LIBBARIES. 202 LIBBABIES. than anything else i= Uiu cause of the practical- ity of this syatem and its wide adoption. In other worda, its popularitj- lias been due to in- telligent practical usefulness." The User of a library as a rule has no interest or knowledge as to theories of the scheme used. His concern is to find quickly any subject wanted* and to find near it other closely allied subjects which he is also likely to consult. Ex- perience proved the proposed cliange from fixed to relative location practical. This solved most of the dilliculties, so that in recent years most libraries careful in deciding on methods .adopt the most important characteristics noted below as recognized essentials in a satisfactorj- library classification. Relative Index. The card catalogue is rapid- ly dis|)lacing the book form. The former could not be indexed like a book by reference to pages, for cards to which additions are made daily could not be numbered like pages. The solution for classed catalogues was to numlxT subjects so that the entry in the relative inde.K was followed by a number which meant not a page of a spe- cial book, but a subject in the complete scheme; e.g. geometry in the index is marked 513, mean- ing Class 5 Natural Science, Division 1 Pure Mathematics, Section 3 Geometry. This relative index number is a key to card catalogue, shelves, pamphlet collections, shelf-list, charging system, newspaper clippings, manuscript notes, in fact to everything arranged on the relative system. Books and other material are arranged in simple 1, 2, 3 order, and the old numbers indicating a fixed location are wholly abolished because this rela- tive index number shows not only what subject the hook treats, but also exactly where it may be found. Old numbers were frequently changed, while relative numbers, being permanent, are gilded on book-backs. Close Classing. All recent experience strong- ly confirms the wisdom of close classification on shelves, and the practice has so changed that what twenty years ago was called extreme close classification would now be considered medium or even coarse. Where libraries depend on cata- logues and indexes for the resources on any sub- ject it usually results, after a few years, in doing the work over at greatly increased cost. The best results can be obtained only by having as far as pliysically practicable material on each definite subject standing together. Lists of books printed as guides to readers are now usual- ly closely classified. Classification in Advance by Experts. By the relative index the number of each minute topic shows the exact place to which it has been as- signed by a recognized authority in that field. A classifier need have no knowledge of the gen- eral subject if he knows exactly what a book is about. Turning to that topic or any of its Bynonyms in the index, he finds its number, which, marked on book, cards, and elsewhere, gives ex pert classification with a minimum of labor. When this book is wanted any assistant who knows what topic is sought repeats the process, finding almost instantly the index number which guides to book or catalogue entry in 1. 2. ."? order. SniPLE Notation. In any plan the system of numbers is of the utmost practical importance. Extreme simplicity i* necessary for rapid and accurate use by readers and attendants. Only Arabic numerals, and Roman letters are suf- ficiently simple and familiar to be available. Must libraries use nothing but figures to number subjects. The expansive system by C. A. Cutter combines letters and figures. The 26 letters allow 071) combinations with two characters or 17,570 with three, and as compared with Arabic numer- als have 18, 40, and 118 times the capacity for three, four, and five characters. This is a great advantage, but produces numbers very unusual and complicated in appearance. The expansive ii 'the best of tile letter systems and the onlj- one used by any number of libraries. Both expansive and decimal systems meet the test of a good classification in insuring that books on the same subject shall be classed together and readily found when wanted. Book NiMnEns. Jlr. Cutter is also author of an ingenious and widely used table to keep books in each class in alphabetical order. The author's initial is followed by a number which is the translation into figures of the rest of the name, so that the system is much simpler in liandling and recording than it would lie if the author's wliole name had to be used. For scientific and other books where chronologic order is prefer- able, W. S. Biscoe, of Xew York State Library, devised a translation scheme by which a letter followed by simple Arabic ' numerals indicates date. Under the subject or class numbers most libraries use Cutter numbers, while many use also Biscoe numbers for science and useful arts. Some use simple accession order. Mnemonics. Both decimal and expansive sys- tems made large use of practical mnemonics. These are of much use, especially in geographic divisions, languages, and form distinctions; e.g. since in the decimal classification (ierman is in- variably 3 and grammars .i, one knows without reference to tables or index that in class 4 phi- lology German grammars must be 435. Library attendants are greatly aided by these mnemonic features and can often construct a number in- stantly. The principle also allows very minute subdivisions of topics where it is needed for de- tailed notes of speciali.sts, as in limited divisions of historical periods. In history. after the number for country means no further geographic subdivision and that the figure following denotes time; e.g. 942 is history of England (class 9 his- tory, division 4 Europe, section 2 England). indicates no further geogiapliic division. 6 the sixth or Stuart periofl. 1 the first Stuart King, so that 942.001 means history of England in the reign of .Tames I. (1003-25)". Or. still further, a specialist may be collecting material on Norfolk imder James I. If so 001. the number for reign of James I., added to 942.01 (Norfolk), making 942.01001. would indicate with absolute precis- ion and in a language understood all over the civilized world exactly the time and place. The wide adoption of this decimal classification for international use was based on the fact that it was a universal language used with equal readi- ness by all nations, while words would often be confused and misinterpreted as they were trans- ferred from one tongue to another. AssioNTNfi Class Ni-iibeks. The content or the real subject of which a hook treats, not the literarj' form or wording of the title, determines its place. A history of mathematics goes with mathematics, not with historj-. Every book should be assigned to the most specific head that will contain it, according to its predominant I
