SKEPTICISM AND IDOLATRY 197 



signifying a tentative generalization which will ulti- 

 mately be accepted as a law or be rejected because of 

 its increasing disagreement with observations and, un- 

 fortunately, it is used very loosely. It frequently 

 means those speculations in which we indulge when 

 we attempt to define the causes of phenomena; when 

 we create fictitious substances, as an atom or ether; 

 and when we construct models to illustrate the actions 

 of forces. None of these cases is a generalization in 

 the sense of Poincare's first statement, because it can- 

 not be verified by experiment. It would greatly sim- 

 plify thought if we would confine the term hypothesis 

 to these speculative ideas, and use the words law or 

 theory for those generalizations which are based on 

 experimental observations of a material world. One 

 has only to note Poincare's own confusion of thought; 

 if we must generalize, and if our generalizations are 

 hypotheses which are subject to, and must be verified 

 by, experiment, then such hypotheses are the truth, 

 so far as can be determined by scientific methods. We 

 certainly have the right to ask : how can they be dan- 

 gerous and how can there be a too great multiplica- 

 tion of them, if they are the truth? If they can be 

 subjected to experiment, then as soon as one is dis- 

 proved, it would necessarily be abandoned. The reason 

 for his caution to us is that Poincare knows that the 

 great majority are incapable of verification, since they 

 deal with purely metaphysical ideas. His own words, 



