for admitting the relativity of knowledge without destroying its reference to reality. The existence of reals other than the self is to him a primary datum of consciousness.
In point of methodology, Renouvier's constant demand is that philosophy shall free itself from the habit of hypostasizing concepts derived by abstracting from experience. The fallacy of mediæval realism he finds infecting "almost the whole of metaphysical speculation, ancient and modern." He is fond of presenting Neo-Platonic emanationism, with its elaborate effort to derive determined being from an absolutely undetermined source, as a completed type of this, which he calls the 'realist method.' For example, Spencer's two concepts, the 'unknowable' (undetermined source of the world), and 'force' (supreme concept of determinate being), are said to correspond respectively to the 'One' and the Nous of Plotinus.
We are now ready to define the relation to this system of the two works under consideration. The Dilemmes de la métaphysique pure is essentially a study of philosophic method—a physiology of philosophy. Its plan is simple. The products of the 'realist method' are summarized in a group of articulated propositions, over against which is placed a corresponding exhibit of the fruit of the 'relativist method.' The 'realist method' endeavors to conceive reality as something 'in itself,' while the 'relativist method' frankly defines every object of thought by means of its relations (Ch. I). Similarly, substance is conceived, on the one hand, as independent of its qualities, and on the other hand, by means of them (Ch. II). The notion of a world infinite in extent, duration, and composition is next placed in opposition to that of a whole which is a determinate number (Ch. III). In the same way, determinism is contrasted with freedom (Ch. IV), and thing with person (Ch. V). The resemblance of all this to Kant's antinomies is only external, for the author constantly brings out the fallacy of the universal invariably hidden in the thesis, and in the end he traces the two sets of propositions to their root in a single problem which he declares to be rationally soluble. This is the problem of freedom. Assume determinism, and there follow the infinite regress of modes, the universal substance, the unconditioned One ; but if freedom be granted, we proceed thence to the notion of personality as reality, the infinite regress is escaped, and with it all the abstractness of the deterministic scheme. The final decision for freedom is based upon the reasoning of Lequier. If we accept the hypothesis of determinism, then the contradictory theories actually held regarding determinism are equally determined. Thus the deterministic hypothesis brings thought to a deadlock, which can be broken only by assuming the truth of its contradictory.
So much for the first of the two volumes. The second, the Histoire et solution des problèmes mètaphysiques, illustrates the fundamental contradiction discussed in the Dilemmes by a critical review of the entire history of philosophy. The intention is to show that 'relativism' and 'personalism' furnish a norm and a corrective for philosophy at all stages of its develop-
