civil actions the direct aim of the judicial intervention of government is to decide disputed points of private right, and give adequate remedies for private wrongs. If, therefore, any individual can obtain the satisfaction he requires otherwise than by the intervention of government, there is no need that the latter should take place: indeed, it is obviously better that the expense, trouble, and probable increase of ill feeling which a formal public trial is likely to entail, should be spared to both parties. Hence, it is desirable that government shall encourage suitors to resort to arbitration; and should enforce the decisions of an arbitrator, if he has been duly appointed and there is no ground for impeaching his bona fides, in case either party refuse to abide by the result of the arbitration. And when the intervention of the judge is necessary, his proper function is obviously to hear both sides and decide between them: no one has ever doubted that the manner of procedure in a civil action should be litigious and not inquisitorial, and that the judge should not initiate a civil action, or carry it on against the will of both parties, or decide any issues of right not raised by the parties.

In all these respects criminal or penal justice presents an important contrast. Punishments, as distinct from damages, are inflicted for the prevention of offences, and therefore primarily in the interest of the community: and where they are inflicted (as is the commonest case) for mischief that falls primarily on some private individual, it is important— at least unless the offence is slight-that the matter should not be settled by private compensation to the individual. This result may be partly attained by making it illegal to "compound a felony": but it seems inequitable that the burden of bringing the offenders to justice should be legally thrown on the person who has already suffered the mischief of the offence, or his nearest relatives: nor is it desirable to rely entirely on revenge-which is an objectionable motive-or safe to rely on public spirit for the performance of this important duty: while, again, it is not in most cases expedient that individuals should be tempted

to take up this invidious task as a trade, for the sake of a pecuniary reward.1 Under these circumstances, the necessity for a public prosecutor or investigator of crime appears incontrovertible.

It is not so clear whether the prosecution or investigation of crime should be regarded as belonging to the executive or the judicial organ. As the business requires energy, discretion, and skill more markedly than judicial impartiality, it seems expedient that the officials engaged in it should be under conditions-in respect of dismissibility—similar to those of executive officials generally : also the work of discovering the perpetrators of crime and proving their guilt is naturally connected with the essentially executive business of forcibly preventing and repressing crime. Moreover, on the assumption that a criminal trial is to take a litigious form, and to be conducted as a dispute between the parties, any close official connection between prosecutors and judges is prima facie objectionable as tending to throw suspicion on the perfect impartiality of the latter. On the other hand, if the prosecution of crime be made a part of executive business, there is some danger of its being performed with undue partiality towards members of the executive. To obviate this, it should be open to private persons to prosecute-judicial permission being obtainedif the public officials decline to do so, or even along with them and the power-which seems necessary-of prohibiting prosecutions as vexatious should be vested in a judicial and not in an executive organ.

It may, however, be doubted whether the litigious form of procedure, which is proper to civil suits, ought to be adopted in a criminal trial. Certainly a public prosecutor, in presenting his case, ought not to show, and would be generally condemned for showing, the partiality which is

1 There are only two cases in which the expedient of repressing mischievous acts by "penalties" recoverable by "common informers" appears to be admissible: (a) in the case of offences at once difficult to discover and important to repress, and yet not directly mischievous to private persons; and (b) as a constitutional security in the case of offences committed by public officials.

tolerated as natural and inevitable in a private litigant. Hence the litigious form of criminal procedure is open to the objection that the litigants cannot really be on equal

e.g. an advocate for the prosecutor would be severely blamed for concealing evidence telling in favour of the accused, but the accused or his advocate would not be similarly blamed for concealing evidence against him. Still, to secure justice, an unbiassed judge who has no responsibility for the prosecution seems indispensable, not only in the final trial, but also to decide whether there are adequate grounds for imprisoning the accused before trial: so far, therefore, the litigious form of procedure seems inevitable, and it is difficult to combine it satisfactorily with a procedure to any extent inquisitorial. At the same time, I admit that where an accused person has in any case to be kept in prison for some time before his trial, there is an obviously convenient opportunity for a private inquisitorial investigation by a judge: and that such an investigation is often likely to be a more effective instrument for finding out the truth than a merely litigious procedure. I think also that the English dislike of inquisitorial examination of accused persons is partly due to a confusion between the sound principle that an innocent person should be allowed every means of proving his innocence, and the unsound principle that a guilty person should be allowed to resist or evade attempts to prove him guilty.

