their oars lashing the waves into white foam. But once

As we considered it necessary in the interests of social self- defence, in the case of criminal law, to combat the individualist excesses of the classical school, so in regard to penal procedure, whilst admitting the irrevocable guarantees of individual liberty, secured under the old system, we think it necessary to restore the equilibrium between individual and social rights, which has been disturbed by the many exaggerations of the classical theories, as we will now proceed to show by a few examples.

The presumption of innocence, and therewith the more general rule, ``in dubio pro reo,'' is certainly based on an actual truth, and is doubtless obligatory during the progress of the trial. Undetected criminals are fortunately a very small minority as compared with honest people; and we must consequently regard every man who is placed on his trial as innocent until the contrary has been proved.

But when proof to the contrary is evident, as, for instance, in the case of a flagrant crime, or of confession confirmed by other elements in the trial, it seems fit that the presumption should cease in view of absolute fact; and especially when we have to do with habitual criminals.

Even the criminals of this class whom I have questioned recognise a presumption of the opposite kind. ``They have convicted me,'' said an habitual thief, ``because they knew I might have done it, without any proof; and they were in the right. You will never be convicted, because you never stole; and if we happen to be innocent once in a way, that must be set against the other times when we are not discovered.'' And the ironical smile of several of these prisoners, condemned on circumstantial evidence, reminded me of a provision which was once proposed in the Italian penal code, under which a person surprised in the attempt to commit a crime, if it was not known what precise form his crime would have taken, was to be found guilty of a less serious offence. This might be good for an occasional criminal, or a criminal of passion, but would be absurd and dangerous for habitual criminals and old offenders.

The exaggerations of the presumption ``in dubio pro reo'' are due to a sort of mummification and degeneracy of the legal maxims, whereby propositions based upon observation and generalisation from existing facts continue in force and are mechanically applied after the facts have changed or ceased to exist.

What reason can there be for extending provisional freedom, pending an appeal, to one who has already been found guilty and liable to punishment for a crime or offence, under sentence of a court of first instance? To presume the innocence of every one during the first trial is reasonable; but to persist in a presumption which has been destroyed by facts, after a first condemnation, would be incomprehensible if it were not a manifestly exaggerated outcome of classical and individualist theories, which can only see a ``victim of authority'' in every accused person, and in every condemned person also.

Another point is that of acquittal in case of an equality of votes, especially where born and habitual criminals are concerned. I think it would be much more reasonable to restore the verdict of ``not proven,'' which the Romans admitted under the form of ``non liquet,'' as an alternative to ``absolvo'' and ``condemno,'' and which may be delivered by juries in Scotland. Every one who has been put on his trial is entitled to have his innocence declared, if it has been actually proved. But if the proofs remain incomplete, his only right is not to be condemned, since his culpability has not been proved. But it is not the duty of society to declare him absolutely innocent, when suspicious circumstances remain. In this case the only logical and just verdict is one of ``not proven.'' Such a verdict would obliterate the shadow of doubt which rests on persons who have been acquitted, by reason of the identical verdicts in cases of proved innocence and inadequacy of proof, and on the other hand it would avoid the tendency to compromise, under which judges and juries, in place of acquitting when the proof is insufficient, sometimes prefer to convict, but make the punishment lighter.

Another case of exaggeration in the presumption of innocence is afforded by the regulations as to contradictory or irregular verdicts, which may be corrected only when there has been a conviction; whilst if the error has led to the acquittal of an accused person, it cannot be put right. The influence of the individualist and classical school is here manifest, for, as M. Majno says, ``the justice of sentences rests as much on just condemnations as upon just acquittals.'' If the individual has a right to claim that he shall not be condemned through the mistake or ignorance of his judges, society also has the right to demand that those whose acquittal is equally the result of mistake or ignorance shall not be allowed to go free.