The truth is that Taine vainly tried to establish a phenomenal doctrine, not purely empirical in character like that of Hume, but a phenomenalism wedded to a necessity which is supposed to be self-explanatory. Such a notion of necessity, however, is formal and abstract. Rather than accept Taine’s view of a law, a formula, an “eternal axiom” at the basis of things, we are obliged to postulate an activity, creative in character, of whose action universal laws are but expressions. Law, formula, axiom without action are mere abstractions which can of themselves produce nothing. Taine’s positivism, however, was not so rigid as to exclude a belief in the value of metaphysics. It is this which distinguishes him from the Comtian School. We see in him the confidence in science complemented by an admission of metaphysics, equivalent to a turning of “positivism” in science and philosophy against itself. Much heavier onslaughts upon the sovereignty of science came, however, from the thinker who is the great logician and metaphysician of our period, Renouvier. To him and to Cournot we now turn. II While Taine had indeed maintained the necessity of a metaphysic, he shared to a large degree the general confidence in science displayed by Comte, Bernard, Berthelot and Renan. But the second and third groups of thinkers into which we have divided our period took up first a critical attitude to science and, finally, a rather hostile one. Cournot marks the transition between Comte and Renouvier. His _Essai sur les Fondements de nos Connaissances et sur les Caractères de la Critique philosophique_ contains some very calm and careful thought on the relation of science and philosophy, which is the product of a sincere and well-balanced mind.[16] He inherits from the positivists an intense respect for scientific knowledge, and remarks at the outset that he is hostile to any philosophy which would be so foolish as to attempt to ignore the work of the modern sciences. [16] See in particular the second chapter of vol. 2, _Du Contraste de la Science et de la Philosophie et de la Philosophie des Sciences_, pp. 216-255.