I have been speaking of such books as should be chosen for profitable reading. A public library, of course, must be far wider in its scope. It should contain something for all tastes, as well as the material for a thorough grounding in all branches of knowledge. It should be rich in books of reference, in encyclopædias, where one may learn without cost of research what things are generally known. For it is far more useful to know these than to know those that are _not_ generally known. Not to know them is the defect of those half trained and therefore hasty men who find a mare's nest on every branch of the tree of knowledge. A library should contain ample stores of history, which, if it do not always deserve the pompous title which Bolingbroke gave it, of philosophy teaching by example, certainly teaches many things profitable for us to know and lay to heart; teaches among other things how much of the present is still held in mortmain by the past; teaches that, if there be no controlling purpose, there is, at least, a sternly logical sequence in human affairs, and that chance has but a trifling dominion over them; teaches why things are and must be so and not otherwise; teaches, perhaps, more than anything else, the value of personal character as a chief factor in what used to be called destiny, for that cause is strong which has not a multitude but one strong man behind it. History is indeed mainly the biography of a few imperial men, and forces home upon us the useful lesson how infinitesimally important our own private affairs are to the universe in general. History is clarified experience, and yet how little do men profit by it--nay, how should we expect it of those who so seldom are taught anything by their own! Delusions, especially economical delusions, seem the only things that have any chance of an earthly immortality. I would have plenty of biography. It is no insignificant fact that eminent men have always loved their Plutarch, since example, whether for emulation or avoidance, is never so poignant as when presented to us in a striking personality. Autobiographies are also instructive reading to the student of human nature, though generally written by men who were more interesting to themselves than to their fellow-men. I have been told that Emerson and George Eliot agreed in thinking Rousseau's "Confessions" the most interesting book they had ever read. A public library should also have many and full shelves of political economy, for the dismal science, as Carlyle called it, if it prove nothing else, will go far toward proving that theory is the bird in the bush, though she sing more sweetly than the nightingale, and that the millennium will not hasten its coming in deference to the most convincing string of resolutions that were ever unanimously adopted in public meeting. It likewise induces in us a profound distrust of social panaceas.