88 SCIENCE AND METHOD. 



strength of the calculation that the opposition would 

 never have a single representative. 



The laws of chance do not apply to these questions. 

 If justice does not always decide on good grounds, 

 it does not make so much use as is generally supposed 

 of Bridoye's method. This is perhaps unfortunate, 

 since, if it did, Condorcet's method would protect us 

 against miscarriages. 



What docs this mean ? We are tempted to attribute 

 facts of this nature to chance because their causes 

 are obscure, but this is not true chance. The causes 

 are unknown to us, it is true, and they are even 

 complex ; but they are not sufficiently complex, since 

 they preserve something, and we have seen that this 

 is the distinguishing mark of " too simple " causes. 

 When men are brought together, they no longer 

 decide by chance and independently of each other, 

 but react upon one another. Many causes come into 

 action, they trouble the men and draw them this way 

 and that, but there is one thing they cannot destroy, 

 the habits they have of Fanurge's sheep. And it is this 

 that is preserved. 



X. 



The application of the calculation of probabilities 

 to the exact sciences also involves many difficulties. 

 Why are the decimals of a table of logarithms or of 

 the number tt distributed in accordance with the laws 

 of chance? I have elsewhere ."studied the question 

 in regard to logarithms, and there it is easy. It is 

 clear that a small difference in the argument will give 

 a small difference in the logarithm, but a great differ- 

 ence in the si.xth decimal of the logarithm. We still 

 find the same criterion. 



