criticist is not as such a sceptic, but he is more likely to fall into scepticism than into dogmatism. The criticist often holds phenomenalism and relativism as narrowly and exclusively as the positivist, but he has always more reason for holding them, and a clearer conception of what he means by them.

The criticist mode of thought has found in France its two most typical representatives in the late M. Cournot and M. Renouvier. Both have occupied themselves with historical philosophy. They have written in entire independence of each other. While both may be regarded as in a general way disciples of Kant, neither has sacrificed to Kant, or any other thinker, his own rights of private judgment.

M. Augustin Cournot (1801-77) had a remarkable capacity both for speculative thought and scientific research. He filled difficult and important educational positions. He wrote valued works on the higher branches of mathematics. The treatises in which he attempted to apply mathematics to economics have been allowed by competent judges to be among the most ingenious and successful of their kind. He expounded his philosophical opinions in the Essai sur les fondements de nos connaissances,' 2 vols., 1851; the 'Traité de l'enchaînement des idées fondamentales dans les sciences et dans l'histoire,' 2 vols., 1861; and Considérations sur la marche des idées et des événements dans les temps modernes,' 2 vols., 1872. These are all most instructive and suggestive books, such as could only be produced by a mind of rare intellectual sincerity, thoroughly disciplined in exact science and in the practice of analysis, and with a grasp of facts at once capacious and firm: books not written with a view to being easily read, and to please, impress, or astonish; not written for a vulgar and thoughtless public, but for the only public worthy of them, one which earnestly seeks truth precisely as it is, truth in its purity, naked, unexaggerated, and unadorned. The last mentioned of them is of most interest for the philosophical historian.

Cournot's conception of philosophy is peculiar. He does not admit it to be a science, inasmuch as he holds it neither to have a definite object nor to be capable of furnishing demonstrative proof or certainty. To represent it as being, or capable

of being, science can only tend, in his opinion, to spread and confirm the pernicious impression that it is nothing real at all, but merely a pretentious illusion. It has no particular object, for whatever objects there may be they are the proper subjects of particular sciences, mathematical, physical, biological, noölogical, or political. Nor does it deal, as Comte taught, with the whole of the generalities of the sciences, the sum of certainties established by the sciences: these generalities and certainties must always belong to the sciences which prove them. Philosophy is an indispensable element of all the sciences, a spirit which inspires and vivifies them. Its conclusions are not certainties. Every philosophy, so far as it embodies itself in doctrines, is only a whole of more or less probable views relative to the order and the reason of things. Cournot's conception of philosophy is thus entirely different from Auguste Comte's. The latter would have all problems which do not admit of a positive solution wiped out; all questions which cannot be definitely settled by experience and scientific proof denied the right of being put. He was by nature and on system intolerant of doubts, questionings, hesitations of belief. Cournot shows himself profoundly conscious that a finite intellect must be a fallible intellect; that man as a conditional being cannot have a strictly absolute certainty; that it is not merely human to err, but that the possibility of error is so involved in the very constitution of the human mind that it cannot be thought of as absent from it; that in all perception, all consciousness, all reasoning, there lurks, and must ever lurk, this possibility; and that we must often resign ourselves to be guided, even in matters of high concern, by low probabilities. In his view all that we can say of the most completely verified laws of nature is that they are infinitely probable; and "speaking physically, infinite probability is equivalent to reality, but logically speaking it is never more than a probability." It is just those questions which most interest and concern humanity which are generally least susceptible of scientific treatment; and therefore it is no disparagement to philosophy to represent it as occupied with such questions.1

1 There is a good study on the general philosophy of Cournot by T. V. Charpentier, in the 'Rev. Phil.,' t. xi.

Cournot's philosophy of history is merely an historical etiology, an analysis and discussion of the causes and concatenations of causes which have concurred to bring about the events of which history presents us with the picture. It is not simply the history either of civilisation or of humanity, for universal history has its etiology just as have the histories of religion, science, morality, policy, art, and industry, or, in other words, the special historical developments which it includes. Nor is it the ambitious and hypothetical teleology of history, to which the name of philosophy of history has been so often given. M. Cournot does not contest that the course of humanity proceeds according to a fixed plan and towards a decreed or designed end; but he thinks that all attempts to trace such a plan and determine such an end are plainly defective and unreliable, and that the most celebrated of them, like those of Hegel and Cousin, although they might be received with applause around a professorial chair, are worthless before criticism, the only good kind of philosophy. He abjures for his own part such venturesomeness. His historical philosophy is critical, not speculative. It allows the use of hypotheses only in so far as they suggest, or are suggested by, inductions.

