The question of justice, however, is not a whit the less essential because it presupposes this social characteristic instead of supplying the primary axioms from which it is to be deduced. It is undoubtedly of the highest importance that every difference in our method of treating different classes should have its sufficient reason, to be assigned as clearly as possible. The preservation of the peace is essential; but that does not settle the methods by which it is to be preserved. On what ground, then, are we to deal with the problem of justice as regards different classes of crime? If the calculation of pain and pleasure, as already stated, seems to be unsatisfactory, what is the right principle of proportioning punishment to offence? I have noticed one argument which Bentham applied, and, as I think, with very good reason. To punish crimes equally, he said, is virtually to put a premium upon the worst. The "in for a penny in for a pound" maxim becomes at once applicable. Moreover, as every one now admits, the old brutal system is condemned by experience. To punish a great number of offences with death led to a mixture of excessive brutality with excessive uncertainty. The cruel punishment of some criminals was balanced by the complete escape of others. But this practical failure clearly resulted, in great measure, from an obscure sense of justice. It was grossly unjust, it seemed, to hang a man for stealing a loaf, when you could only hang another for the brutal murder of his wife. The penalty in the first case, was, it was felt, altogether out of proportion to the offence. This instinctive sentiment was, as I think we all feel, substantially right. In any case, it would have to be taken into account by the legislator, for the obvious reason that punishments which outrun public opinion, tend to make martyrs of criminals. They are either not inflicted, or they set the sympathy of the people on the side of the offender. But to say this, is not to prove the sentiment to be just, only to take account of its existence. And the question, therefore, remains, how it is to be logically justified, for it may seem to imply the theory to which I have objected--the hypothesis of a sort of debtor and creditor account--of the old "eye for an eye" doctrine, which, as I have argued, involves a misconception of the true doctrine. My reply would be, in general terms, that the doctrine requires restatement, and, if properly stated, will not lose but acquire new forces.