tants. Speculators on the forms of government have, in fact, generally treated the problem in this manner; and though they have used a generic term, they have considered it as designating a limited species, which served in their mind as a representative of the entire genus. Their universal idea was embodied in a real type. Thus, Aristotle, in describing the characteristics of oligarchy and democracy, has in his mind exclusively the oligarchies and democracies of the small states, which existed at and before his time on the coast of the Mediterranean. Montesquieu, again, in speaking of despotisms and monarchies, appears to have mainly in view the Oriental despotisms and the French monarchy of his own time.(138) So M. de Tocqueville, in treating of democracy, confines his regards almost exclusively to the United States, without thinking of the democracies of antiquity and of the middle ages. We shall say more upon this mode of investigating this problem in the next chapter, when we come to the consideration of theories of limited application.

If, in considering the three forms of government (monarchy, aristocracy, and democracy), we confine ourselves to what is strictly involved in the idea of each, namely, the proportionate number of persons who share in the sovereign power; and if we then attempt to lay down general theorems upon the characteristics of each, we shall find that the mere numerical difference in the governors is a narrow foundation for such inferences. If we are investigating the character of a severe or a mild system of punishments, of a free or restricted system of trade, of a direct or indirect system of taxation, we may reasonably expect that the effects of each system would be different; but are we entitled to expect that the conduct of one man will, uniformly or generally, be different from that of a few men, or that the conduct of a few men will, uniformly or generally, be different from that of many men, all other circumstances in which they

(138) See, for example, the chapter of the Esprit des Lois, on the delegation of power, (v. 16,) where the contrast between despotisms and monarchies, though expressed in general terms, is, in fact, between the satrapical system of the Oriental states, and the provincial government of the old French monarchy.

are respectively placed, save their numerical ratio to the community, being supposed to be identical? Again: are we entitled to expect, upon the same supposition, that the conduct of different men, though their numerical ratio to the community may be similar, will always be the same. Unless we can make these two assumptions, it is impossible to lay down any universal attribute of monarchy, of aristocracy, and of democracy, considered as governments in action, and influencing the condition of the community. Unless we can affirm that, other circumstances being the same, the ruling one will always act differently from the ruling few or many; that the ruling few will always act differently from the ruling one or many; that the ruling many will always act differently from the ruling one or few; and, again, unless we can affirm that all persons standing respectively, in the relation to the community, of the ruling one, few, or many, whether simultaneously in different states, or successively in the same state, will always act in the same manner, we cannot infer that any constant operation belongs to either form of government. Before we can arrive at universal propositions on this subject, we must be able to say, not only that every monarch, and every oligarchical and democratic body, will, in the same position, respectively act in a different manner if compared with one another, but also that every monarch and every oligarchical and democratic body will, in the same position, respectively act in the same manner if compared with themselves. For example, if the political com

munity is represented by the number 500,000, then we must assume not only that one ruler, 500 rulers, and 250,001 rulers would, in similar circumstances, always act differently from each other, but that the successive persons in the position of the one ruler, the 500 rulers, and the 250,001 rulers would, in similar circumstances, always act in the same manner as their predecessors; and also, that persons standing to other communities in the same relation as the one ruler, the 500 rulers, and the 250,001, would all act respectively in the same manner, when placed in similar circumstances.

Now the proposition, already adverted to, that 'human nature is always the same,' might lead us to expect that every monarch, and every oligarchical and democratic body, would, in the same circumstances, act in the same manner. But it could not lead us to expect that the mere variation in the numerical ratio to the community would lead the ruling one, few, and many, always to act differently from each other, when placed in similar circumstances. On the contrary, it would rather indicate the probability that their conduct would respectively be similar.(139)

The question, therefore, resolves itself into this—whether, all the circumstances in which the ruling one, few, or many, are placed (with the necessary exception of their numerical relation to the community) being supposed to be identical, we can affirm that they will severally act in a certain manner; and whether, therefore, we can lay down any constant and invariable characteristics of monarchy, aristocracy, and democracy? Does the variation in the numerical ratio of the rulers to the community necessarily involve corresponding variations in the character of the government? As the index in the scale moves along from one to the opposite extreme, indicating the sum of the community, can we predicate a simultaneous change in the operation of the government, as we can point to certain constant changes in outward nature as the thermometer rises from zero to temperate, and from temperate to summer heat?

