There is thus no lack of demonstration from all quarters that mental facts and brain processes belong together; and yet, however much we may cumulate such popular and scientific observations, they would never by themselves admit of the sweeping generalization that there cannot be any mental state which is not accompanied by a process in the central nervous system. Someone might say, to be sure, the perceptions and memory images, the volitions and instincts and impulses, have their physiological basis, but there remain after all acts of attention, or decisions, or subtle feelings, or flights of imagination, which are independent of any brain action. Here, indeed, observation cannot settle such a general principle. Its real hold lies in the fact with which we started: there is no causal connection in the mental states as such. If we want to understand mental facts as such in a chain, of causal events, we have first to conceive them as parallel to physical events. The principle of psychophysical parallelism, that is, the principle that every psychical process accompanies a physiological change is thus not a mere result of observation. It is simply a postulate. Every science begins with postulates and only that which fulfills such postulates has the dignity of truth in the midst of that scientific realm. The astronomer cannot find by observation that there is no star the movements of which are not the effects of foregoing causes. He knows it beforehand, he demands it, he does not recognize any movement as understood until he has found the causes, he presupposes that such causes exist, that no star moves simply by a magic power, and that nowhere in the astronomical universe is the chain of causality broken. He postulates it, and where he does not discover the causes, he is sure that he has not solved the real problem.