operation. The order followed in the genesis of the science of geometry is typical of the whole course of scientific development. In each case practical exigencies, or exceptional conspicuousness, have called attention to phenomena of a special kind-to the movements of the heavenly bodies, to the play of mechanical forces, to the composition of material substances, to the structure or functions of the human body—from the investigation of which have arisen the sciences of astronomy, of mechanics, of chemistry, of anatomy, of physiology. Each science, called into existence by the anxiety to explain striking experiences, or to provide and justify practical expedients, has taken in charge some special and limited order of relations, has detached these from the mixed and heterogeneous body of physical phenomena, and has made them the subject of isolated and special study. The laws of the various orders of physical relations have thus been determined; and the rays of scientific light extracted from the separate investigations of perfectly independent workers have been made to converge in elucidation of the actual composite facts of the outer world.

This has been the course of development in physical science, the method by which the secrets of external nature have been unlocked. It has been a method, not of study in the ensemble, but of study through the elements-of analysis followed by synthesis. In perfect analogy with this mode of proceeding is the political economist's conception of the path of inquiry to be followed in dealing with the facts of social life. He proposes to break them up into their elementary groups, and he takes one of these groups-the phenomena of wealth as the subject of his special investigations. It may be remarked, moreover, that, in selecting this particular group of phenomena, he has been influenced by considerations in all respects analogous to those which have determined the separate treatment of the various classes of physical phenomena. Political Economy, like Geometry, Astronomy, Mechanics, Chemistry, had its origin in practical exigencies, and made its début as an art. It aimed at the practical object of enriching particular nations by means of trade. For this purpose complicated machinery-encouragements for particular industries, prohibitions of others, bounties, drawbacks, in a word the whole body of commercial regulations known as the Colonial and Mercantile systems-was put in force. These expedients, if they favoured some interests, damaged others: the conflict of interests brought on discussion; and the argument rapidly passed from attack and defence of practical plans, to examination of the natural laws governing the order of relations, which it was the purpose of these plans to control. The limits of the debate were not at first, perhaps, very distinctly defined, but by degrees they grew clear. The facts of wealth became detached for the purposes of discussion from the other classes of facts with which in actual existence they were

blended; and Political Economy, as the science of those facts, emerged. As regards origin and, mode of development, therefore, the parallel between Political Economy and the physical sciences is complete; nor have I any reason to suppose that M. Comte would dispute the general correctness of the description I have given: indeed, he frankly admits that the precedents of physical science are against him. What, then, is his line of argument? It is this: he contends that the cases are not similar; that the problems presented, on the one hand by physical nature, on the other by social life, are so radically discrepant that the method applicable to the one must be, not only modified, but reversed, in dealing with the other. To follow in social inquiry the precedents of physical research is, according to M. Comte, in oblivion of essential distinctions, to practise a "blind imitation." This is the position which we are now called upon to consider.

Most people who take an interest in questions of the kind we are now discussing, are familiar with M. Comte's classification of the sciences. As is known, it proceeds upon the plan of arranging the various branches of scientific knowledge in the order indicated by the degree of complexity of their subject-matter. Thus it places first in the scale the sciences which deal with the most simple order of relations-number and extension. After these comes mechanics, as involving relations one degree more complex; next to mechanics, astronomy, which is followed by physics, and so on through the entire encyclopædia of scientific knowledge; each science, according to its place in the scale, representing a degree of complexity greater than those preceeding and less than those following it. It results from the principle of the arrangement that the organic sciences, having for their subject-matter the complex phenomena of the vegetable and animal world, should occupy the later portion of the scale, and that Sociology, or the science of human society, as concerned with the most complex of all phenomena, should conclude and crown the whole. As regards the merits or demerits of this classification-a question on which the highest authorities are not agreed-it would be unbecoming in me to pretend to express an opinion. I only refer to it in order to render M. Comte's argument against Political Economy intelligible. As has been said, then, the sciences are arranged in the order indicated by the degree of complexity in their subject-matter; those occupying the first or lower portion of the scale embracing phenomena but little complex, while the phenomena embraced by the sciences in the later portion are complex in a high degree. It is on this distinction that M. Comte grounds his argument for disregarding in sociological speculation the precedents furnished by physical research. According to him, the method of investigation that has been followed in the study of physical nature, the method, that is to (1) Philosophie Positive, vol. iv. pp. 353-54. 8139.Ed.

say, which proceeds by breaking up composite phenomena into the elementary groups composing them, studying apart the elementary groups, determining their laws, and afterwards combining these laws. in explanation of the original aggregates-this method, according to M. Comte, owes its efficacy to the uncomplex character of the phenomena submitted to the process. As phenomena become more complex, the method, he contends, becomes less suitable, less efficacious, till at length a point is reached at which it fails altogether, and it becomes necessary to adopt a contrary mode of procedure, the mode of procedure, namely, which he describes as investigation through the ensemble. This point in the scale of the sciences coincides, he tells us, with that at which the transition is made from inorganic to organic nature. The method of investigation by disintegration and separate study should thenceforth give way to that which proceeds by treatment in the ensemble. Accordingly, he holds that the organic sciences generally should be cultivated in conformity with this principle; but in the study of social phenomena, the most complex and intricate of all, the rule becomes absolute and imperative.

