A modern

to have greater stability than ordinary laws. civilised state-we may now assume-will be normally a constitutional state, in which all the organs of government carry on their work under certain fundamental laws, which at once assign and limit their powers. We may assume that such laws should be somehow legally alterable: it is prima facie reasonable that it should be beyond the power of the ordinary legislature to change them; and it is obviously desirable that so long as they exist they should have the support of popular acceptance. If so, there seems to be no way so effective to secure the desired stability and popularity at once, as to place them formally under the guardianship of the people at large. This may, no doubt, be done indirectly, according to the method adopted in several European constitutions, by requiring any change in the constitution to be approved by two successive legislatures, so that a general election may always intervene between the proposal of the change and its final adoption. But, as I hold it undesirable that legislators should be elected solely on account. of their opinions on one particular question, I think it better that, where the assent of the people at large is required for the validity of any legislative change, this assent should be sought and given in a direct and simple form.

I pass, therefore, to consider (1) whether it is really expedient to give special stability to a certain portion of the law of the state; and (2) if so, on what principles the portion thus rendered exceptionally stable should be determined.

§ 4. I may begin by observing that the terms "constitution," "constitutional," are ambiguous; they may either signify "rules relating to the structure of government and the distribution of powers among its parts," or, "rules not alterable by the ordinary process of legislation." To avoid the confusion that the double meaning might cause, I shall here use the term "structure" for "constitution" in the former sense. It is obvious that a constitution," in the latter sense, may include rules that do not relate to the structure of government, but merely restrain or prescribe its action. On the

other hand, while every state must have some governmental structure since there must be some rules, expressed or implied, determining the appointment and powers of the organs of government-still, the process of changing such structural rules may be simply that of ordinary legislation, as it is in England. In this case we may say that the structure of government is Flexible: while, if it cannot be altered by the ordinary process of legislation, we may call it more or less Rigid, according to the difficulty of the process of change.

The grave and I think-decisive disadvantage of flexibility lies in the ease with which fundamental changes may be made; so that valuable rules and institutions may be abolished in a transient gust of unpopularity, and thus lose irreparably the stability given by antiquity and unbroken custom. The corresponding drawbacks of rigidity are partly that clearly expedient changes are prevented, partly that, by making change more difficult, the danger of violent revolution is increased. These drawbacks perhaps render it dangerous to require for a constitutional change a preponderance of votes so great that a majority, large enough to be plausibly described as "the nation," is still constitutionally unable to change its fundamental laws. But this objection could hardly be serious if the approval of no more than an absolute majority of the electorate were required for the validity of constitutional changes, since a restraining rule can hardly be felt as an intolerable burden by citizens who will not take the trouble to walk to the polling-places in order to get rid of it; at the same time this requirement is likely to be an important barrier against hasty changes. Perhaps a constitution protected by this barrier-and by the requirement of a similar majority in at least one of the chambers of the central legislature,1 before any change could be pro

1 In a two-chambered legislature, in which the members of both chambers are appointed by periodical election, an absolute majority in each of the two chambers might reasonably be required to concur in bringing forward any change in the constitution. But to allow a non-elective senate permanently to obstruct constitutional change, would obviously not be consistent with the popular basis here advocated for the constitution. The refusal of

posed-would have as high a degree of stability as it is desirable to aim at in a unitary state.1

There are, however, certain further disadvantages attaching to a rigid constitution which require careful consideration. Firstly, when changes are proposed, the attention of statesmen and of the public, which should be concentrated on the difficult task of weighing considerations of expediency for and against the change, is liable to be inconveniently distracted by the question of its legality; secondly, it is hard to find a satisfactory authority for deciding this latter question. The former disadvantage is to some extent inevitable; but in order to minimise it, minute and complicated constitutional restraints should be avoided as far as possible. The difficulty of finding an unexceptionable organ for deciding disputed questions of constitutional. interpretation may be met in various ways; none of which, however, is free from objection.2

Prima facie, this interpretative function belongs to the judiciary; at the same time, an arrangement by which the judiciary has to sit in judgment on the legislature involves certain difficulties and disadvantages. In the first place, in the ordinary administration of justice, the function of the judge comes into play when a breach of law re

consent by such a senate should therefore only have the effect of delaying for a time a change supported by an absolute majority of the house of representatives; and any power of veto vested in the supreme executive should be similarly limited.

