ties is useful in keeping alive in the minds of the citizens at large the sense that the prevention of crime is a public duty which ought not to be entirely resigned into the hands of officials; since circumstances may at any time arise in which the aid of private citizens in performing this function is valuable and even indispensable. On the other hand, it is of great importance to the whole community that no part shall be allowed to harbour law-breakers; and, assuming that the rules of law are in the main determined by a central legislature, it seems almost indispensable that the coercive organisation for their enforcement should be under the ultimate control of the central executive; since, if it is placed in the hands of locally appointed organs of government, there is a serious danger that laws locally unpopular will not be effectively enforced. At the same time local legislation should on the same principle be enforced by local police; and the existence side by side of two police organisations separately directed is likely to be a troublesome complexity. It seems, therefore, on the whole best that the main part of the police force should be supported from national funds and in the service of the central government, but that, together with any additional police locally appointed and paid, it should be normally left under the control of local executives, who should for this purpose act as subordinates of the central executive; it being always in the power of the latter to resume the control of the force paid by it, in case of necessity.

In the case of the machinery for the punishment of crime, the considerations in favour of a central organisation for any area over which a common system of law is established seem more decisive. The impulses of indignation that occasionally prompt private persons to take the punishment of crime into their own hands, call, in a normal state of society, for repression rather than encouragement; while the importance of uniformity in order to avoid injustice is very manifest. The task of justly apportioning punishment to crime is in any case a profoundly difficult one; but a needless and preventible kind of arbitrary inequality is introduced

into the application of punishments, if-e.g.-a period of imprisonment, nominally the same, involves very different degrees of privation and discomfort in different localities.

§ 4. So far I have chiefly had in view the executive or administrative work of government, together with the financial business which this entails. But perhaps the most important question that belongs to the present chapter -and one of the most important questions in the whole discussion of the structure of government-relates to the extent to which Legislation should be allowed to be localised. We have seen before that some power of laying down general rules, to be obeyed by others besides the servants of government, cannot without inconvenience be denied to the central executive and in the same way the local executive work that we have been considering will naturally involve some exercise of legislative functions. Thus we may assume that local governments will have a limited1 power of making general regulations for the common use of streets, bridges, parks, and other public property, of which the use is necessarily confined in the main to the inhabitants of certain localities, and of which, therefore, as we have already seen, the expense may properly be localised. So again the sanitary intervention of government and protective measures against noxious plants and insects, and against destructive floods, will usually involve a certain amount of general coercive regulation. But the peculiar interest that a man's neighbours have in his right behaviour is obviously not restricted to his observance of such rules as these. Even as regards the fundamental rights of personal security, property, and contract, it is indefinitely more important to a Yorkshireman that they should be properly defined and protected in Yorkshire, than that they should be properly defined and protected in Kent. Nay, further, it is to be expected that differences of physical conditions and industrial development-if not of race and political history -will render the special needs of Yorkshiremen in respect of protection from mutual mischief, enforced co-operation for

1 If the power were unlimited it might be abused for the oppression of classes locally unpopular.

common benefit, or regulated use of natural resources, somewhat different from the special needs of Kentishmen; thus Yorkshire may require factory acts, but be indifferent to the regulation of hop-picking, or compulsory insurance against fruit disease, which may be prominent objects of concern in Kent. Hence localised legislation will tend to be more fully and closely adapted to these varying requirements than centralised legislation is likely to be. Even where the real requirements do not materially vary, it is not unlikely that there may be wide differences between the views prevalent in different localities as to the nature and limits of desirable legislative interference; and if this is the case, the gratification of the consequently divergent demands-even at the cost of some amount of bad legislation may avoid the grave evil of spreading a general aversion to law in any district when the wishes of the inhabitants are over-ruled in deference to the opinions prevailing in other districts. It is true that the total amount of coercion involved in the enforcement of law does not necessarily tend to be reduced by extending the legislative power of local governments, even if such extension results in considerable local variations in law; since the aggregate of local minorities opposed to the different local laws may be as large as the previous minority opposed to the law of the state. But so far as there are any causes tending to make different opinions prevail in different districts, we may assume that legislation will be more closely adapted to the wishes of the inhabitants of all the districts taken together, in proportion as it is localised.1

