wish the librarian would work here a few minutes. There would be a change." This recommended method has also been tried successfully by some librarians.

A third method is to visit departmental seminar meetings or special classes of students and engage in a frank discussion of the failures of readers to obtain service. This method at one institution

resulted not only in improvements in the routine work of the library, but also in recommendations from instructional departments to university administrative officers for increased library appropriations

Summary

Means of increasing usability of libraries.—(1) More attention to the individual inquirer. Abandonment of such phrases as, “Look it up in the catalogue," "We do not have the book," and "The book isn't in." The assumption by librarians that any failure to furnish material requires explanation, that the burden of proof is upon the library in case of failure in service or refusal of request.

(2) Adequate assistance in use of card catalogue.

(3) A provision that every freshman by the end of his first semester understands the use of bibliographic tools, the card catalogue, and the most important reference books.

(4) Active interest by faculty members in the use of books by their students. More consultations of faculty and library staff to improve instruction in connection with reading.

(5) Prompt messenger service to faculty.

(6) Libraries open not fewer than 14 hours a day and several hours on Sunday as use requires.

(7) The availability on open shelves of 10,000 volumes, including books for general reading. Display for one week of all new books for general reading.

(8) Skilled assistance to students, faculty, and research workers in search for informational material.

(9) Liberal policy in use of library material by citizens of the State.

(10) Participation by chief librarian in work at loan desk and reference desk, if only for a few hours each week, in order to obtain direct contact with readers and first-hand information of failures in service.

(11) An occasional use by chief librarian of other libraries where he is not known in order to ascertain by actual experience the difficulties encountered in the use of a library.

Few, if any, libraries of land-grant institutions reach the ideal in service rendered. Many are failing in a majority of the items mentioned. It is obvious that the ideals of the library service which have been outlined in this section require ample provision for pur chase of books and periodicals, an adequate building, and a sufficient library staff with special bibliographic and educational qualifications The needs for these provisions are discussed in the three following chapters,

Chapter IV.-Books and Periodicals

False Standards

Quantity v. quality.-The inadequacy of book collections is noted by faculty members as the most important cause of difficulties in the use of libraries. For years administrators and librarians have emphasized in catalogues and publicity circulars the number of volumes in libraries. On the other hand, little emphasis, in print at least, has been given to such subjects as the type of book needed, the type purchased, methods of selection of books for university libraries, the relationship of books purchased to educational needs, and objective tests as to the use of material.

The North Central Association of Colleges and Secondary Schools requires a minimum of 8,000 "live, well-distributed" volumes. Any secondhand bookstore would be glad to furnish what might be called such a collection at 10 cents per volume. A search of old attics would doubtless yield a much larger number at much less cost. The addition of such material in the past has resulted in the swelling of collections to meet a mechanical standard, rather than an addition of usable books. The president of a small college remarked to one librarian, "We now have 10,000 volumes. We more than meet the demands of the North Central Association; we need to buy no more." No fixed number of volumes is satisfactory as a standard. The number and type of books needed will vary as instruction and research in different institutions vary and, even within an institution, the books needed will change continually as new material is published. The actual number of volumes should have less weight than the adequacy of the collection to meet the need of every instructional course and every research project on the campus. The amount spent annually for additions of suitable material and the number of subscriptions for current periodicals are more significant than the number of books in the library.

A new standard has recently been emphasized by the proposed compilation of lists of "best books" which should presumably be purchased by all college libraries. It is clear that such lists, per se, will not provide a permanent, adequate solution of the problem of the selection of books for college libraries. Lists may be a guide as to whether the selection has or has not been well done in an

institution. They can aid selection but can not replace the initiative of the individual instructor and the librarian. Methods of teaching and use of supplementary reading matter will and should vary with different instructors as subject matter of courses varies in different types of institutions. One of the most important desiderata for the land-grant college is to develop and retain the interest and the initiative of the individual instructor. Reading lists may tend to deaden two of the most valuable features of American higher education-the individual characteristics of the American colleges and the deep interest of the instructors in their work. Standard lists of books to be purchased by all institutions, valuable as they may be as guides, are therefore dangerous if they are interpreted as offering an easy substitute for methods which will insure an adequate book collection for the specific programs of instruction and research conducted by individual institutions. Each institution should attempt to determine for itself the adequacy of its book collections rather than rely blindly upon lists or attempt to meet a standard based on number of volumes.

Selection of Books

The selection of suitable books for public libraries has received much deserved attention in professional literature. Practically no consideration has been given in printed literature to the question of methods for the selection of books for college libraries and the mutual relationship of instruction and library staffs to the purchase of books. The lack of material on this subject justifies the following extended discussion of (1) the faculty and book selection; (2) the librarian and book selection; (3) allotment of funds; and (4) summary of recommended methods.

The faculty and book selection.-If all instructors and all research workers are actively concerned with the availability in the college library of the published material in their fields, and if funds are available for the purchase of this material, the library collections will be adequate and will be kept alive. The process will be a continuing one. The initial recommendation for books and periodicals needed by students is a duty of the instructors responsible for the teaching. The instructor may be aided in his selection, but his interest in reading in his field can not be replaced.

Many departments in some of the larger institutions have appointed library representatives whose special duties are to see that the library meets adequately the needs of the department. This ar rangement works well if it supplements and does not curtail the activities of other members of the department in recommending desirable purchases. Heads of departments should realize their