But here an objection occurs: precisely the same course seems to be followed by the mind in the establishment of science. In a scientific period man feels himself as profoundly _dependent_ as in a religious period, and this feeling of dependence is accompanied by a no less vivid reaction in the one case than in the other. The man of science and the believer alike aim at enfranchisement, but by different means. Must one be content, then, with an external and negative definition, and say with M. Darmesteter: “Religion embraces all knowledge and all power not scientific”?[1] A knowledge not scientific possesses all the attributes of a contradiction in terms, and, as for a power not scientific, it is indispensable to distinguish it in some positive way from the power which is afforded us by science. Well, to keep close to the facts, the power of religion is that which we frankly do not possess, while the power of science is that which we do possess and know that we possess. One might indeed fall back on the distinction between belief and certainty; but the man of science also has his beliefs, his preferences for such and such a cosmological hypothesis, which, however, is not a religious belief, properly so called. Religious and moral “faith,” as opposed to scientific “hypothesis,” is an ultimate and very complete manifestation of the religious sentiment, which we shall examine later, though it carries with it no suggestion of its primitive origin. [1] See an account given of the _Prolégomènes_ of M. Albert Réville, by M. Darmesteter, _Revue philosophique_, seventh year, vol. i. p. 76.