There is not a single truth of science upon which we ought to bet more than about a million of millions to one.