Still more recently, however, a revulsion of opinion has taken place. Against the physicists, the mathematicians in particular have risen up, and taking their stand on science, have demonstrated that all the mechanisms invented have crowds of defects. First, in each particular case, there is such a complication that that which is defined is much more simple than the definition; then there is such a want of unity that quite special mechanisms adapted to each phenomenal detail have to be imagined; and, lastly--most serious argument of all--so much comprehensiveness and suppleness is employed, that no experimental law is found which cannot be understood mechanically, and no fact of observation which shows an error in the mechanical explanation--a sure proof that this mode of explanation has no meaning. My way of combating the mechanical theory starts from a totally different point of view. Psychology has every right to say a few words here, as upon the value of every kind of scientific theory; for it is acquainted with the nature of the mental needs of which these theories are the expression and which these theories seek to satisfy. It has not yet been sufficiently noticed that psychology does not allow itself to be confined, like physics or sociology, within the logical table of human knowledge, for it has, by a unique privilege, a right of supervision over the other sciences. We shall see that the psychological discussion of mechanics has a wider range than that of the mathematicians. Since our cognition cannot go beyond sensation, shall we first recall what meaning can be given to an explanation of the inmost nature of matter? It can only be an artifice, a symbol, or a process convenient for classification in order to combine the very different qualities of things in one unifying synthesis--a process having nearly the same theoretical value as a _memoria technica_, which, by substituting letters for figures, helps us to retain the latter in our minds. This does not mean that figures are, in fact, letters, but it is a conventional substitution which has a practical advantage. What _memoria technica_ is to the ordinary memory, the theory of mechanics should be for our needed unification.