THE LAWS OF SCIENCE 49 



is important because it gives a reason why we do actually 

 select for the study of science those portions of experience 

 which are actually selected ; but it is not the criterion 

 which we consciously apply. The conscious criterion 

 for the subject-matter of science is rather that it has been 

 regarded hitherto as connected together by a relation 

 of invariable association such as is asserted by a law. 



DO LAWS STATE CAUSES AND EFFECTS ? 



So far we have only considered half of the problem of 

 the laws of science. We have expanded and made more 

 precise the conception of a law of nature, have considered 

 such laws are of such supreme importance for science, 

 and have inquired how they can be at once its starting 

 point and its goal. A law, we have concluded, is the 

 assertion of an invariable association, and the events or 

 properties or other things that it declares to be invariably 

 associated are themselves collections of other invariably 

 associated things. But we have not attempted to ask 

 fun t is meant by " invariable association." We 



noticed in passing at the outset that it was often thought 

 that laws were concerned characteristically with rela- 

 s of cause and effect. A cause and its effect are invari- 

 ably as- The view is therefore suggested that 

 by invariable association we mean simply the relation of 

 a cause to its effect. Is that what we mean ? This is 

 r problem and to it the rest of the 

 chapter must be devoted. 



We must naturally start by asking ourselves what 



i (>r what we should mean) by "cause 



and is is a matter on which there has been 



which underlies most 



;<c of the terms seems to b We 



B happens, it happens 



only oecaust b^n pi- y some vent 



