The library problems of the 20th century sum themselves up in one, the problem of expansion, and we may perhaps best regard them from the point of view of the obstacles to expansion, these obstacles constituting the problems. First, we must notice our library buildings, and admit that many of them, and most of the ideas heretofore cherished about the building of libraries, present such an obstacle. When we note that since the plans were drawn on which nearly all of our most recent large library buildings have been erected, three new ideas in library administration have come into general acceptance which must powerfully affect library construction, we can but feel that great foresight and wisdom are needed to erect libraries that shall not very soon be obstacles to proper and necessary expansion. These three new ideas are, first, access of readers to the bookshelves; second, children's rooms, and third, the distribution of books through schools, branches, delivery stations, home libraries, and inter-library loans, this third new idea involving provision for business offices, packing rooms, etc., unthought of formerly. To meet not simply these new ideas, but others with which the new century is pregnant, care must be taken that great sums of money, leaving the securing of more for a long time hopeless, are not expended on structures in which instead of provision for expansion we seem to have provision against it. Another obstacle to expansion is found in elaborate systems of shelf-marks connected with systematic schemes of classification, representing carefully arranged subordination and co-ordination of the parts. For two things are certain: first, accepted classifications of books rapidly become obsolete, and second, no library will long be content with an out-of-date arrangement. Especially will my successor, or yours, be sure to feel the necessity of signalizing his accession to office by introducing what is in his day the latest classification. And in this he will be right. Now, if we have a fair sense of our duty to our successor, which is merely an extension forward of our duty to the library itself, we shall be unwilling to tie the library by an intricate notation to a present system of classification. I think we must take more pains than is done by either the Decimal or Expansive schemes to provide a somewhat elastic notation. I regard the classification of the University of California Library as the best (available in print) for libraries of our class, because it employs designations which indicate mere sequence of classes. A little thought will, I am sure, show you how this is true. At any rate, a little experience in attempting more or less reclassification with, for example, the Decimal classification, will prepare you to believe that a less highly involved and articulated method of designation would be in the interest of reasonable expansion, and save such expansion from the odium of upsetting the classification.