THE LAWS OF SCIENCE 53 



of a substance, the answer is undoubtedly that it has 

 not been recognized sufficiently that such propositions 

 are laws. For they are not usually called laws. But 

 the fact that the name is not applied to them is largely 

 the result of history. As ^we rioted^ laws of_ this type 

 are among the results which science accepts in the 

 first instance from the experience of common sense, 

 although it subsequently refines them and may change 

 them almost beyond recognition. Knowledge is dignified 

 by the imposing name of law only when it has been 

 arrived at by Hftlifafiratft anrUxmscious investigation, and 

 not when, like Topsy, it simply "^growedT* But it is 

 more difficult to explain why numerical laws, to which 

 the name " law " is applied characteristically, have not 

 been recognized as providing instances to^show that 

 causejmd effect is jjoLthej>ri1y relation Wltk^MchJaws 

 are concerned . 



I think the real reason is to be found in a confusion 

 between the method by which knowledge is attained 

 and the content of the knowledge once it is attained. 

 What I mean is this. Suppose we were seeking to 

 discover whether Ohm's Law is true. We shall set up 

 instruments for measuring the current and the pressure, 

 and shall then watch how the current changes when we 

 change the pressure. In . makmg juch^experiments, what 

 we shall actually observe^ that a change in current 

 follows a change in pressure ; we shall first make 

 deliberately a change in the pressure and then observe 

 a change in the current ; in other words, during the 

 experiment the change in current appears as an effect 

 of which the change of pressure is the cause. But, 

 though it may be maintained that it is by observing 

 such relations of cause and effect that we discover 

 the truth of Ohm's Law, it is not these relations which 

 are stated by the law. I It is the numerical relation, and 

 not the relation in time, thai is stated by the law. | For, 



