132 SCIENCE PROGRESS 



first sight it seems to support the idea that the popular will is 

 as vacillating and unstable as the strongest critics of democracy 

 have maintained. But the very regularity of these periods 

 suggests that something more is at work than the arbitrary 

 whims of a mere inconstant mob ; we seem to be in presence 

 of a periodic force whose working, although concealed, is con- 

 tinuous and recurrent, and which should therefore be sufficiently 

 rational to yield its underlying law to analysis. 



A close examination of consecutive election figures soon 

 furnishes a working hypothesis. In 1885 and 1892 the Liberals 

 had a majority ; in 1895 and 1900 the Conservatives ; in 1906 

 the Liberals again secured a majority. 



The actual Conservative vote in 1885 was 1,935,216 



n n f* „ 1895 „ 1,780,753 



b „ „ „ 1900 „ 1,676,020 



„ „ „ „ 1906 „ 2,463,608 



We have therefore the apparent anomaly that the Conserva- 

 tives polled more votes in 1885, when they lost the election, 

 than in 1895 or 1900, both of which elections they won ; while 

 their largest poll was in 1906, when they lost by a very large 

 majority. 



The Liberal figures explain this anomaly. 



The actual Liberal vote in 1885 was 2,156,952 



„ „ n „ 1895 „ 1,657,856 



>y „ . „ „ 1900 ^ „ 1,520,285 



„ „ Liberal-Labour vote in 1906 was 3,111,929 



The total electorate in 1885 was 5,707,531 ; of whom 

 4,092,168 voted. At the two following elections the natural 

 increase of population involved a small increase in the number 

 of possible voters — there was no change in the franchise — but 

 the total number of votes cast was smaller in 1895 than in 1885, 

 and smaller in 1900 than in 1895. In 1906 the number of 

 possible voters had again increased, but this time there was 

 a very large increase in the actual number of votes polled ; 

 with the curious result that the Conservatives suffered their 

 greatest defeat at the election at which they polled more votes 

 than they had ever done before.^ 



^ Consideration of the 191 8 election is omitted, as the conditions were 

 abnormal. The electorate was largely increased, the number of potential 

 voters being 21,392,322 ; but 106 constituencies were uncontested, which 

 deprived nearly 7,000,000 persons of a vote, on the average calculation 

 that one member represents 65,000 persons. Many voters were away at the 

 war, many more were far from their homes ; thousands who were on the 

 register were already dead, the election having followed immediately the 

 outbreak of influenza. In these circumstances, the fact that 10,781,025 

 electors voted must be considered extremely high. 



