This appears to me an insoluble question, and probably, even, a factitious one. The idea that an object can be modified in its nature or in its aspect comes to us through the perception of bodies. We see that, by attacking a metal with acids, this metal is modified, and that by heating a body its colour and form become changed; or that by electrifying a thread it acquires new properties; or that when we place glasses before our eyes we change the visible aspect of objects; or that, if we have inflammation of the eyelids, light is painful, and so on. All these familiar experiments represent to us the varied changes that a body perceived can undergo; but it must be carefully remarked that in cases of this kind the alteration in the body is produced by the action of a second body, that the effect is due to an intercourse between two objects. On the contrary, when we take the Kantian hypothesis, that the consciousness modifies that which it perceives, we are attributing to the consciousness an action which has been observed in the case of the objects, and are thus transporting into one domain that which belongs to a different one; and we are falling into the very common error which consists in losing sight of the proper nature of the consciousness and making out of it an object. If we set aside this incorrect assimilation, there no longer remains any reason for refusing to admit that we perceive things as they are, and that the consciousness, by adding itself to objects, does not modify them. Phenomena and appearances do not, then, strictly speaking, exist. Till proof to the contrary, we shall admit that everything we perceive is real, that we perceive things always as they are, or, in other words, that we always perceive _noumena_.[26] After having examined the relations of the consciousness with its objects, let us see what concerns the perception, by the consciousness, of the relations existing between these objects themselves. The question is to ascertain whether the _a priorists_ are right in admitting that the establishment of these relations is the work of the consciousness. The role of synthetic power that is thus attributed to consciousness is difficult to conceive unless we alter the definition of consciousness to fit the case. In accordance with the definition we have given and the idea we have of it, the consciousness makes us acquainted with what a thing is, but it adds nothing to it. It is not a power which begets objects, nor is it a power which begets relations.