I propose to bring fresh evidence here for my theory of knowledge and expand it in new directions. We shall arrive most swiftly at the centre of the theory, by going back to the point from which I started about twenty years ago. Upon examining the grounds on which science is pursued, I saw that its progress is determined at every stage by indefinable powers of thought. No rules can account for the way a good idea is found for starting (...) an inquiry; and there are no firm rules either for the verification or the refutation of the proposed solution of a problem. Rules widely current may be plausible enough, but scientific enquiry often proceeds and triumphs by contradicting them. Moreover, the explicit content of a theory fails to account for the guidance it affords to future discoveries. To hold a natural law to be true, is to believe that its presence may reveal itself in yet unknown and perhaps yet unthinkable consequences; it is to believe that such laws are features of a reality which as such will continue to bear consequences inexhaustibly. (shrink)