had been stationed there by caprice or by favour. If our gods seem nowadays to be purely honorary, the fact was otherwise at a previous period. Religions are not the work of caprice; they correspond to an invincible tendency in man, and sometimes in the lower animals, to try to understand what passes before his eyes. Religion is nascent science, and it was with purely physical problems that it at first essayed to grapple. It was a physics _à côté_, a paraphysics, before becoming a science _au delà_, a metaphysics. CHAPTER II. RELIGIOUS METAPHYSICS. I. Animism or polydemonism—Formation of the dualist conception of spirit—Social relations with spirits. II. Providence and miracles—The evolution of the dualist conception of a special providence—The conception of miracles—The supernatural and the natural—Scientific explanation and miracles—Social and moral modifications in the character of man owing to supposed social relations with a special providence—Increasing sentiment of irresponsibility and passivity and “absolute dependence.” III. The creation—Genesis of the notion of creation—The dualistic elements in this idea—Monism—Classification of systems of religious metaphysics—Criticism of the classification proposed by Von Hartmann—Criticism of the classification proposed by Auguste Comte. _I. Animism._ [Sidenote: Legitimate and illegitimate metaphysics.] The upshot of the preceding chapter is that every religion in its beginning consisted of a mistaken system of physics; and between a mistaken system of physics and certain forms of metaphysics there is often no difference but one of degree simply. Magnify some scientific error, reduce it to a system, explain heaven and earth by it, and it will be a metaphysics—in the bad sense. Whatever one universalizes—error or truth—acquires metaphysical significance, and possibly it is more easy to universalize in this way the false than the true; truth possesses always a greater concreteness than error, and therefore offers greater resistance to arbitrary fashioning. Let a modern man of science develop his knowledge as he will, and enlarge the circle of known phenomena; so long as he holds vigorously by scientific methods he will never be able to pass at a bound from the sphere of phenomena to the sphere of things in themselves. The conscientious man of science is prisoned within the limits of knowledge, his thought has no outlet. But let him once break the chain of logic which confines him, and behold him free. His false hypothesis grows without obstacle or check from reality; he lands at a bound up to his neck in metaphysics. The fact is, one may arrive at a system of metaphysics in two ways—incontinently, by a logical solecism and an exaggeration of some false premise _ad infinitum_, or by following the chain of known truth to the point at which it disappears in eternal night, and