I shall take up these points in order. First, the needs of the community. These are not necessarily to be measured by its demands, otherwise the librarian’s labor would be considerably lightened. Unfortunately, when a community needs a given class of books very desperately it is often serenely unconscious of the fact. To the librarian falls the task not only of determining what the need is and of filling it, but also of arousing a wholesome consciousness of it. In this educational work he may be, and often is, aided by the teacher, the clergyman, or even by the users of the library themselves. Hence the importance of getting in touch with all the agencies that may do work along this line. There is nothing that calls for more tact. With the children it is comparatively easy to point out a deficiency, but a direct attempt with a self-respecting adult may end in disaster, and a season or two of well-meant effort may result in weakening the librarian’s influence or even in losing him his position. But one can rarely teach tact to the tactless, and tact is something that every librarian must have, so that this lopping-off process, after all, may simply be regarded as a phase of nature’s elimination of the unfit. One way of ascertaining the proportional demand for various classes of literature in a community, is by examining the class-percentage of circulation. By comparing these with the corresponding volume percentages we may see whether the demands of the community are being met, and by comparison with the percentages of an ideal library we may see whether such demand ought to be met or not. Of course, the ideal is somewhat indefinite. One may accept the suggested proportions in the A.L.A. catalog, or average those of several libraries of high class; or one may construct an ideal of one’s own. In any case, the ideal proportions will evidently vary with conditions of place and time. To show how this test may be applied, consider the percentage of science circulated last year in the New York Public library. This varied from 3 to 28 per cent in the various branches, and was 9 per cent for the whole library. The percentage of science on the shelves similarly varied from 6 to 18 per cent, and was also 9 for the whole library. In our library sociology and philology are included in the science report, and the percentage of these three classes combined in the old A.L.A. catalog is 17. If this is to be taken as the standard, therefore, the library as a whole falls below it, though individual branches approach or even exceed it. As a whole, however, the demand and the supply balance pretty well. There is no doubt, however, that in this and most other libraries the demand in this class is too small and needs stimulation. Of course, this is brought up merely as an instance of how fertile this comparison of percentages is in information, and how valuable in ascertaining whether the demands of a community are supplied, and whether they ought to be