Wagner's gives the feeling of an atmosphere of light, changing its colors in succession.
Having established the mental nature of the impressions of color, we come now to seek the cause of their apparition. We know pretty well what one means when he declares that a is red; but we have not explained how the idea or perception of a sound can awaken the idea of a particular color. Our ideas have generally a logical origin; we are at least in the habit of believing this, and it often occurs, in analyzing our representations, that we find the cause that brings them out and connects them. If I hear a bell, and, without seeing it, conceive its roundish form, its clapper, and its dark-green color, the connection of ideas is understood to be natural, useful, and true; it is derived from previous experiences. It is a piece of the outer world registered in my mind. But these associations of colors with sounds are factitious, have a purely individual character, and correspond with nothing in the order of external facts. A sound is a sound, and has nothing in common with a color. The human voice is grave or sharp, and is not yellow or green. How has such an association been created and developed in the face of good sense? It is evident that the act of establishing tenacious associations between impressions that have nothing in common is the sign of some intellectual form which is not everybody's. We are disposed to attach some importance to the quality of the illusions evoked. They are of a visual character, which seems to indicate that there exists in colored audition an intense rush of visual images and a tendency to think as well as to feel with them; in short, we suppose that those who have colored audition belong to the category of visuals, or persons who, according to the classification of M. Charcot and many physiologists following him, have visual memories. As the case of M. Inaudi enabled us to study a high development of the auditive memory, another category in this classification, colored audition, will perhaps permit us to study visual memory. This is only a hypothesis; for it is not absolutely certain that colored audition always agrees with the type of visual memory, and that there is a causal relation between the two things, but we do not advance it without the support of good reasons.
First, we have the testimony of the subjects whom we have had, opportunities to question. We addressing them in a tone of indifference and without trying to dictate their responses, they have' remarked that colors and forms are the things they remember most easily. A young woman to whom I sent my requests in writing to avoid the unconscious suggestions of accent, answered me, i "You ask me if I more easily recollect things seen or things] heard; things seen. When I recollect a conversation, the gestures
