THE APPLICATIONS OF SCIENCE 179 



general misunderstanding of statistical laws, renders 

 them peculiarly liable, not only to their special fallacies, 

 but to all those which we have discussed before. False 

 theories and prejudices lead men to notice only those 

 instances which are favourable to the law they want to 

 establish ; they fail to see that, if it were a true law, a 

 single contrary instance would be sufficient to disprove 

 it, while, if it is a statistical law, favourable instances 

 prove nothing, unless contrary instances are examined 

 with equal care. They forget, too, that a statistical 

 law can never be the whole truth ; it may, for the time 

 being, represent all the truth we can attain ; but our efforts 

 should never cease, until the full analysis has been 

 effected, and the domain of strict law carefully separated 

 from that of pure chance. The invention of that method 

 of analysis, which leads to a possibility of prediction and 

 control utterly impossible while knowledge is still in the 

 statistical stage, is one of the things which makes science 

 indispensable to the conduct of the affairs of practical 

 life. 



CONCLUSION 



Our examination of the errors into which uninstructed 

 'reasoning is -liable to 'mislead mankind, affected by so 

 many prejudices and superstitions, shows immediately 

 how and why science is indispensable, if any valuable 

 lessons are j to , be drawn from the most ordinary 

 experience, jln the first place, science brings to the 

 analysis of such experience the .conception of definite, 

 positive, and fixed law. For the vague conception of a 

 law as a predominating, though variable, association, 

 always liable to be distorted by circumstances, or even, 

 like the laws of mankind, superseded by the vagaries of 

 some higher authority, science substitutes, as its basic 

 and most important conception, the association which 

 13 



