December 19, 1913] 



SCIENCE 



869 



tial threads of interconnection among them. 

 Therefore he does not seek absolute cer- 

 tainty in his theories, at least when he 

 realizes the fundamental limitations of his 

 methods ; but he understands his theories 

 rather as the most convenient means by 

 which he may summarize for himself and 

 others the actually observed interrelations 

 in nature. 



Now, let us suppose that an experimental 

 scientist attempts to attain absolute cer- 

 tainty in his conclusions, and enquire as to 

 the kind of difficulty which he will en- 

 counter. 



An analysis of the matter shows, first of 

 all, that he must make one fundamental 

 assumption — that involved in the hypoth- 

 esis of the uniformity of nature. If phe- 

 nomena have no laws it is futile to ascribe 

 laws to them ; and therefore a fii-st requisite 

 for the existence of experimental science is 

 the supposition that laws exist. It must be 

 assumed that the universe will not suddenly 

 depart to-morrow from its previous way of 

 behaving ; it must not be a thing of caprice. 



But what ground have I for believing 

 that to-morrow will not put forth a set of 

 phenomena totallj'' different from those 

 which I have observed before? None at all, 

 except what comes through my belief in the 

 uniformity of nature. It is clear that this 

 is not the way by which the principle is to 

 be established. In fact, we can go further 

 and say with confidence that there is no 

 absolute certainty, but only a high degree 

 of probability, that nature is uniform. 



There is also another fundamental as- 

 sumption at the basis of experimental 

 science — one that is curiously related to the 

 mind that has made the assumption. 



A fundamental property of mind is mem- 

 ory ; without it mind can not exist in its 

 usual state. What one does to-day is 

 colored, modified, perhaps determined by 

 one's memory of past acts. No experiment 



on a thinking sub.ject can be performed for 

 the second time ; for the presence of memory 

 in the second event is a factor of determin- 

 ing importance and can not be left out of 

 account. 



And yet mind, of which this is a char- 

 acteristic and fundamental property, has 

 chosen to assume that matter is without 

 memory. If I desire to experiment with a 

 falling stone, I need not enquire whether 

 the stone has gone through the same experi- 

 ence before. In other words, I assume that 

 the stone has no memory of its previous 

 existence ; and consequently its previous 

 historj;- will not affect my present experi- 

 ment. 



If it is true that experimental science is 

 so shot through with basic assumptions, 

 what is to be said of our confidence in its; 

 results? What measure of certainty 

 attaches to them and how do we come to 

 that certainty ? Clearly, the evidence must 

 be indirect ; but it need not on that account 

 be less trustworthy. 



We may arrive at one phase of this evi- 

 dence by noticing what change has taken 

 place in man's relation to natural phe- 

 nomena since the dawn of the modern era in 

 scientific investigation. It is patent to 

 every one that there has been an immense 

 gain in control; man has harnessed the 

 forces of the world and is using them for 

 his purpose. A thousand and one new in- 

 struments of power and pleasure attest to 

 his more profound understanding of the 

 relations among phenomena. For hundreds 

 of miles he can transfer the immense power 

 of Niagara along a slender wire, and then 

 use it to run his machinery and light his 

 cities and warm his houses. In every con- 

 ceivable direction he is making progress 

 decade by decade ; and the momentum of his 

 progress increases as the years pass. 



But even this is not the chief reason for 

 believing that he is essentially right in his 



