It is possible, but not, I hope, probable, that some hasty reader may suppose that in this and the preceding lectures I am recommending a new method or instrument of discovery. “If you want to reach truth, follow your unreasoned inclination,” may be his summary of my doctrine: brief—but also unjust. Of the manner in which discoveries are going to be made I say nothing, for I know nothing. I am dealing with the past: and in the historic movements of scientific thought I see, or think I see, drifts and currents such as astronomers detect among the stars of heaven. And, as the law of gravitation will hardly (I suppose) explain the last, so observation, experiment, and reasoning will hardly explain the first. They belong to the causal, not to the cognitive, series; and the beliefs in which they issue are effects rather than conclusions. Those who feel little sympathy for such a view may be inclined to regard the relatively faint inclinations dealt with in this lecture as ordinary scientific hypotheses confirmed by ordinary scientific methods. This view, as I have already observed, is not applicable to the inevitable beliefs dealt with in earlier lectures. Whatever philosophers may say after the event the conviction that we live in an external world of things and persons, where events are more or less regularly repeated, has never been treated as a speculative conjecture about which doubt was a duty till truth was proved. Beliefs like these are not scientific hypotheses, but scientific presuppositions, and all criticism of their validity is a speculative after-thought. The same may be said, though with less emphasis and some qualification, about beliefs fostered by the intellectual tendencies considered in this chapter. These, as we have seen, are many. They are often inconsistent; they are never inevitable; and they perpetually change their form under the pressure of scientific discovery. Atomism in one shape follows atomism in another; doctrines of conservation rise, fall, and rise again; incredulity about “action at a distance” breeds explanations whose failure (in the case of gravity) leaves the hope of final success untouched.