56 Heredity, Variation and Genius 



While admiring the zeal and industry of the 

 students of the new science of biometry founded 

 by Karl Pearson and hoping much from the 

 diligent pursuit of its exact method, it is easy 

 to over-rate the values of its present labours 

 to obtain statistical averages reducible to exact 

 mathematical formulas ; pardonable perhaps to 

 feel an uncomfortable distrust of statistical calcu- 

 lations of ancestral resemblances. How safely 

 trust the applications of measurements to such 

 vague and uncertain characters or compositions 

 of characters as vivacity, introspection, temper, 

 conscientiousness, assertiveness, and the like, 

 where the particular character cannot possibly 

 be separated and defined with any approach to 

 exactness, and where the data collected at large 

 rest on the observations of several persons widely 

 differing in temper and qualifications. That is 

 surely but a poor basis on which to rest exact 

 measures of quantity. 



When in regard of a complex and difficult en- 

 quiry involving unknown conditions and factors 

 statistics are collected at large by different 

 observers with all sorts of different tempers, 

 characters and qualifications, mutual intelligence 

 and collaboration by right method are practi- 

 cally impossible. Every observation is a product 

 of the fact to be observed and of the observer 

 who makes it, and of the two factors the personal 

 factor is vastly the more important. A single 

 observation by one who knows how to look 



