this, he claims that it is quite possible to give fair representation to the main parties and to small sections at the same time.  In illustrating the system he avoids the issue as to the character of these sections by giving them a “scientific” nomenclature, such as Colour, Place, Pursuits, Qualities, &c.  These abstractions are very misleading, as attention is diverted from the fact that they refer to voluntary groups of men united for some political purpose.  The real question is, on what basis are these groups likely to be formed?  When the element of human nature is taken into account it must be apparent that they will be formed for the propaganda of some sectional interest; some on a religious basis, others on a class basis, &c.  Now, if we were to ask each candidate to declare his religion, we could easily take religions as the basis of representation and allow proportional representation to each religion; and similarly with classes, races, and so on.  But we could only take one basis at a time, and the important deduction is that if we were to take religions as the basis of representation, the people would be induced to vote according to religion; if we were to take classes, according to class, and so on.  Now, no one but the fanatic or the demagogue will claim that the majority is entitled to rule where religions only or classes only are represented.  The questions then arise—­What is the correct basis of representation?  How should the people be induced to vote?  And the answer is clearly that the people should be induced to vote on questions of general public policy, on the leading questions of the day which decide the party lines, and that, therefore, the policies of the two main parties should form the primary basis of proportional representation.  But the Hare system, by taking individual candidates as the basis of representation, induces the elector to vote on any basis or on sectional lines.  It promotes dissension instead of repressing it, and instead of encouraging all sections to express their opinion as to what is best for the general well-being, it encourages them to express their opinion as to what they imagine to be best for themselves.  Public opinion expressed on these lines would be worse than useless.  But Professor Nanson thinks that the electors would still have regard for the main parties, even though they grouped themselves into small sections.  He declares that “any party amounting to anything like a quota would not only have two candidates of its own—­one Liberal and one Conservative—­but would also be wooed by candidates of both leading parties.”  We may well question whether factions would trouble themselves about the main parties; but, granting the assumption, the small parties might just as well be single electorates as far as the main parties are concerned.  The Liberal candidates might be successful in all of them, and the Conservatives be unrepresented.  The peculiar feature is that the defeated Conservatives are expected to transfer their votes to the Liberals to make up the quotas for the small parties!