And here one is led to ask why the method of specialisation should lose its efficacy as problems become more complex? The very opposite is what one would naturally expect. If a problem involving no more than two or three distinct elements can only be resolved by the process of analysis and separate consideration of the parts, the necessity for this would seem to be still more urgent as the elements engaged become more numerous. M. Comte's reason for reversing this inference is very peculiar.1 He says that as phenomena become more complex, the elements composing them become more solidaire. In the physical universe, the complexity of the phenomena is not great, and consequently their "solidarity" is but "slightly pronounced:" "the elements are here better known. to us than the ensemble." But the reverse is the case with the organic world, and more especially with that portion of the organic world which constitutes the social organism. The phenomena are here characterised by a very high degree of complexity, and therefore, says M. Comte, by a very high degree of solidarity: "the ensemble of the subject is better known to us and more accessible than the parts." On the fundamental principle, then, of inductive logic, which requires us to proceed from the known to the unknown, from

(1) This argument has appeared to me so weak-indeed, M. Comte's whole case against Political Economy is, as it seems to me, so weak, that I have felt it difficult at times to repress the suspicion that his reasons for rejecting it were not purely and simply of a philosophical kind. "Il s'agit malheureusement," he says in one passage, "et sans que rien puisse m'en dispenser, de tenter une création philosophique qui n'a jamais été jusqu'ici ébauchée ni convenablement conçue par aucun de mes prédécesseurs." "Sociology could not be constructed in its entirety by M. Comte if Political Economy were a legitimate speculation. But M. Comte felt it to be his mission to construct Sociology in its entirety. The conclusion seems evident.

the better to the less known, we are bound, in dealing with the phenomena of organic nature, but more especially with the phenomena of society, to begin our investigations with the study of aggregates, and only after we have determined their laws to address ourselves to that of the less known elements. M. Comte admits that this mode of proceeding must "gravely augment" the fundamental difficulties already incident to the extreme complication of the subject-matter; but this, he conceives, is only a reason for reserving the study of society for "the highest scientific intelligences."

In attempting to criticise this argument, it becomes necessary to assign a distinct meaning to its several propositions. We encounter, in the first place, the expression, "the ensemble of society," and the statement that this is better known to us than the "elements." In the most obvious meaning of the word the statement is manifestly not true. By the ensemble of society most people would, I think, understand the aggregate of the human beings composing society -of those human beings considered in their social relations; and by the "elements," the individual social men and women. In this sense I say it is manifestly untrue that we know society better in its ensemble than in its "elements,"-so manifestly so, that it cannot for a moment be supposed that this was M. Comte's meaning. When, for example, an Englishman travels in France, it is not with the ensemble of French society that he comes into contact, but with certain railway officials and hotel proprietors exemplifying a very limited range of French social existence. As he prolongs his residence he may extend his knowledge; but the course which his acquisitions take will, I need scarcely say, be in the opposite direction of that which M. Comte's maxim affirms. Nor can a French philosopher attain a knowledge of French social existence by any different path; he, too, must proceed from individuals to classes, and from classes to the social whole. But there is another sense in which M. Comte's language may be understood. Social phenomena, like all other phenomena, meet us not simple, but composite. We do not encounter purely religious, or purely industrial, or purely political men and women. Social acts, social situations, can rarely be referred to any single influence. Human beings, as they exist, are not abstract, but historical, human beings, in a greater or less degree, under the influences of all the causes that have been affecting the race from its origin down to the present time. Thus regarded, society, or more properly social phenomena, may be said to present themselves to us in the ensemble; and thus understood, the statement that we know society through its ensemble, not through its elements, is undoubtedly true. If this be M. Comte's meaning, the proposition cannot be disputed; but then it must be remarked that the assertion is equally true as applied to the phenomena of the physical universe. Physical forces also act in constant conjunction.

Unless we effect the separation by artificial means we encounter no purely chemical, or purely optical, or purely mechanical phenomena; but phenomena in the production of which a variety, greater or less, of physical forces concur-that is to say, we know physical nature also through its ensemble. We are thus brought back to the point from which we started, why are we-the phenomena of social life and those of physical nature being made known to us under similar conditions to reverse in our study of society the method of investigation which has been found efficacious in dealing with the physical world ?

M. Comte's reply at this stage of the argument resolves itself into the doctrine I have already stated, that the solidarity of phenomena varies directly with their complexity. It is true, he seems to admit, that we know physical nature equally with social through its ensemble; but the ensemble, in the former case, is composed of fewer elements, and these, in proportion as they are fewer, are less solidaire, are therefore more easily broken up and submitted to separate examination. Hence arises an increased facility of applying the method of disintegration and separate study in their case. But, in the first place, this does not meet the difficulty, since the answer admits that physical nature is known to us through its ensemble—an admission which, on M. Comte's principles seems to draw with it the obligation of studying physical nature through this, its most familiar manifestation. Waiving, however, this point, I wish to examine M. Comte's position, which is really the root of his whole argument against Political Economy, that phenomena in proportion as they are more complex are more solidaire. If this assumption be not wellfounded, there is absolutely nothing for his reasoning to rest upon.

To test the doctrine, let us consider it in a concrete case. I take the instance of water, a composite physical phenomenon exemplifying a variety of physical laws. Considered chemically, its complexity is of the lowest degree, containing as it does but two elements, oxygen and hydrogen. According to M. Comte's doctrine, water, being chemically of the lowest degree of complexity, ought to exhibit, in the relation of its chemical elements, the lowest degree also of solidarity. The fact, I need scarcely say, is exactly the reverse. As every one knows, the solidarity-by which I understand intimacy of relationship, closeness of interdependence-existing between the elements composing water is of an extremely intense kind, so much so that the analysis of water constituted an epoch in chemical history. On the other hand, if we take a phenomenon of greater complexity, say water in combination with lime, we find the solidarity diminish as the number of the elements is increased; the water or the lime being much more easily detached from the hydrate of lime than the elements composing the water, or than those composing the lime, are from each other. Nor is this a solitary