THE BEGINNINGS OF SCIENCE 37 



If under new conditions what he predicts from his 

 theory fails to happen, he knows that it is wrong some- 

 where. He therefore modifies it by changing some 

 assumptions or adding new factors until he can account 

 for the phenomenon which invalidated the theory. 

 It may be that the theory is shown to be absolutely 

 wrong, so that no modification can possibly make it 

 explain the things that actually happened. In that 

 case it must be discarded at once. 



As time goes on new conditions arise and the theory 

 is subjected to new tests. Sometimes it is not neces- 

 sary to wait until new conditions arise naturally, for 

 they may be devised and arranged. A scientist thus 

 checks his theory by experiment. It may be that he 

 can form a rule, expressing what to expect under any 

 given set of conditions. This relation between cause 

 and effect he states as a law. Sometimes the experi- 

 ments will show unexpected relations, not inconsist- 

 ent with the theory and yet not covered by it. The 

 scientist may then verify this relation in other experi- 

 ments and make a statement of it as a purely empirical 

 law. 



By theories formulated on the basis of experiment, 

 by experiments devised to test them, and by their 

 consequent verification, modification, or rejection, 

 science has advanced. It has developed through a 

 succession of theories. The man who furthers it, 

 except by some accidental discovery, is the one who has 

 imagination for the hidden processes by which natural 

 phenomena occur. 



The danger to science is not in the suggestion of new 

 theories but in the failure of their adherents to follow 



