I have shown elsewhere, by experiments on hysterics, that it is possible by the intermediary of their insensibility to touch to suggest ideas on the value of which the patients make mistakes. For instance, you take the finger in which they have no sensation, you touch it, you bend it. The patient, not seeing what is done, does not feel it, but the tactile sensation unfelt by their principal consciousness somehow awakes the visual image of the finger; this enters into the field of consciousness, and most often is not recognised by the subject, who describes the occurrence in his own way; he claims, for instance, that he thinks of sticks or of columns. In reality he does not know of what he is thinking, and we know better than he. He is thinking of his finger, and does not recognise it. All these examples show that the clearly defined characteristics into which it is sought to divide extrospection and introspection do not exist. There is, however, a reason for preserving the distinction, because it presents a real interest for the psychology of the individual. These two words introspection and extrospection admirably convey the difference in the manner of thinking between those who from preference look, and those who from preference reflect. On the one hand, the observers, who are often men of action; on the other, the speculators, who are often mystics. But it would be no more legitimate by this means to separate psychology and physics than to say, for instance, "There are two kinds of geology: one is the geology of France, for one is acquainted with it without going from home, and the other is that of the rest of the world, because in order to know it one must cross the frontier." We reject, therefore, the definition drawn from the difference of method. At bottom there is no difference of method, but only differences of process, of _technique_. The method is always the same, for it is derived from the application of a certain number of laws to the objects of cognition, and these laws remain the same in all spheres of application. Here is another difference of method which, if it were true, would have an incalculable importance. Psychology, we are told, is a science of direct and immediate experiment; it studies facts as they present themselves to our consciousness, while the natural sciences are sciences of indirect and mediate experiment, for they are compelled to interpret the facts of consciousness and draw from them conclusions on nature. It has also been said, in a more ambitious formula, "The science of physical objects is relative; logical science is absolute."