noticed that there is room for much further evolution, and we could go on splitting up the spectrum indefinitely if only we had the power to distinguish these finer differences, but as a matter of fact I have never met with a normal-eyed man who could see more than twenty-nine monochromatic patches in the spectrum, and there are really millions, though by monochromatic patches I do not mean twenty-nine separate colours. Not only are all the details of the process of the evolution of the colour-sense supported by all the facts that we can obtain from literature and museums, but the theory accounts for facts which were previously inexplicable. The distinction between light-sensation and colour-sensation is explained, and all facts of colour-mixing, complementary colours, and simultaneous contrast. We can understand how, as in many cases which have been recorded, a man may lose his colour-perception and still have an unaltered sense of luminosity and visual acuity. The explanation of complementary colours is a fundamental part of the theory. It is obvious that the two colours of the dichromic are only recognised as different because they are seen in contrast to each other, and that when they are mixed they neutralise each other. It is the same with the other colour-sensations, when they are developed they replace the colours occupying their positions. Therefore green which replaces the grey of the dichromic should be, and is, complementary to the other two colour-sensations, red and violet combined. In the same way when the yellow sensation replaces the red-green of the trichromic it should be possible to compound it of both. Also, when the green sensation is in a feeble state of development it will not have the value that it has at a subsequent stage, and, therefore, yellow will be a much redder colour to those persons than the normal, and in a colour match of red and green forming yellow, more green will be required. Simultaneous contrast is also explained. When two colours are contrasted each appears to be a colour higher or lower, as the case may be, in the spectrum scale; that is to say, the close comparison exaggerates the difference. As the colour-blind have fewer colours, simultaneous contrast should be greater with them, and this I have found to be the case.