It is not needful, nor indeed is it possible, to define Romance. In the mathematical sciences definitions are all-important, because with them the definition is the thing. When a mathematician asks you to describe a circle, he asks you to create one. But the man who asks you to describe a monkey is less exacting; he will be content if you mention some of the features that seem to you to distinguish a monkey from other animals. Such a description must needs be based on personal impressions and ideas; some features must be chosen as being more significant than the rest. In the history of literature there are only two really significant things--men, and books. To study the ascertained facts concerning men and books is to study biography and bibliography, two sciences which between them supply the only competent and modest part of the history of literature. To discern the significance of men and books, to classify and explain them, is another matter. We have not, and we never shall have, a calculus sufficient for human life even at its weakest and poorest. Let him who conceives high hopes from the progress of knowledge and the pertinacity of thought tame and subdue his pride by considering, for a moment, the game of chess. That game is played with thirty-two pieces, of six different kinds, on a board of sixty-four squares. Each kind of piece has one allotted mode of action, which is further cramped by severe limitations of space. The conditions imposed upon the game are strict, uniform, and mechanical. Yet those who have made of chess a life- long study are ready to confess their complete ignorance of the fundamental merits of particular moves; one game does not resemble another; and from the most commonplace of developments there may spring up, on the sudden, wild romantic possibilities and situations that are like miracles. If these surprising flowers of fancy grow on the chess- board, how shall we set a limit to the possibilities of human life, which is chess, with variety and uncertainty many million times increased? It is prudent, therefore, to say little of the laws which govern the course of human history, to avoid, except for pastime, the discussion of tendencies and movements, and to speak chiefly of men and books. If an author can be exhibited as the effect of certain causes (and I do not deny that some authors can plausibly be so exhibited) he loses his virtue as an author. He thought of himself as a cause, a surprising intruder upon the routine of the world, an original creator. I think that he is right, and that the profitable study of a man is the study which regards him as an oddity, not a quiddity.