216 Methods of Election. 



same state of the poll as in the first scrutiny of the proposed 

 method. Thus the second way of applying the method will 

 give the same final result as the proposed method. This 

 second way is very convenient, for if there be an absolute 

 majority for or against any candidate, it is made obvious at 

 the first or second count, and the election is settled with as 

 little counting as possible. The two counts are conducted on 

 well known plans, and if the circumstances are such that 

 either of these necessarily gives a correct result, that result 

 is adopted. But if it is not obvious that a correct result can 

 be arrived at, then we take the mean, or what comes to the 

 same thing, the aggregate of the two counts. This might 

 appear to be a rule of thumb, and on that account may per- 

 haps commend itself to some persons. This is not the case, 

 however ; and it is remarkable that that which might 

 suggest itself as a suitable compromise in the matter should 

 turn out to be a rigorously exact method of getting at the 

 result in all cases. The view of the proposed method which 

 has just been given shows exactly what modifications require 

 to be made in Condorcet's practical method in order to make it 

 accurate. 



Laplace's Objection. 



It may be said that the proposed method is open to the 

 objection raised by Laplace to the method of Borda. To 

 this I think it a sufficient answer to say, that if we have a 

 method which will truly interpret the wishes of the electors, 

 as expressed by their voting papers, we need not trouble 

 ourselves whether they vote honestly or not ; that is their 

 own concern. If we provide a method which will bring out 

 a correct result for honest electors we need not try to go 

 further, and endeavour to construct a method which will 

 force dishonest electors to vote honestly. Nevertheless, it 

 may be pointed out that Laplace's objection is not of so 

 much force in this ca,se as in the case of Borda's method. 

 For if an -elector vote otherwise than according to his real 

 views it will be at the risk of having his vote at the final 

 scrutiny counted against the candidate whom he considers 

 most fit for the office to be filled. This risk would be suffi- 

 cient to deter most electors from voting otherwise than 

 according to their real opinions. If. in spite of this risk, an 

 elector persists in voting otherwise than according to his 

 real views we must take him at his word. To illustrate this 

 objection, let us suppose that B and C are two formidable 



