which will already have presented itself to one who has followed the argument thus far.

In dwelling upon the services which Constitutions may render, by fostering the centripetal forces, or by restraining the violence and softening the action of the centrifugal forces, we must not forget that no scheme of government can hope permanently to resist the action of either tendency if either develops much greater strength than it possessed when the Constitution was framed. If the centripetal forces grow, the Constitution whose provisions have recognized and given scope to the centrifugal will be practically, in some of those provisions, superseded. If the centrifugal grow, it may be overthrown. It is where the forces are nearly balanced, that the weight of the Constitution may turn the scale, and avert conflicts which would have rent the community, or caused a violent subjection of one part of it to the other. And in any case the Constitution ought, where dissimilative and disruptive forces are feared, to be so drawn as to enlist all available motives of interest, to shelter the law behind popular sentiment where possible, to oppose it to sentiment as little as possible, and to avoid challenging at the same time the hostility of several kinds of sentiment.

VI. THE PROBABLE ACTION OF THE AGGREGATIVE AND THE DISJUNCTIVE TENDENCIES IN THE FUTURE. Whether in the long run it is the centripetal or the centrifugal force that will prevail in politics, or, in other words, whether large States or small States are more likely to commend themselves to mankind, is a question

which belongs rather to history than to the doctrine of constitutions, and which could be adequately discussed only after a long investigation. History shows us first one force dominant, then the other, though no doubt the centrifugal is usually more powerful in rude times and in hilly or mountainous countries, the centripetal in countries comparatively advanced in civilization, and in level and fertile regions where wealth is more easily acquired and stored, and where military operations are easier. When the mists of antiquity begin to rise sufficiently to show us the Mediterranean and south-west Asiatic world, we discover both a few great States and a multitude of small ones. The former have a low, the latter a high and intense political vitality. From the time of Menes down to that of Attila the tendency is generally towards aggregation: and the history of the ancient nations shows us, not only an enormous number of petty monarchies and republics swallowed up in the Empire of Rome, but that empire itself far more highly centralized than any preceding one had been. When the Roman dominion began to break up the process was reversed, and for seven hundred years or more the centrifugal forces had it their own way. Europe and Western Asia were divided up among innumerable petty potentates, and even the large monarchies, such as the two Khalifates, the Romano-Germanic Empire, the kingdoms of France and Hungary, possessed so feeble a royal authority that the real organs of government and centres of attraction were to be sought rather in the vassals than in the nominal sovereign. From the thirteenth century onwards the tide begins to set

the other way. One great State indeed-the Empirefirst decays and then disappears under the action of centrifugal forces, but all the other chief States expand, absorbing their smaller neighbours, and giving themselves a compact and well-knit organization which makes the central power effective through the whole sphere of its action. This process culminates in the despotic monarchies of the eighteenth century, when the strength of feudal localism has been completely broken, though the picturesque relics of it still cumber the ground, and when at the same time the foundations are laid in the West of a gigantic State which proceeds to cover the temperate area of North America between the two oceans, and, in the East, of the dominion of a European nation which has absorbed the numerous and populous principalities of India. Immediately afterwards the doctrine of popular selfgovernment and the doctrine of nationalities come upon the scene, threatening a disruption of some existing political aggregates. In point of fact, however, these new principles have done as much to unite as to sever, for though five States-Greece, Rumania, Servia, Montenegro and Bulgaria-have been cut off from an effete monarchy, and sixteen republics have been carved out of the American dominions of Spain and Portugal, the doctrine of nationality has substituted two new great States, more important than all the lastmentioned twenty-one put together, for the multitude of kingdoms and principalities which so late as 1859 filled Italy and Germany.

Thus neither Democracy nor the principle of Nationalities has, on the balance of cases, operated to check

the general movement towards aggregation which marks the last six centuries.

It may, however, be said-and this question should be faced before we proceed to inquire whether the aggregative movement is likely to continue-that in all this inquiry we have been ignoring two potent factors. One is Conquest-that is to say, military power. We have been examining the forces of Interest and Sympathy, which cover a number of influences social or economic, racial or sentimental. But after all it is Conquest, i.e. the might of the strongest, which has created most States as we find them. Is Conquest one of the centripetal forces? and if so, is it not the greatest of them?

The other factor is Family Succession, which both during the Middle Ages and since has done a great deal to consolidate principalities and kingdoms. The United Kingdom owes much to this agency, Austria and France

even more.

Conquest and Dynastic Succession are hardly fit to be classed among the centripetal forces, because they are not susceptible of scientific treatment like the other influences. The disposition of the stronger to subdue and annex the weaker neighbour is of course a permanent fact in human nature, and therefore in history. But in each particular instance the success of one or other combatant depends on what may be called historical accidents-on the numbers or the discipline of troops, on the possession of a commander of military genius, on alliances with other States, on the internal dissensions of one State as compared with the unity of another. Physical force belongs to a

different sphere from that in which political constitutions work. Constitutions may result from a conquest or may be maintained for a time by arms; but if they are obliged to rely on and have constant recourse to physical force in order to prevent their overthrow, they are, considered as Constitutions, failures; because the very nature and object of a constitutional Frame of Government is so to express and so to adjust to existing conditions the wishes and aims of the citizens as to make the majority, and if possible the vast majority, of the people desire to support it. According to the proverb, you can do anything with bayonets except sit down on them. Physical force is of course needed to punish occasional infractions of the Constitution or to quell revolts against it. But the system of government which ex hypothesi corresponds to the permanently strongest among the moral forces, else it has no right to prevail in a free country, ought not to be surrounded by cannon.

Similarly, the devolution of princedoms or kingdoms by marriage and inheritance, much as it has done to bring States originally independent under one government, lies outside political science in the proper sense of the term. Like conquest, it brings about a new state of things by an event with which the ordinary political and constitutional phenomena of national life have nothing to do, coming into these phenomena as an incommensurable and (so to speak) irrational factor1.

1 The fact that the custom of a country permits or forbids succession through females makes a great difference in the importance of succession. The union of Castile with Aragon, like the union of England with Scotland, would not have occurred under a different rule of succession. So it may make a difference whether the throne of the larger country passes to the