When we go into metaphysics we are continually astounded to see how different conceptions of things which have a classic value are independent of each other. In general, phenomenism is opposed to substantialism, and it is supposed that those who do not accept the former doctrine must accept the latter, while, on the contrary, those who reject substantialism must be phenomenists. We know that it is in this manner that Berkeley conquered corporeal substantialism and taught phenomenism; while Hume, more radical than he, went so far as to question the substantialism of mind. On reflection, it seems to me that, after having rejected phenomenism, we are in no way constrained to accept substance. By saying that we perceive things as they are, and not through a deluding veil, we do not force ourselves to acknowledge that we perceive the substance of bodies--that is to say, that something which should be hidden beneath its qualities and should be distinct from it. The distinction between the body and its qualities is a thing useful in practice, but it answers to no perception or observation. The body is only a group, a sheaf of qualities. If the qualities seem unable to exist of themselves and to require a subject, this is only a grammatical difficulty, which is due to the fact that, while calling certain sensations qualities, we suppose a subject to be necessary. On the other hand, the representation which we make to ourselves of a material substance and its role as the support of the qualities, is a very naive and mechanical representation, thanks to which certain sensations become the supports of other and less important sensations. It would suffice to insist on the detail of this representation and on its origin to show its artificial character. The notion we have of the stability of bodies and of the persistence of their identity, notwithstanding certain superficial changes, is the reason for which I thought proper to attribute a substance to them, that is to say, an invariable element. But we can attain the same end without this useless hypothesis; we have only to remark that the identity of the object lies in the aggregate of its properties, including the name it bears. If the majority of its properties, especially of those most important to us, subsists without alteration, or if this alteration, though of very great extent, takes place insensibly and by slow degrees, we decide that the object remains the same. We have no need for that purpose to give it a substance one and indestructible. Thus we are neither adherents of phenomenism, nor of substantialism.] [Footnote 27: I borrow from RABIER this argument, which has thoroughly convinced me (see _Psychologie_, p. 281).] [Footnote 28: PILON is the psychologist who has the most forcibly demonstrated that resemblance acts before being perceived. I refer the readers to my _Psychologie du Raisonnement_, where I have set forth this little problem in detail.]