When the problem is presented in this form, we see that it admits of no universal solution. Nobody can persuade himself that a uniform and invariable character can be separately traced

(139) The reasoning of Spinoza, in his Tractatus Politicus, c. 7. § 27, implies that the few and the many would not act differently if placed in the same circumstances. It is, he says, a frequent mistake to confine to the populace the vices which are common to all men. Human nature is the same in all. We are deceived by power and culture. 'Atque hæc quæ scripsimus, usu forsan excipientur ab iis, qui vitia, quæ omnibus mortalibus insunt, ad solam plebem restringunt; nempe quod in vulgo nihil modicum, terrere ni paveant, et quod plebs aut humiliter servit aut superbe dominatur, nec ei veritas aut judicium, etc. At natura una et communis omnium est. Sed potentiâ et cultu decipimur,' &c. Spinoza here refers to Tacit. Ann. i. 29: Hist. i. 32; ii. 29: Livy, xxiv. 25.

in all monarchies, all aristocracies, and all democracies-that every monarch exactly resembles every other monarch—and that every aristocratic or democratic body is the exact counterpart of every other such body. Let us, then, lay aside this inquiry, and attempt to render the problem more tractable, by omitting to look for absolute effects, and seeking only for tendencies-by substituting a presumptive proposition for one declaratory of invariable consequences.

Without, therefore, looking for invariability of actual effects, we may ask whether the influence of the numerical relation of the rulers to the community, of the interests, feelings, subordinate institutions, usages, and manners, which that relation engenders, is so decisive and powerful as to outweigh the variety of human character and disposition, and to impel all monarchs, and all aristocratic and democratic bodies respectively, in the same direction, though the impulse may be often counteracted or accelerated, or otherwise modified?

§ 8 It is, perhaps, difficult to maintain that all monarchs and all aristocratic and democratic bodies are, by the mere force of their position, and by the circumstances in which they are necessarily placed with respect to the community, urged to adopt a certain course of conduct. Nevertheless, a broad line may be drawn between all despotic or absolute monarchies on the one hand, and all republics, or governments of several (whether aristocracies or democracies), on the other, in respect of the position of the rulers.(140) The despotic monarch, being alone, and legally uncontrolled, has the utmost facilities for abuse of

(140) According to Plato (Leg. iii. 12, p. 693) there are two parent governments-monarchy and democracy. The Persian government is the type of the former, the Athenian of the latter. All other governments are variations from these two.

Sir W. Temple has a similar observation:- Though the old distinctions run otherwise, there seem to be but two general kinds of government in the world; the one exercised according to the arbitrary commands and will of some single person; and the other according to certain orders or laws introduced by agreement or custom, and not to be changed without the consent of many-Essay upon the Original and Nature of Govern ment; Works, vol. ii. p. 30. See also Montesquieu, quoted above, p. 50, n. 74. The government of an absolute prince is contrasted with popular government generally in Eurip. Med. 119-30.

power.

The only restraints upon him are his own humanity, sense of justice, moderation, and prudence, the influence of relations, friends, favourites, advisers, and of public opinion, and the fear of offending powerful persons or bodies. A republican body, on the other hand, however composed, and whether consisting of few or many persons, is, indeed, considered in its corporate capacity, as free from legal control as the despotic monarch; but the persons who compose it, not being individually sovereign, act as legal checks upon one another, and control each other's conduct, not only by their mutual opposition and dissent, but also by their want of sympathy in the passions and appetites of the others. This effect is produced even in a small number; so that, as long as the republican constitution can be preserved, in reality as well as in form, and a substantial division of the sovereign power among co-ordinate members of a body can be maintained, there is a legal restraint upon the conduct of each person, which never can exist in the case of a despotic monarch. No single member of a sovereign body can exercise the supreme power for his own personal gratification. In whatever abuses of power the body may indulge, they must be for a common advantage, real or supposed. Whereas a single despotic ruler may seize women, destroy life vindictively or capriciously, and use the powers of government for satisfying his sensual appetites, his frivolous fancies, or his love of selfish amusement. Gratifications of this sort are essentially individual; the cruelties, the lusts, the caprices, the follies, of a Caligula, a Nero, a Commodus, a Helagabalus, an Eccelino, a Cæsar Borgia, a Tamerlane, a Nadir Shah, or a Louis XV., cannot be put in commission.

An absolute monarch is, as such, necessarily, and by the force of his position, free from all legal restraint or control in every relation of life.(14) Every command of his is legally binding, and he is subject to no law; he is, according to the Roman expression with respect to their princeps or emperor, legibus solutus. On the other hand, it is of the essence of the

(141) The divorce of Josephine by Napoleon affords a good example of the immunity of the despotic monarch from obligations which would