THE LAWS OF SCIENCE 45 



found relations between the objects of which they are 

 the laws, and so arrived at new and more complex laws. 

 In the second process, science takes these simple laws. I 

 analyses them, shows that they are not truly laws, and ( 

 divides them up into a multitude of yet simpler laws.) 

 These two processes have been going on concurrently* 

 throughout the history of science ; in one science at one 

 time one of the processes will be predominant ; in another 

 science at another time, the other. But on the whole 

 the first process is the earlier in .time. Science started, 

 as weFhave seen, from" the ordinary everyday knowledge 

 of common sense. Common sense recognizes kinds of 

 objects and kinds of events, distinguished from particular 

 objects and particular events by the feature we have just 

 discussed ; they imply trie, assertion of a law. Thus all 

 " substances," iron, rust, water, air, wood, leather, and 

 so on, are such kinds of objects ; so again are the various 

 kinds of animals, horses, sparrows, flies, and so on. Simi- 

 larly common sense recognizes kinds of events, thunder 

 and wind, life and death, melting and freezing, and so 

 on ; all such general terms imply some invariable asso- 

 ciation and thus are, if the association is truly invariable, 

 proper matter for the study of science. And science in 

 its earlier stages assumed that the association was invari- 

 able, and on that assumption proceeded to build up laws 

 by the first process. It found that iron in damp air 

 produced rust ; that poison would cause death. 



But, as soon as this process was well under way, the 

 second process of analysis began ; it was found that the 

 association stated by the laws implied by the recognition 

 of such objects was not truly invariable. This discovery 

 was a direct consequence of the first process. Thus, 

 until we have discovered that steel in general rusts, we are 

 not in a position to notice that there are some steels 

 which do not rust. When we have found that there are 

 certain substances, otherwise like steel, but differing 



