Moreover, in subjects which belong to moral proof, such, I mean, as history, antiquities, political science, ethics, metaphysics, and theology, which are pre-eminently such, and especially in theology and ethics, antecedent probability may have a real weight and cogency which it cannot have in experimental science; and a mature politician or divine may have a power of reaching matters of fact in consequence of his peculiar habits of mind, which is seldom given in the same degree to physical inquirers, who, for the purposes of this particular pursuit, are very much on a level. And this last remark at least is confirmed by Lord Bacon, who confesses "Our method of discovering the sciences does not much depend upon subtlety and strength of genius, but lies level to almost every capacity and understanding;"[113:1] though surely sciences there are, in which genius is everything, and rules all but nothing. 4. It will be a great mistake then to suppose that, because this eminent philosopher condemned presumption and prescription in inquiries into facts which are external to us, present with us, and common to us all, therefore authority, tradition, verisimilitude, analogy, and the like, are mere "idols of the den" or "of the theatre" in history or ethics. Here we may oppose to him an author in his own line as great as he is: "Experience," says Bacon, "is by far the best demonstration, provided it dwell in the experiment; for the transferring of it to other things judged alike is very fallacious, unless done with great exactness and regularity."[113:2] Niebuhr explains or corrects him: "Instances are not arguments," he grants, when investigating an obscure question of Roman history,--"instances are not arguments, but in history are scarcely of less force; above all, where the parallel they exhibit is in the progressive development of institutions."[113:3] Here this sagacious writer recognizes the true principle of historical logic, while he exemplifies it. The same principle is involved in the well-known maxim of Aristotle, that "it is much the same to admit the probabilities of a mathematician, and to look for demonstration from an orator." In all matters of human life, presumption verified by instances, is our ordinary instrument of proof, and, if the antecedent probability is great, it almost supersedes instances. Of course, as is plain, we may err grievously in the antecedent view which we start with, and in that case, our conclusions may be wide of the truth; but that only shows that we had no right to assume a premiss which was untrustworthy, not that our reasoning was faulty. 5.