PEOCEEDmrtS FOE 1894. LXXV 



totals are 64, 62, 69, 67 and 38, respectively. Thus Dr. Fleming's method would elect C and D 

 whereas obviously the majoritj- of the voters prefer A and B to all the other candidates. 



With twenty electors voting, a similar complete miscarriage may occur if there are three 

 vacancies, jjrovided theie be six candidates. I need not give an example, for the assertion is easily 

 verified. 



The committee, in asserting that the candidates elected by Dr. F.'s method in such cases are the 

 best all-round candidates, give no adequate grounds for their assertion. They obviously regard the 

 numbers by which a voter indicates the order of his preference of the candidates as somehow giving a 

 measure of his opinion of their merit. That they do not give any such measure, however, even of the 

 roughest kind, is clear from the fact that though one voter may regard the candidates as all very good, 

 and another as all very poor, he yet uses the same numbers in indicating his order of preference. In 

 the former case No. 1 means : very good, but not so good as the others ; in the latter case it means : very 

 poor and pooier than the others, but still fit to be a Fellow. Hence, to conclude from the fact that 

 though a candidate gets few high numbers awarded him ho also gets few low ones, that if his total is 

 greatest he is the best all-round candidate, seems to me to be a complete noii tsequitur. 



The possibilitj', according to Dr. F.'s method, that a candidate who is considered best by a major- 

 ity of voters may lose his election if a sufficient minority place him low down on their lists, gives a 

 minority a very dangerous power. It seems imjDrobable that with four or five candidates in the field 

 one so eminent as to be considered best by a majority should, on grounds of fitness for a fellowshij) 

 alone, be ranked sufficiently low by a sufficiently large minority to lose his election. But it may readily 

 haj)pen that he is placed sufficiently low to give this result on other grounds, say because the minor- 

 ity are not familiar with his work and have not taken the trouble to make proper inquiry, or because 

 he is connected with an institution, or is resident in a province already in their opinion sufficiently 

 represented in the Society. Thus Dr. F.'s method enables a minority of voters to reject, on what I 

 may call unworthy grounds, a candidate whom the majority wish to elect. This seems to me a very 

 dangerous power to confer upon a minority. It may be right enough within certain limits to elect a 

 candidate of less eminence in order to secure a more equable representation of institutions, provinces, 

 etc., but this should be done by the will of the majority of voters, not of a minority. 



Dr. F.'s method makes no provision for the case in which a voter may regard two or more 

 candidates as equal in their claims, a case which may frequently arise. If a voter considers A best, B 

 and C equal and second best and D the least desirable of four candidates, is he to attach to their names 

 the numbers 4, 3, 3, 2, or 4, 2, 2, 1, or 4, 3, 3, 1, or 3, 2, 2, 1, or is he to toss up as between B and C 

 for the numbers 3 and 2 ? The result of an election on Dr. F.'s method may obviously turn upon the 

 mode of notation in such cases. 



The committee report that Dr. F.'s method is much the simpler of the two. It should be noted, 

 however, that it is the simpler onlj' so far as the work of the scrutineers is concerned. Both methods 

 use the same voting paper, except for the defect in Dr. F.'s method mentioned in the last paragraph. 

 The simplicity which is desirable in an election is simplicity in the voting paper. Scrutineers can 

 easily be found to do the work of collating the voting papers; but it is often difficult to get even 

 intelligent electors to fill up properly a complicated voting paper. As, excej^t for the difficulty 

 mentioned in the last paraijraph, the voting papers are the same according to the two methods, Dj-. 

 F.'s has no advantage in this respect. With my method the work of the scrutineers is less simple 

 than with Dr. F.'s, but the number of voters being small, the difference is hardly appreciable. After 

 many trials, I find that to determine who are elected in any election such as we hold, my method 

 requires say fifteen minutes more than Dr. F.'s, while the process is within the capability of any 

 school boy. The difference of simplicity is therefore negligible. 



I dare say it was the form of my proposed amendment of rule 6 that made my method appear so 

 complex to the committee. But if Dr. F. will prepare an amendment embodying his method and 

 providing for all cases that may arise, I think it will be found to assume an equally complex and 

 repellent form. 



