This method is independent of his metaphysic, which is substantially that of Locke, and distinguishes between the properties of matter which are primary and given by the objects of touch, and those which are secondary like sounds and colors, and subjective, i.e. effects of the motions of atoms, etc., on consciousness. This assertion leads to a conflict with the principle of the conservation of energy, since the physical forces are already equated in the motions of the atoms, and the soul, therefore, must create the subjective colors, tones, etc., out of nothing.
But this subordination of the data of the other senses to those of touch is a practically convenient method of treating them, rather than a valid derivation, and makes the generation of the qualitative differences of the senses from the quantitative differences of atomic vibrations, etc., incomprehensible. This difficulty is created by ignoring the psychological co-ordination of the various senses. In the order of knowledge, moreover, tones and colors are prior to atomic and ethereal vibrations. Lastly, it is on evolutionist grounds improbable, that if vibrations devoid of color, light, and temperature had been the original fact, they would have been so variously transformed by the senses.
Hence, as a matter of metaphysics, the physicist has either to become an idealist and admit the subjectivity of the data of touch also, or to assert the objectivity also of sounds and colors. The latter view is more natural, and harmonizes with the method of physics, if a parallelism between, e.g. vibrations and colors is substituted for the causation of the second by the first.
The second realistic dogma, that reality belongs only to the normal appearances of things, leads the physicist to assert the dependence of some sounds, colors, etc., on the physiological condition of the organs of sight, hearing, etc., in order to account for abnormal appearances like after-images, fused tones, etc. Science cannot ignore them simply because they are practically unimportant, and so physiology is appealed to. Thus the physiological treatment of perception is throughout based on the physical explanation which it supplements, and so admits neither of a special physiological method nor of a metaphysic. These points are treated in a very detailed and complicated discussion, which shows that physical explanations will go much further in physiology than has generally been admitted, and that the assertion of peculiar and specific physiological activities of the sense organs is untenable. Nevertheless, the physical theory fails to explain, e.g. the dimness of indirectly seen objects (for which a psychological explanation must be adduced) and can provide no satisfactory theory of color. In the last resort this failure is due to the defects of the mechanical theory itself.
The inference from all this is that the metaphysic, though not the method, of physical science has broken down. But on the other hand
