number of Liberal members rose from 8 to 40, and the number of
Conservative members fell from 52 to 20. The violence of these changes
was attributed to a similar change on the part of the electors, but it
was much more largely due to an electoral method that exaggerates any
changes in public opinion beyond all reason.

If, however, the results--not of two but of the eight General Elections,
1885-1910--are considered it will be seen that the current belief, that
the single-member system invariably yields a large majority, rests on a
very precarious foundation. The General Election of 1892, for example,
gave to the Liberals (inclusive of the Nationalists) a majority of 44
only. In England (which, excluding Wales and Monmouth, returns 461
members) the Conservatives in 1895 and 1900 had majorities of 233 and
213; in 1906 the Liberals had a majority of 207; but in the elections of
January and December 1910, the Conservatives had on each occasion a
majority of 17 only. If Wales and Monmouth are included, it will be
found that in the 1910 elections the Liberal majorities were 13 and 11
respectively. Single-member constituencies do not therefore guarantee
large majorities. It can with greater truth be said that they guarantee
wrong majorities, for, as the following table shows, there is no
constant relation between the size of the majority in votes and the size
of the majority in seats:--

General Election. Majority in Seats. Majority in Votes.

1885 Liberal 158 Liberal 564,391
1886 Conservative 104 Liberal 54,817
1892 Liberal 44 Liberal 190,974
1895 Conservative 150 Conservative 117,473
1900 Conservative 134 Conservative 157,417