1 For the reason why a federal constitution is naturally more stable, see chap. xxvi. § 4.

of

2 It is to be observed that this difficulty may conceivably arise even when the structure of government is flexible; since any structural rules, limiting the powers of governmental organs, may lead to disputes as to the power one or other organ, even though such rules are as alterable as ordinary laws; and it may not be possible to settle such disputes by new legislation, if ordinary legislation requires the agreement of several differently appointed bodies or individuals. Especially where the Executive is not dismissible by Parliament, and has a veto on new legislation, it is not unlikely to disagree with the legislature as to the exact limits of its powers under the existing law. In such a case the question must arise who is to settle the point at issue; and it will be a disadvantage if the constitution--whether fixed by precedent or by statute does not clearly assign, either to the ordinary judiciary or to some other body, the function of decisively interpreting its rules.

quiring reparation or punishment is alleged to have been committed; but if an act is passed by the legislature in excess of its constitutional powers, it is hard that an ordinary citizen should suffer either for obeying or for disobeying it. To prevent this, an authoritative decision on the validity of the act might be obtained from a judicial. tribunal before it was finally passed; but this arrangement is open to the objection that such a decision is less likely to be right than one arrived at in the ordinary way of litigation, after the court has heard the arguments of professional advocates on both sides. Perhaps it would be, on the whole, best that a short interval should be allowed in which objections might be taken to the constitutionality of any new law passed by the legislature, and that if they were taken the law should not come into operation until the points raised had been argued before and decided by a tribunal. Where no such objections were taken the decision of some high judicial organ without litigation-perhaps the Law Council before suggested-should be final.

But the plan of making the ordinary judges interpreters of the constitution is open to more fundamental objections. Since it confers on the judges a final and supreme power of practically determining the law, wherever the constitution is ambiguous and a majority sufficient to change it unattainable, it introduces a danger that the judges may be drawn into party conflict, and the confidence in their impartiality thereby be impaired. And from the same cause there arises a further danger that the legislature or the executive may be tempted to misuse its control over the appointment and dismissal of judges, in order to obtain a tribunal subservient to its wishes; while yet the withdrawal of all control of this kind would leave the judges in too independent a position.

On the other hand, if, to avoid these difficulties, we leave the interpretation of constitutional rules to the ordinary supreme legislature, we can hardly expect an impartial

1 This objection obviously applies still more strongly to the central legislature in a federal state.

decision in any dispute as to the limits of constitutional restraints imposed on legislation. If, therefore, in the case of any legislation that excites strong opposition, the charge of transgressing constitutional limits is urged against the parliamentary majority by their opponents, it is likely to be a permanent source of resentment, and seriously to aggravate any discontent that the disputed legislation may cause on other than constitutional grounds.

(I may notice a simpler method of avoiding the evils of a conflict among the organs of government,-viz. by depriving all constitutional rules of strictly legal force as against the supreme executive, and making it the duty of the judges to regard its commands as valid in the last resort: since this is an interpretation of "Constitutional Monarchy still surviving in Germany. It might doubtless sometimes enable the monarch to meet a dangerous crisis successfully, by a salutary extension of executive power which would otherwise be illegal and disorderly. But this advantage appears to me too dearly purchased by a sacrifice of the normal relation between the legislature and the executive.)

Another possible method would be to refer to the citizens at large any question of legislative change of which the legality is disputed by a sufficiently large minority in the legislature. This plan, however, is open to the objection that the citizens at large are obviously not qualified to decide disputed points of interpretation: and it is liable to make the constitutional rules practically more elastic than they were designed to be, if the assent of a mere majority of those voting is sufficient, and more stringent than they were designed to be, if it is not sufficient.

On the whole, I am inclined to prefer the judicial method, with the modifications above suggested, in a unitary no less than in a federal state; though the disadvantages of it appear to me to constitute a strong reason for reducing to a minimum the restraints on legislation which it is thus left to the judiciary to interpret and apply.

In any case, changes in the constitution should be