DECEMBER 11, 1914] 



SCIENCE 



837 



to their degree of combustibility. The opera- 

 tion of burning was simply equivalent to the 

 liberation of the phlogiston. This theory 

 dominated chemical thought for more than a 

 generation, notwithstanding its inherent de- 

 fect due to the fact that the products of com- 

 bustion were heavier than the original sub- 

 stance, whereas the theory demanded that they 

 should be lighter. 



I have purposely illustrated the influence 

 of preconceived opinion by means of some of 

 the older examples. Many others might be 

 given. In fact, in nearly all our theories rel- 

 ative to experimental phenomena we intro- 

 duce important elements not suggested by 

 our observations, but by our own esthetic sense. 

 Witness the introduction of the ether in so 

 much of physical theory. A man sometimes 

 feels that he is putting into his theory noth- 

 ing except what observation has directed. 

 This, I believe, is always a delusion. More- 

 over, I think that it is an undesirable thing 

 to attempt. It is not true that observations 

 compel any one theory. In fact, as Poincare 

 has shown, there is an infinite number of ex- 

 planations of any finite set of facts. From 

 among this enormous totality we must select 

 the explanation which is most satisfying for 

 us from considerations of convenience or from 

 the demands of the esthetic sense. This is 

 actually what we always do. It should be 

 done consciously. 



Now it is clear that any body of doctrine 

 built up in this way is in danger of being 

 seriously in error, and therefore it is neces- 

 sary for us often to reexamine our theories 

 with a view to ascertaining whether the pre- 

 conceptions which were wrought into them 

 still appear to be justifiable. This is one of 

 the hardest tasks of the scientist. Accord- 

 ingly he often waits long in the presence of 

 his difficulties before he tries to overcome them 

 by this heroic method. He is usually more 

 averse to the surgical knife operating among 

 his ideas than on the members of his body, 

 however hard he may try to overcome this 

 disposition. 



It is no surprise that this is so. The race 

 was too long practical before it sought to be- 



come scientific for us to make the change 

 readily. Some one has defined the practical 

 man as one who practises the errors of his 

 forefathers. He is tied down to his precon- 

 ceived opinions, not being enough of a 

 dreamer to get away from them. He will be 

 able to get through the world without receiv- 

 ing many hard knocks; but he will not inau- 

 gurate profound changes and advances in hu- 

 man life. That will always be left for the 

 scientist who refuses to be satisfied with what 

 is and who is always seeking a new sort of 

 fact to destroy his own and his contempo- 

 raries' equilibrium. 



But this will be harder for him to do as the 

 years pass. In fact it is true in one respect 

 that the problems of the scientist are increas- 

 ing in difiSculty. As the mass of accumulated 

 observations grows larger there are fewer es- 

 sentially new facts to be discovered. And 

 when it becomes necessary to devise a new 

 theory it is harder to make it fit into and 

 explain the great array of recorded phenom- 

 ena. But this ailords no ground for pessi- 

 mism, as we shall show in a moment. More- 

 over, it carries with it a reward of its own. 

 If a theory can be made to fit into the facts 

 as now known it has a good chance of doing 

 service for some time, and this from the rea- 

 son that it has been made to explain so many 

 things already. 



But there was a real advantage to be gained 

 from the meagerness of data in the old time. 

 It was not so difficult to theorize with some 

 appearance of success, and therefore men the 

 more readily conceived the possibility of re- 

 lating things according to law and the more 

 easily set up a tentative explanation. I have 

 no doubt that speculative philosophy, for in- 

 stance, has profited in times past by the 

 meagerness of the data on which its specula- 

 tions were based. The very fact that no large 

 body of observed occurrences stood in the way 

 of speculation emboldened men to launch 

 forth upon what otherwise would have been a 

 forbidding sea. 



But confidence in setting forth did not save 

 from danger and shipwreck. For some time 

 we have known that no conclusion in science 



