VI 



EOYAL SOCIETY OF CANADA. 



to writing, so as to be jji-inted in the ' Ti'ansactions ' with Dr. MacGregor's communication of May 

 !i3rd.^ 



With following result : 



Total value of votes. 



18 votes cast for D=91 



18 " " " C=89 



18 " " " B=85 



17 " " " E=75 



By Dr. MacGregor : 



D elected ]st vacancy. 



C " 2nd 



B " 3rd 



E " 4th " 



18 

 11 



" " F=64 

 " " A=35 



Highest possible value of votes, 18 x 6=108. 

 Two-thirds highest possible value, 72.] 



' It is thought best to place here both Dr. MacGregor's letter and Dr. Fleming's proposed amendment, as 

 afterwards laid before the third section. In revising the proofs, Professor MacGregor makes the following commen- 

 tary on Dr. Fleming's suggestion, which the editor thinks it advisable to add : 



" Dr. Fleming's proposed method of treating the voting papers may be expected occasionally to elect the wrong 

 man, because it makes the unwarranted assumption that the members in the voting papers give a quantitative 

 estimate of the appreciation of the voter for the candidate. That it might do so is obvious from the following 

 tabulated voting papers in an election to one vacancy with three candidates. As in Dr. Fleming's table, the larger 

 numbers indicate greater preference. 



" Thus according to Dr. Fleming's method, B would be elected ; and yet as between A and B ten members 

 vote for A and only eight for B. Hence in this case Dr. Fleming's method would elect the wrong man." Ed. Trans. 



