=An Aid to the Librarian.=—Another person who is greatly benefited by classification is the librarian, and it is just as important that he be helped as that the reader be helped. He is, however, helped in a different way. He knows what the system of classification in use in the library is and with the outlines of this scheme in mind he goes through the library and finds out where it is strong and where it is weak and can plan future purchases accordingly. If, for example, he finds on the shelves little of value on photography he will make a note of it and buy more books on that subject when funds are available. If he finds that there is an undue supply of travel on hand, he will note that also and buy fewer books in that class in the future. Without the help of classification the librarian would overlook many such irregularities. In an unclassified library they would be discovered only through a long and tedious investigation. His only recourse would be the catalog and that is not so well adapted to answer such questions. =Basis of Classification.=—The next question is, what shall be the basis of classification. It is obvious that this basis should be sought in the character of the books themselves and should be applied with constant reference to the reader and his needs. In regard to the first point, character of the books, we know that books have been written on all kinds of subjects,—religion, law, history, medicine, etc. and that those subjects form the only rational basis for classification. A classification based on these distinctions is the only one that helps the reader. If a man comes to the library to investigate a particular point in medicine, it is clear that it will help him if he finds all the medical books together rather than all the books grouped according to their date of purchase by the library. =Present Tendency.=—Many schemes have been devised for the classification of books, some very simple, others extremely elaborate. The present tendency is to adopt the more elaborate classification. Formerly most libraries were not classified at all, but the books were arranged in the order in which they were received, the only grouping of the books being in such cases one based on size,—folios in one place, quartos in another, in order to save shelf room. Assuming that the books in the library were numbered according to the date of their accession from one up to, say, twenty thousand, it is clear that the reader could find a book by a particular writer quickly enough by looking up its number in the catalog, but if he had wished to consult thirty books on one subject, it would be a very tedious operation and most readers would not take the time for it.