386 Report S.A.A. Advancement of Science. 



tlie surplus was 2824, the total number of papers on which next pi'e- 

 ferences were marked was 4876, and the numbei- of next preferences 

 recorded for Sir A. Acland-Hood was 77. Sir A. Acland-Hood's sliare 

 of the surplus was therefore f H-g^ x 77 {i.e. 44*6). In calculating the 

 number of votes to be transferred it is desirable to ignore fractions. 

 When this is done theie will be a slight difference between the total 

 of the votes transferred and the surplus. In this e.xperiment the 

 numbers were adjusted so as to make the total votes tmnsferred 

 correspond with the surplus. The correct proportions were maintained 

 as accurately as possible. 



The next operation was the transfer of Keir Hardie's surplus votes, 

 but without going through all the process I ma^^ say that after adding 

 all the transferred votes of Sir Hem-}' Campbell-Bannerman and Keir 

 Hardie, none of the other ten were made up to the necessary quota. 



The next operation was the distribution of the votes of the 

 candidate lowest on the poll beginning from the bottom and working 

 upwards. The election of these candidates was hopeless, but as their 

 supportei'S had indicated to whom in that event they wished to transfer 

 their support their votes were not vvaste<l. This process involved no 

 calculations, as all the votes of each candidate had to be transferred, 

 and those papers which showed no further pi-eferences were naturally 

 set aside as exhausteil. 



The final result was : — 



elected. 



12,398 



The next step should have been the allocation of Mr. Chamberlain's 

 surplus votes, but even although Mr. Snowden had got the lot he would 

 still have been behind Mr. Churchill, who accordingly was declared 

 elected. 



It would be most interesting to analyse the results of this election, 

 but my object is served by indicating the siniplieity of the whole pro- 

 cess. If the votes recorded for the candidates in the first place be 

 totalled in accordance with their party views it will !)(> found that tlie 

 election results weie not quite in acc<»id with those figures, which are 

 as follows : — 



5810 Liberals. 



2572 Lalxnn-. 



3016 Unionists 



However, no fairer distribution could have been made, because the 

 Liberals had two "quotas" with a remainder of 1676, the Labour one 



