This conspectus is not necessarily complete, nor is all the material named of equal value or equally in demand; but every librarian should review these headings in relation to his reference department. To place rigid limits upon the stock is absurd, seeing that utility and not mathematical or other precision is the object of the work. If, therefore, a librarian finds some other special field of material is demanded he should add it without hesitation if he is convinced that the demand is not frivolous or very restricted. For example, to add an expensive and recondite (say) archæological treatise in Modern Greek at the behest of one reader is a case in which his decision might wisely be for refusal.
412. Classification.
412. Classification.—In its classification the reference library presents more physical problems than the lending library. The most minute classification is the best undoubtedly, and this should be used; but the size problem is a real one. While the greater number of reference books are of octavo size, quartos, folios, and even larger books are many. They cannot stand together in class sequence without an impossible loss of space. The simplest method is to have three sequences, for octavos, quartos and folios respectively, in appropriately sized shelves, in three different parts of the room. But this means various journeys across the room when all the books on a subject are required. The distance is abbreviated if the octavos, quartos and folios follow one another in each class. A third method which has proved most successful is to divide every tier (which is presumed to have adjustable shelves) into three parts and to run three parallel sequences in each, the octavos occupying the top part, the quartos the middle, and the folios the bottom. The parallel can be only approximate, but it is sufficiently close for the reader or the staff to review any subject completely and readily. With any of these methods broken order may be resorted to if it is thought well. The arranging of all quick-reference books in a separate complete sequence nearest the entrance or the place of service is a case in point; and special separate classifications may well be given to periodicals, to local literature, to the predominating industry, and so on without limit. Again, convenience is the supreme law.
Fig. 156.—The Main Reading Room of the Royal Society of Medicine, a modern open access reference library.
413. Cataloguing.
413. Cataloguing.—The catalogue of the department should aim at the maximum of fulness and be in as many forms as are necessary to bring out the entire resources of the department; there should be no retrenchment of time or labour in producing the best here, as a small collection of books adequately catalogued will give greater service than a larger one catalogued poorly. It is not an unfair paradox to say that the smaller (within reason) a reference library is the more detailed should be its catalogue. This being so, whatever kind of catalogue may publish the basal stock, the general current needs of the library can be kept supplied only by a card or slip catalogue of unlimited expansibility. The reference catalogue, even for books already in stock, can rarely be complete, and any fixed form of printed catalogue, unless it is supplemented by a MS. catalogue, will soon fail signally as a guide to the collection. As to the cataloguing form, experience proves that a mere author catalogue has a very limited value in reference work. It should be provided in some form, of course, but for one reader who inquires for a book by its author, a score require something about subjects, usually specific subjects such as the Horse, Verdun, Violin strings, Election Law, Tithes, Date of a Battle, Arms of Sussex, Birthplace of Douglas Haig, Words of a Poem, etc. There must therefore be some form of subject catalogue, and there is much virtue in the fully classified card or sheaf catalogue, with author and subject indexes. These, if carried out efficiently and minutely, will do the work that is required. By fulness of entry we mean that titles should be abbreviated as little as is possible within common-sense limits; that all bibliographical particulars, number of volumes, size, pagination, date, illustrations, maps, diagrams, glossaries, indexes, bibliographies and date and places of publication (except when London) should be indicated. Moreover, annotations of obscure books, and indicating sequences, commentaries, missing parts, and so on, are of special value here. Added entries may, and should, be carried as far as the cataloguing resources of the library allow, all books of composite character, miscellaneous works, transactions, many periodicals, etc., being analysed and displayed in the catalogue under their class headings. It is also most useful to collect in the catalogue references to bibliographies of all kinds contained in works which are in the lending library. For examples, the Home University Library and the Cambridge Manuals of Science and Literature are not books to be found in the average reference library, but they contain excellent little select bibliographies which the reference librarian will find useful, and an entry of each of these should be made. Every item which goes into the library should be catalogued, pamphlets and excerpts from other works, however small, included, if it is intended to preserve them, as well as maps. Photographs and prints probably need a separate catalogue, as certain considerations, dealt with later on, enter into their cataloguing, but in a card catalogue provision can be made for nearly every kind of material. Temporary material may be entered on a coloured card, which permits of rapid revision of the entries.
Fig. 157.—Reading Table, Chair, and Accessories, Royal Society of Medicine.