"[H]e sinned the sin of excessive generalization. It may be true that in certain cycles or shorter periods of time the sums of human acts are strangely alike. It may be true also that statistics afford to history one of its most sure and instructive auxiliaries. But it is no less certain that such tabular records are not only in their infancy, but as regards former times, either do not exist, or are most scanty and precarious aids to truth. At the best, also, they represent a few only of the elements of social life, and probably centuries of exact observation must elapse before they can be permitted to supersede the other grounds, moral, intellectual, and religious, on which history hitherto as been constructed."