At the outset it should be said that the importance of this study of beginnings is in every science quite out of proportion to the importance of the objects studied. Beginnings are by nature small. The highest and best things are by nature the most complex and latest, but the study of the earliest and simplest libraries, like the study of the simplest cell life, is not only useful from several points of view but vital to a right understanding of the more complex. The great vice of technical education of all sorts is its tendency to fix attention on the latest and best only. It is true of course that man’s ideas and methods are an evolution—just as his body is. The fact of the accumulation of human experience is the central significant fact of human civilization. It is the glory of libraries that by reason of this fact they are an indispensable tool of progress in civilization. On the whole, by and large, the latest ideas are in fact best, for they tend to sum up in themselves the total of the useful variations of all preceding ideas, and the main time and attention of a course of education must of necessity therefore be given to the latest and best experience, because it does sum up all that has gone before. This does not, however, lessen the value of the study of earlier ideas on any subject back to the very beginnings, for at any given time and place, the latest idea or method in any field is not necessarily the best. It might be the best: it is in position to build on all previous experience and so become best. We all know, however, that the latest book on a subject is not always the best book. So it is, too, of individual ideas or methods. This frequent failure of the latest to be best comes chiefly from lack of knowledge of previous experience. Every year sees library methods put in operation which were tried and found wholly wanting in the last century or it may be, two, three or even five thousand years ago. On the other hand again, every now and then we find that some method or idea, discovered long ago but neglected meantime, is far better than those in common use. This has often been true of great scientific ideas and we have in Mendelism a striking recent example. One must needs therefore study earlier ideas in any field, both in order to be sure that so-called new ideas are not exploded old ones and in order to find whether common practice in any field at a given time is not really the development of an inferior line of evolution.