I may note briefly the characteristics of criminal procedure at least in grave cases-which seem to me to follow properly from the sound principle just mentioned. The accused person should be allowed complete information as to the charges against him, full time for preparing his defence, and the advice of experts. He should be allowed to hear all the evidence given against him and to crossexamine the witnesses personally or by counsel to make the latter privilege completely effective, it seems desirable, generally speaking, that only oral evidence should be admitted; unless the accused himself wishes to put forward 1 i.c. in the final trial.

written testimony on his behalf. He should be allowed to summon witnesses on his side who should be bound to attend, and, if he wishes, to give evidence personally, subject to cross-examination. As a final guarantee against official oppression, his trial should be in a place to which the public are normally admitted; though exceptions to this rule of publicity are needful in special cases, in the interest of morality; and the judge should always have such powers of exclusion as may be necessary to maintain order, and repress demonstrations of popular feeling dangerous to the independence of the tribunal.

§ 7. A right of appeal against the decision of any court of first instance should generally be allowed, in order to correct mistakes and preserve uniformity in the judicial interpretation of law. It is a more difficult question how far an appeal should be allowed from a decision on a question of fact, which does not form a precedent. In particular, such an appeal does not harmonise well with the jury system; except in cases where there is fresh evidence, of which the absence in the first trial was not due to the negligence of the party whom it favours. For it hardly seems consistent with the principle of the jury system to appeal from a jury to a court of professional judges; while, if the appeal be made to a second jury, it is hard to see why this should be generally expected to judge more correctly than the first. It would seem therefore, that, in civil cases at least, a decision of a jury should not be set aside by a court of appeal because it is against the weight of evidence, except when an unmistakable and scandalous miscarriage of justice has occurred. case, the independence of the court that has the power of overruling the jury should be guarded with especial care: and, if the principle of the jury system is thoroughly maintained, the result of the overruling must be not a decision on the question at issue, but a new trial by another jury.

In any

The question of appeal-on other than purely legal issues in criminal trials is beset with peculiar difficulties, especially where the punishment is severe, and such as either could not, or practically would not, be increased on appeal:

since, in such cases, one would expect criminals always to appeal, unless prevented by cost: while if cost alone prevented the appeal, the inequality between rich and poor would be flagrant. A widely accepted way of meeting this difficulty is not to grant a formal right of appeal to convicted criminals, but to give to a high executive official the power of remitting or mitigating punishment at his discretion; a power which can also be used in cases where the infliction of the full legal punishment would be for special reasons impolitic, as well as in cases of judicial error. The objections to this are (1) that an executive individual or council is not likely to be particularly well qualified for difficult judicial functions, and (2) that the executive thus acquires a dangerous power of weakening the restraints of law. On these grounds it seems desirable that if this power be vested in the executive its exercise should be subject to the approval either of carefully selected judicial advisers or of parliament: judicial approval being required when the "pardon" or commutation of punishment is granted on strictly judicial grounds, parliamentary approval when it is granted on extrajudicial grounds. The latter kind of intervention should, however, be extremely rare, lest the deterrent effect of legal penalties be dangerously weakened; and the initiative should in no case be taken by parliament, so that the responsibility for the exceptional procedure may be concentrated on the executive.

Finally, the consideration of expense renders it expedient to deal with cases of minor importance in a more summary manner, with a procedure in which simplification for the sake of economy overrides to some extent precaution against error. The same consideration may reasonably prevent the separation of executive and judicial functions from being carried out to the extent that would otherwise be desirable.

§ 8. So far I have taken note of the differences of grade in the judiciary required on grounds of economy-tribunals of the lowest grade being confined to cases of minor importance; and also of the differences in organisation corresponding to the fundamental distinction between civil and criminal