In preparing a lecture I find I always have to work hardest on the things I do not say. The things I am sure to say I can easily get up. They are obvious and generally accessible. But they, I find, are not enough. I must have a broad background of knowledge which does not appear in speech. I have to go over my entire subject and see how the things I am to say look in their various relations, tracing out connections which I shall not present to my class. One might ask what is the use of this? Why prepare more matter than can be used? Every successful teacher knows. I cannot teach right up to the edge of my knowledge without a fear of falling off. My pupils discover this fear, and my words are ineffective. They feel the influence of what I do not say. One cannot precisely explain it; but when I move freely across my subject as if it mattered little on what part of it I rest, they get a sense of assured power which is compulsive and fructifying. The subject acquires consequence, their minds swell, and they are eager to enter regions of which they had not previously thought. Even, then, to teach a small thing well we must be large. I asked a teacher what her subject was, and she answered, "Arithmetic in the third grade." But where is the third grade found? In knowledge, or in the schools? Unhappily it is in the schools. But if one would be a teacher of arithmetic, it must be arithmetic she teaches and not third grade at all. We cannot accept these artificial bounds without damage. Instead of accumulated wealth they will bring us accumulated poverty, and increase it every day. Years ago at Harvard we began to discuss the establishment of a Graduate School; and I, a young instructor, steadily voted against it. My thought was this: Harvard College, in spite of what the public imagines, is a place of slender resources. Our means are inadequate for teaching even undergraduates. But graduate instruction is vastly more expensive; courses composed of half a dozen students take the time of the ablest professors. I thought we could not afford this. Why not leave graduate instruction to a university which gives itself entirely to that task? Would it not be wiser to spend ourselves on the lower ranges of learning, covering these adequately, than to try to spread ourselves over the entire field? Doubting so, I for some time opposed the coming of a Graduate School. But a luminous remark of our great President showed me the error of my ways. In the course of debate he said one evening, "It is not primarily for the graduates that I care for this school; it is for the undergraduates. We shall never get good teaching here so long as our instructors set a limit to their subjects. When they are called on to follow these throughout, tracing them far off toward the unknown, they may become good teachers; but not before."