Let us dwell a little longer on the principle of substances. The question is to know whether the idea of subject, of substance, precedes or follows the exercise of the principle. Upon what ground could the idea of substance be anterior to the principle that every quality supposes a substance? Upon the ground alone that substance be the object of self-observation, as cause is said to be. When I produce a certain effect, I may perceive myself in action and as cause; in that case, there would be no need of the intervention of any principle; but it is not, it cannot be, the same, when the question is concerning the substance which is the basis of the phenomena of consciousness, of our qualities, our acts, our faculties even; for this substance is not directly observable; it does not perceive itself, it conceives itself. Consciousness perceives sensation, volition, thought, it does not perceive their subject. Who has ever perceived the soul? Has it not been necessary, in order to attain this invisible essence, to set out from a principle which has the power to bind the visible to the invisible, phenomenon to being, to wit, the principle of substances?[33] The idea of substance is necessarily posterior to the application of the principle, and, consequently, it cannot explain its formation. Let us be well understood. We do not mean to say that we have in the mind the principle of substances before perceiving a phenomenon, quite ready to apply the principle to the phenomenon, when it shall present itself; we only say that it is impossible for us to perceive a phenomenon without conceiving at the same instant a substance, that is to say, to the power of perceiving a phenomenon, either by the senses or by consciousness, is joined that of conceiving the substance in which it inheres. The facts thus take place:--the perception of phenomena and the conception of the substance which is their basis are not successive, they are simultaneous. Before this impartial analysis fall at once two equal and opposite errors--one, that experience, exterior or interior, can beget principles; the other, that principles precede experience.[34] To sum up, the pretension of explaining principles by the ideas which they contain, is a chimerical one. In supposing that all the ideas which enter into principles are anterior to them, it is necessary to show how principles are deduced from these ideas,--which is the first and radical difficulty. Moreover, it is not true that in all cases ideas precede principles, for often principles precede ideas,--a second difficulty equally insurmountable. But whether ideas are anterior or posterior to principles, principles are always independent of them; they surpass them by all the superiority of universal and necessary principles over simple ideas.[35]