70 SCIENCE AND METHOD. 



impulse we give to the needle. I assume that the 

 needle will make ten or twenty revolutions, but it 

 will stop earlier or later according to the strength 

 of the spin I have given it. Only a variation of a 

 thousandth or a two-thousandth in the impulse is 

 sufficient to determine whether my needle will stop 

 at a black section or at the following section, which 

 is red. These are differences that the muscular sense 

 cannot appreciate, which would escape even more 

 delicate instruments. It is, accordingly, impossible for 

 me to predict what the needle I have just spun will 

 do, and that is why my heart beats and I hope for 

 everything from chance. The difference in the cause 

 is imperceptible, and the difference in the effect is 

 for me of the highest importance, since it affects my 

 whole stake. 



III. 



In this connexion I wish to make a reflection that 

 is somewhat foreign to my subject. Some years 

 ago a certain philosopher said that the future was 

 determined by the past, but not the past by the 

 future ; or, in other words, that from the knowledge 

 of the present we could deduce that of the future 

 but not that of the past ; because, he said, one cause 

 can produce only one effect, while the same effect can 

 be produced by several different causes. It is obvious 

 that no scientist can accept this conclusion. The laws 

 of nature link the antecedent to the consequent in 

 such a way that the antecedent is determined by the 

 consequent just as much as the consequent is by the 

 antecedent, liut what can have been the origin of 

 the philosopher's error ? We know that, in virtue 

 of Carnot's principle, physical phenomena are irrevers- 



