96 INHERITANCE IN ANIMALS 



experiences, with very fair accuracy, by a theory which 

 enables us to predict the general characters of any long 

 series of similar experiences. 



Now, if we attempt to describe our experience of any 

 character of living things, we find the same difficulties 

 with individual cases, the same regularity when we deal 

 with large series of cases, as we find when we attempt to 

 describe the results of games of chance. In each case we 

 are unable to predict the result of any single observation, 

 before we have made it, but we can predict, with very 

 considerable accuracy, the result of a long series. 



The first important difference between the methods 

 used by most astronomers, by most chemists and physi- 

 cists, and the methods which ought to be used by students 

 of biology, is determined by this fact that while astrono- 

 mers, physicists, and chemists have by great and success- 

 ful efforts reduced the limits of uncertainty concerning 

 the results of their observations until for many purposes 

 they can neglect the discrepancies between the results of 

 individual experiments, and treat their experience as 

 uniform, biologists have not yet gone so far, and they are 

 still forced to base such general statements as they can 

 make on the characters of long series of observations. 



The method by which long series of variable results 

 can be described, in such a way that the mind can easily 

 remember and form a useful picture of each series as a 

 whole, is provided by the science of statistics, and the 

 application of these methods to biological problems, com- 

 menced by Quetelet, has been enormously extended of 

 late years by the work of Francis Galton and Karl 

 Pearson. The details of the method are logically un- 

 important, and I shall not trouble you by asking you to 

 consider them, except in so far as it is necessary to do so 

 in order to put before you one of the generalizations 



