870 



SCIENCE 



[N. S. Vol. XXXVIII. No. 990 



interpretation of the relations of phenom- 

 ena. His strongest ground of confidence is 

 in the multiplicity and the accuracy of his 

 predictions — predictions which he verifies 

 by further tests in the laboratory. 



Probably the severest test of a physical 

 theory is the requirement that it predict 

 accurately a phenomenon which has not yet 

 been observed; and this is a test to which 

 theory is constantly subjected — and it 

 comes out successful. This is the ground 

 of our confidence in physical theories. It 

 is this which lends the strongest possible 

 •credence to such a general hypothesis, for 

 instance, as that of the uniformity of 

 mature. 



This ultimate test of prediction finds its 

 anost extensive exemplification in the results 

 'obtained by the apparatus of abstract 

 mathematical ideas. From a few funda- 

 mental laws, as for instance those of static 

 electricity, an immense body of doctrine is 

 built up by the processes of mathematical 

 analysis. The results so obtained are exact 

 and are stated with careful precision. Not- 

 withstanding their great variety and the 

 absolute precision with which they are 

 stated, they are found to be always in ac- 

 cord with new experiment however the con- 

 ditions may be varied. It is this which fur- 

 nishes our strongest ground of confidence 

 in physical theory ; it is not the argumenta- 

 tion or inference by which the theory was 

 first discovered or created. 



The success of this prediction through 

 mathematical or other argumentation is so 

 great that we can not escape the conclu- 

 sion that science is on the right track; im- 

 provements will come, to be sure, but we 

 have certainly made some fundamental 

 progress. In fact, the ground for this con- 

 clusion is so strong that the burden of 

 proof must rest on whoever disputes its 

 validity. If our theories are essentially 

 erroneous, it requires careful explanation 



to understand why our attempt to put them 

 in mathematical language has issued in 

 such a remarkable success in the way of 

 relating and predicting phenomena. 



Even though we are still left face to face 

 with the conclusion that there is no abso- 

 lute certainty in our scientific theories, we 

 see nevertheless that our ground of confi- 

 dence in them is such as to justify our lay- 

 ing out our life and its activity as if they 

 were so. "We shall accept them as our guide 

 in getting around among external phe- 

 nomena. And we can do this even with 

 more confidence than we can plan those 

 things which depend on our own acts. 

 Indeed there is much greater certainty 

 attaching to the prediction of physical phe- 

 nomena than to the prediction of our own 

 acts; and what more could one reasonably 

 demand of science? 



Now of the two methods which we have 

 ■considered, the mathematical and the 

 experimental-scientific, which is the better? 

 You will probably expect me to say that 

 the mathematical method is the better; but 

 I do not say it. Neither is the better ; the 

 question is meaningless. Each method is 

 of profound importance and each is suited 

 to its proper purposes; each will be im- 

 proved as time passes and will be carried 

 over more and more into all fields of 

 thought and conduct; and each will con- 

 tinue to add new conquests to human 

 achievement. But we shall not say that 

 one is better than the other. 



Most of you to whom I have spoken this 

 evening are at the threshold of life. The 

 future lies before you. You will doubtless 

 choose some definite work to do in it. "Would 

 you like to have a part in promoting those 

 fundamental ends of human development 

 which may be secured through the use of 

 one or the other of these great methods of 

 advancement ? 



But what is it to have a part in using 