Cournot rejects the Comtian law of the three states, and, succinctly but conclusively, shows its inconsistency with facts. He does not attempt to replace it by another; he does not even venture to affirm that there is any law of history. Defining a law of nature to be "a constant mathematical relation between two variable quantities," he finds nowhere in history laws corresponding to his definition. It is not laws, therefore, which he seeks in history, but causes or reasons, connections and relations. "Whether there are or are not laws in history, it is enough that there are facts, and that these facts are sometimes subordinate to one another, sometimes independent of one another, in order that there may be room for a criticism designed to trace out in the one case the subordination and in the other the independence. And as this criticism cannot pretend to irresistible demonstrations, such as produces scientific certainty, but is restricted to the setting forth of analogies and inductions, like those with which philosophy must be content (otherwise it would be a science, as so many people have vainly

pretended it to be, and not philosophy), it follows that we are quite entitled to give this criticism of which we are speaking, and which, notwithstanding its uncertainties, is of so much interest, the name of 'philosophy of history.' The same holds of the history of peoples as of the history of nature, which is not to be confounded with the science of nature, seeing that the one has chiefly for object facts and the other laws, but facts which may be on so great a scale, and have consequences so vast and durable, that they appear to us to have, and really have, the same importance as laws. None the less reason recognises a radical difference between laws and facts: the former valid always and everywhere, by a necessity inherent in the permanent essence of things; the latter brought about by a concurrence of anterior facts, and determining in their turn the facts which are to follow them."1

Cournot considers it essential to a correct understanding of history to distinguish between necessary and fortuitous events, and to assign a considerable place to the latter. He holds that the idea of chance or hazard is not a mere phantom evoked by the mind to hide from itself its own ignorance, or to express the imperfection of its knowledge in certain circumstances and conditions, but the notion of a fact true in itself, demonstrable in some cases by reasoning, and more commonly confirmed by observation. The fact which it implies is the independence of series of causes which, although unrelated, do in fact concur to produce certain phenomena or events, which are on this account appropriately termed fortuitous. Such independence of series of causes Cournot regards as quite consistent with belief in their common suspension to a single primordial ring beyond, or even within the limits to which our reasonings or observations can attain. There is, in his view, no opposition between chance properly understood and Providence, between hazard and Divine Will or Fate. An accidental fact does not mean an effect without a cause, or a fact which human wisdom cannot in any measure foresee or provide against, but a fact brought about by the interaction of chains or groups of facts which are not naturally connected. Were there no facts of this kind there could be no history, but only science. Were all facts of this

kind there could equally be no history, but only annals. History properly so called implies the commingling of fortuitous and necessary facts. The part of fortuity, according to Cournot, is especially large in political history, as the action of exceptional and superior personalities has there most effect; it diminishes, however, as general causes, the collective reason and will, attain ascendancy. Inasmuch as the efficiency of fortuitous events may be extensive and even permanent, particularly in the political sphere, the student of historical etiology must be on his guard against overlooking them; at the same time, political history, in which hazard has most influence, is for the historical etiologist not the first but the last department of history, the most superficial, particular, and external. On this very account, however, political history is always the chief object of interest to the ordinary historian, constitutionally incapable of general and philosophical views.

With characteristic caution M. Cournot refrains from attempting to survey the course of history as a whole, and confines his reflections chiefly to modern times. He has, however, some introductory chapters on the medieval period; and in these he characterises with remarkable sagacity its general spirit, its scientific condition, its scholastic philosophy, its ecclesiastical organisation, and its feudal constitution. He shows very clearly how it ought to be differentiated from ancient and modern history. It is to be regretted that the late Professor Freeman did not become acquainted with his observations on the division of history into "ancient" and "modern." He could hardly have failed to learn from them that there was more to be done in relation to that division than simply to assail it and condemn its abuses; that it was also necessary to inquire how far it is legitimate, and what the terms ancient and modern, old and new, when applied to history and historical phenomena, really mean.

Even of the limited period of history selected by him for investigation, Cournot does not attempt to give a systematic survey, to trace in it the operation of laws, or to formulate its characteristics and results. His treatment of it is comprehensive, but not deductive or constructive; it has no other unity than that which arises from sameness of spirit and method. His