A few convinced materialists and parallelists, to whom I have read the above criticisms on their systems, have found no answer to them; my criticisms have appeared to them just, but nevertheless they have continued to abide by their own systems, probably because they were bound to have one. We do not destroy an erroneous idea when we do not replace it by another. This has decided me to set forth some personal views which, provisionally, and for want of better, might be substituted for the old doctrines. Before doing this, I hasten to explain their character, and to state openly that they are only hypotheses. I know that metaphysicians rarely make avowals of this kind. They present their systems as a well-connected whole, and they set forth its different parts, even the boldest of them, in the same dogmatic tone, and without warning that we ought to attach very unequal degrees of confidence to these various parts. This is a deplorable method, and to it is perhaps due the kind of disdain that observers and experimentalists feel for metaphysics--a disdain often without justification, for all is not false, and everything is not hypothetical, in metaphysics. There are in it demonstrations, analyses, and criticisms, especially the last, which appear to me as exact and as certain as an observation or experiment. The mistake lies in mixing up together in a statement, without distinction, the certain with the probable, and the probable with the possible. Metaphysicians are not wholly responsible for this fault of method; and I am much inclined to think that it is the natural consequence of the abuse of speculation. It is especially by the cultivation of the sciences of observation that we foster in ourselves the precious sense of proof, because we can check it any minute by experimental verification. When we are working at a distance from the facts, this sense of proof gets thinner, and there is lost that feeling of responsibility and fear of seeing one's assertions contradicted by a decisive countervailing observation, which is felt by every observer. One acquires the unbearable pride which I note in Kant, and one abandons one's self to the spirit of construction. I am speaking from personal experience. I have several times detected within me this bad spirit of construction, I have been seeking to group several facts of observation under the same idea, and then I have discovered that I was belittling and depreciating those facts which did not fit in with the idea. The hypothesis I now present on the relations of the mind and the brain has, for me, the advantage of bringing to light the precise conditions which a solution of this great problem must satisfy for this solution to be worthy of discussion. These conditions are very numerous. I shall not indicate them all successively; but here are two which are particularly important.