THK LAWS OF SCIENCE :,i 



as these undoubtedly suggest that by cause and effect 

 we mean something rather more abstruse and certainly 

 more obscure than the simple invariable sequence of one 

 event after another, which seems usually to be regarded 

 as constituting the causal relation. 



But this is not what we have to consider. For those 

 who have seriously maintained that the business of laws is 

 to state relations of cause and effect have always regarded 

 such relations as consisting merely of invariable sequences. 

 It is possible that if the terms are used in this sense they 

 do not coincide exactly with common usage, but, if that 

 is so, it will only be one more of the innumerable examples 

 where science has diverted a term slightly from its sense 

 in popular discourse. What we have to ask is whether, 

 in discovering scientific laws, we are simply establishing 

 in variable sequences in which one event or set of events 

 follows after another. 



It may be admitted without further discussion that 

 some part of the laws of science do actually consist of 

 statements of invariable sequences. So much follows at 

 once from our previous discussion. For, though we have 

 spoken hitherto more vaguely of invariable association 

 rather than of invariable sequence, it is obvious that, if 

 there is such a thing as an invariable sequence, it is 

 one form at least of invariable association and possesses 

 the qualities which we concluded were necessary to make 

 a relation the proper subject-matter of science ; invariable 

 sequence is a relation concerning which universal agree- 

 ment might be obtained, just because it is invariable. 

 On the other hand, it seems certain that there are 



gB as invariable sequences, for it is doubtless within 



the province of science to predict future events, for 



motions of the A the changes of the 



her ; and how could prediction from present to future 



be possible, unless it were possible to discov nces 



n variable and which always rccu 



