HUMAN FACULTY 283 



Then, through the courtesy of Mr. Muir, the 

 Chief Examiner at the Education Office, I was 

 allowed to examine a large number of results from 

 the Civil Service Examinations, and found that the 

 average value of the first prize should be ^74. 

 Taking groups of 50 cases, each group gave that 

 value pretty closely, no one differing as much as ^4 

 from it. 



The subject has since been generalised and dis- 

 cussed in Biometrika with far more mathematical skill 

 than I possess, by both Professor Karl Pearson and 

 Mr. W. F. Sheppard (a former Senior Wrangler), 

 with practically the same result, so that if only two 

 prizes are to be given, whatever be the character of 

 the competition, and whatever the number of candi- 

 dates, the first prize should in round numbers be 

 three times the value of the second. 



Professor Max Muller had, in a work dated 1886 

 or 1887, laid an exaggerated stress, as I considered, on 

 language as a means of thought, upon which I wrote 

 some remarks in Nature [97], entitled "Thought 

 without Words," which led to a short newspaper 

 controversy, June 2, between us two. My point 

 was that I myself thought hardest when making no 

 mental use of words. Professor Max Mliller's 

 definitions of what he considered "words" seemed 

 to me to vary, and therefore to be elusive, so I did 

 not and will not pursue the matter farther. 



It led, however, to the idea of an experiment that 

 seemed worth making, which I described [127] as 

 "Arithmetic by Smell." When we propose to add, 

 and hear the spoken words "two" and "three," we 