On the other hand, we may assume that, speaking generally, the average statesmanship of the aggregate of local legislators will be inferior to that of the members of a single central legislature; and further, that the danger of mischievous legislation in the interests of a predominant class will be on the whole greater in the bodies responsible to sections

1 Also, assuming that there is less aversion to migration within a country than to emigration from it, a quiet reduction of dissidence through the departure of the dissidents will be more probable in the case of local laws.

of the community. But the strongest reasons for limiting narrowly the legislative powers of local governments are drawn from the bad effects of diversity in the laws enforced in different parts of a coherent civilised community. In certain important cases, the protection from mischief which the law is designed to afford cannot be effectively given except by rules enforced throughout the whole country; thus, if Yorkshire refused to protect industrial inventions from imitation, any encouragement to inventors given by patent-laws in Lancashire would become practically worthless. But even in cases where uniformity is less indispensable, serious inconveniences must be expected to result from allowing the definitions of important legal rights to vary from district to district, within a country where migration from one district to another is unfettered and frequent. In the first place, such variations are likely to bring the doubtful equity and expediency of the varying laws prominently before the minds of ordinary citizens; so that it will become more difficult to maintain in due strength the habit of obedience to law and the sentiments condemnatory of illegal conduct. Further, the intellectual labour necessary to acquire adequate knowledge of the law for practical purposes will be seriously increased, at any rate for traders and other persons whose callings bring them frequently into legal relations with inhabitants of different districts: And finally, since the law courts of any one district will be continually called upon to recognise the validity of legal relations determined by the laws of other districts, the judicial administration of law must become more laborious, through the bewildering variety of legal rules which judges and legal practitioners are required to know and apply; and the same cause will have an unfavourable effect on the average quality of judicial decisions.

These disadvantages make it generally expedient to avoid any extensive devolution of legislative powers on local governments, at any rate in a community of which the parts are not strongly divided by marked differences of race or civilisation, or the habits and sentiments surviving from

1

previous political independence, and are not precluded by distance or physical obstacles from active mutual communication. It should, however, be noted that the disadvantages above mentioned are likely to be very different in degree in different departments of law. Thus, we have before seen that local variations are especially to be deprecated in the law regulating family relations, on account of the special importance in this department of harmony between legal rules and moral sentiments; and also in the law regulating commercial relations, on account of the natural and desirable tendency of such relations to relations to extend over the lines of local division in a modern civilised community. On the other hand, we may expect the disadvantages of mere variation-apart from any consideration of the goodness or badness of the different rules-to be decidedly less in the case of laws regulating the tenure of agricultural land: since the purchase or occupation of such land is an important act that is not likely to occur very frequently in the life of an ordinary member of the community; and the disputes arising out of it will rarely involve any conflict of laws.

One argument for allotting legislative functions to local organs deserves special notice from a student of politics, since it interests him-if I may so say-professionally; viz. that it would assist the progress of political science, through the greater opportunities that it would give for the trial of legislative experiments. And certainly in matters on which the opinions of experts appear to be evenly divided as to the expediency of a given legislative measure, valuable instruction might often be gained by allowing it to be partially introduced by way of experiment in a district favour

1 See chap. xiv. § 5: where I have quoted from an American writer (Professor Munro Smith) an illustration of the bad effects of local variations in family law. It is, however, to be noted that the inconveniences due to local variations in law have been much reduced in the United States by the fact that most of the States have inherited the English common law, and both Federal and State Courts, in interpreting and applying this law, have co-operated to keep it practically the same in the different States. See Professor Munro Smith's Article (II) on State Statute and Common Law, in the Political Science Quarterly for March 1888.