150 



Life and Letters of Francia Galton 



relation'" between smallness of head and energy. Now if we distribute 

 40 cases uniforndy among Galton's 99 individuals, we Hud : 



A modern statistician would not be quite happy in asserting without 

 further investigation that the owners of large heads were the less energetic. 

 But actually the odds against {li) l)eing a sample of (C) are about 50 to 1 and 

 Galton's conclusion is reasonably justified. I have cited these results be- 

 cause, I think, Galton's statistical method does raise doubts in the minds of 

 some readei-s as to its axlequacy to support the superstructure'. I have 

 been surprised myself on testing his data to find what a reasonable prob- 

 ability there is for the validity of his conclusions. After all the higher 

 statistics are oidy common-sense reduced to numerical ap])reciation, and 

 their deductions are what "les esprits justes sentent par une sorte d'instinct, 

 sans qu'ils puissent souvent s'en rendre compte." Galton's judgments in liis 



for the scientists wh-s 22-"5" + 050, whicli is not a sifjnificant diffi-ronce, 

 Galton's })ocket-book. The following are perhaps worth extracting: 



1 deduce these from 



' This is the first occasion on which I have noticed Galton using this word, of course not 

 yet with reference to the 'coefficient of correlation'. 



* I feel considerable doubts myself as to the validity of the map of the distribution of 

 birthplaces of 100 English men of science on p. 20. It may moan nothing more than a rough 

 density of population map. The local distribution of each typo of ability in the British Isles 

 with reference to density of |)opulation would be well worth undertaking. 



