The first step towards natural science is, therefore, to dissolve the real experience into thing and personality; that is, into object and active subject, and to eliminate in an artificial abstraction the activity of the subject, making the object material of merely passive awareness, and related no longer to the will but merely to other objects. It may be more difficult to understand the second step which naturalism has to take before a natural science is possible. It must dissolve the object of will into an over-individual and an individual part and must eliminate the individual. That part of my objects which belongs to me alone is their psychical side; that which belongs to all of us and is the object of ever new experience is the physical object. As a physicist, in the widest sense of the word, I have to ignore the objects in so far as they are my ideas and have to consider the stones and the stars, the inorganic and the organic objects, as they are outside of me, material for every one. The logical purpose of this second abstraction may be perhaps formulated in the following way. We have seen that the purpose of the study of the objects is to find out what we have to expect from them; that is, to what effects of the given thing we have to submit ourselves in anticipation. The ideal aim is thus to understand completely how present objects and future objects--that is, how causes and effects--are connected. The first stage in such knowledge of causal connections is, of course, the observation of empirical consequences. Our feeling of expectation grows with the regularity of observed succession; yet the ideal aim can never be fulfilled in that way. The mere observation of regularities can help us to reduce a particular case to a frequently observed type, but what we seek to understand is the necessity of the process. Of course we have to formulate laws, and as soon as we acknowledge a special law to be expressive of a necessity, the subsumption of the particular case under the law will satisfy us even if the necessity of the connection is not recognized in the particular case. We are satisfied because the acknowledgment of the law involved all possible cases. But we do not at all feel that we have furnished a real explanation if the law means to us merely a generalization of routine experiences, and if thus no absolute validity is attached to the law. This necessity between cause and effect must thus have its ultimate reason in our own understanding. We must be logically obliged to connect the objects in such a way, and wherever observation seems to contradict that which is logically necessary, we must reshape our idea of the object till the demands of reason are fulfilled. That is, we must substitute for the given object an abstraction which serves the purpose of a logically necessary connection. That demand is clearly not satisfied if we simply group the totality of such causal judgments under the single name, Causality, and