In complete contrast with such a view-point, the empiricist came in the end to make experience the sole test of truth, even of such truth as the mathematician possessed. If the issue is between taking thought respecting all things with the rationalist, or everywhere trusting to observation with the empiricist, it is clear the latter has plausibility on his side. Who, closing his eyes and reasoning it out, could learn that there were just eight planets, and not seven or nine? If we must do one thing or the other exclusively, is it not easier to imagine that the axioms of geometry embody the experience of the ages and nothing more, than to suppose that equipped only with the pure reason, _i.e._, with the principles of logic, one could discover the one thinkable world to be that in which a person that is “I” should exist with a sheet of paper this moment before him and a fly buzzing by his ear? So it seemed more and more as empiricism was developed at the hands of Locke, Berkeley and Hume. Belief, we find Hume maintaining in the end, is all of a kind; it is the inference from an actual impression (A) to an expected impression (B), based on the remembered experience that A has always in the past been followed by B. Since this past experience is limited and since the remembrance of it may be defective, the belief based on the two can never amount to certainty. Such an attitude may well be called sceptical when contrasted with the older rationalism, in that it denies the possibility of complete certainty in any field of science, substituting as the ideal of scientific evidence an ever-increasing balance of probability in favor of the opinion we are constrained to accept. But though to think of our body of accepted opinion after this manner is to induce an extreme flexibility of the imagination, which must be prepared to conceive that the firmest truth may be untrue and has only a more or less inadequate array of facts behind which to defend itself, yet it does not follow that nature is a fantastic dream, without order and coherence. Indeed, _that_ experience which is to be our guide from now on, assures us of just the contrary, and the new evidence that would be required to make us admit that an event in exception to any well-founded law had really occurred would have to be overwhelming. * * * * * Nowhere does Hume’s faith in the evidence upon which the best tested uniformities of experience base their claim to acceptance as nature’s laws, show itself more clearly than in his treatment of miracles. To an analysis of the evidence for such miracles as history records he devotes an entire section of his “Enquiry Concerning Human Understanding,” (1748):