MKTliODS FUR COL'NTlNt; IN ELECTIONS. 66y 



Where, in place of the Hmited expression of the voters' 

 preference that is possible when each voter marks one choice 

 only, we can get the opinion of every voter as to the order of 

 his preferences for the members he most prefers by means of 

 the use of the transferable vote, Droo]^'s (juota fails to 

 indicate the members most preferred by the voters. This can 

 only be done when we make use of Hare's quota, by which 

 alone we can, to give ]Mr. Humphreys' own words, " give effect 

 to the first and higher preferences before making use of lower 

 preferences " to ithe largest j^ossible extent. We can do this, 

 because Hare's (jUota is the largest |x)ssible quota that can be 

 used. In this way only, by the use of Hare's quota, can Mr. 

 Pim's fifth rule, on which a correct system of counting must be 

 founded, be carried out. 



The (transferable vote gives us grades of preference — a new 

 feature. First-choice votes express a higher grade of pre- 

 ference than second-choice votes, second-choice votes a higher 

 grade of preference than third-choice votes, and so on. To 

 make use of the grades of i)reference thus expressed b\ the 

 voters at the poll, we must make use, to the utmost possible 

 extent, of all the tirst-choice votes firstly before second-choices 

 are considered, and so on successively, with each grade of choice. 

 Under these circumstances. Droop's quota no longer elects the 

 members mos: ]:)ref erred by the voters. It is therefore inadmiss- 

 able : it is inconsistent with the use of the transferable vote — 

 it fails to show the members mos:: preferred — when by tlie 

 transferable vote the voters express fully their preferences. 



The highest number of votes a member can obtain is Hare's 

 quota — the highest number all // members can each obtain. In 

 this paper I have pointed out that in all rules of counting, where 

 Droop's quota is .used to transfer the votes, we can only say 

 that certain members are elected by the Continental plan of the 

 absolute majority of one vote, but that we have no certainty 

 that these are the members most preferred by the voters. I have 

 shown that this was the very system Andrae, Hare, and John 

 Stuart Mill wished to supersede by the use of the transferable 

 vote, which enables every A'oter to help in the selection of the 

 member he most prefers. I have demonstrated that the use of 

 Droop's quota is utterly inconsistent with this object, and affords 

 no security that the wishes of the voters marked on tlieir ballot- 

 papers are carried out, and is inconsistent witli securing l'r(j])or- 

 tional Representation. 



I have shown that by the use of Hare's (|Uota, (Gregory's 

 .system of surplus distribution, and Mr. Harold Pim's method 

 of the simultaneous distribution grade by grade of all available 

 choices as marked by the voters, their wishes can be carried out 

 exactly as they marked them with impartiality, certainty, 

 accuracy, and despatch. 



I have drawn up a set of suggested rules which will, 1 

 believe, secure these results ; and 1 give here the working out 

 of two elections under both systems, and a table of choices in 



