To the financial section of this discussion belongs also the question of editions. Shall the librarian choose the best or the cheapest? Which is the best and which is the cheapest for his purpose? In the first place, we may exclude the extremes. Editions de luxe have no place in the ordinary free library, and, on the other hand, we should not think of offering to a self-respecting reader books printed on bad paper with worse type, simply because they can be purchased at a phenomenally low figure. But between these two there are many grades of beauty and durability. Here, as elsewhere, there is safety in the golden mean. As far as bindings of exceptional durability go, the question of paying extra for them depends on the use that is to be made of the book. If it will circulate so little that the ordinary binding will last twenty years, why spend money for anything stronger? Again, if it get such hard treatment that it must be replaced in a year’s time, why put on it a binding that would outlive ten years of such vicissitudes? Still again, with current books of popular interest, the library cannot wait to have them put into special bindings, but for standard, popular works, which will have steady but not hard use, and which can be ordered three months before they are to be used, money spent on special bindings may be economy in the end. Here, however, we are drifting a little way from our subject. The three points that we must take into consideration in selecting books, namely, the community’s need, the determination of what books will satisfy it, and the consideration of how far the library’s financial condition will allow it to go in that direction, have been treated separately, but it must be evident that they are in reality so closely connected that they act and react on each other. No one of them can in practice be considered apart from the others. Thus the first necessity of the library may be books on music, and a secondary need may be books on water supply. It may so happen, however, that a complete and up-to-date work on the latter subject, we will say, has just been issued at a moderate price, while the works on music most needed are expensive. The result would be quite different from that reached by a consideration of the first point alone. Again, we will take the case of a large library with a book appropriation large enough to buy practically all that it wants in current literature. This fact drops point third out of consideration entirely and modifies both the others considerably. If the library wants both music and hydraulics, and has money enough for only one, we must consider carefully which can best be spared; but if the funds are at hand for both, all this thought is not needed. In like manner, even if there are funds for both, but only for one or two books on each subject, we must select the books we need most, which we need to do if we have money to buy all we want on both subjects.