Here we have both the colours which the green-blind calls white side by side. If the brightness of each be the same, he would see no difference in them. Is it possible that on any theory this can be correct? To explain this apparent paradox, and without reference to the mathematical proof that white subtracted from white leaves white, we have only to look at our diagram (Fig. 16), and it is immediately apparent how it arises. The red and the blue curves cut at this point; and if we take away the green sensation entirely, the residue will be a mixture of the red and blue, which is the identical purple colour forming the patch. If we are wishful to ascertain the colour that the violet-blind calls white, we have only to cover up the violet slit and a yellow is left behind as the result. I would have you remark that these colours which are seen as white would only be of the hues shown you, supposing the colour sensations were identical with those in normal vision. Whether this is the case we cannot absolutely say, and the only way in which this can be authoritatively settled is by examining some person who has normal colour vision in one eye and defective colour sense, _not due to disease_, in the other. One such person has been examined abroad, but in what way I am unable to say. It is recorded that he sees the red end of the spectrum as yellow with the eye that is defective. Another person I have heard of in England, but so far have not had the good fortune to get hold of him for examination. When I can lay my hands on him, he will be able to help to confirm or disprove what should be a general rather than a particular case. So far I have only met with what appears to be one genuine case of violet blindness. It is very remarkable, on account of the eccentricity of the colour nomenclature. The only two colours which the subject saw were red and _black_. He named all greens and blues as black, the distinction between the two being that the former was “bright black” and the latter “dark black.” Yellow he called white, and a glance at Fig. 16 will show that at this place in the spectrum the neutral point of a violet-blind should occur. By shifting the slit gradually into the green, he called it grey, instead of “bright black,” though it did not match the white patch when darkened. He called a green light a “bright black” light. We shall have to refer to this case when we are describing other investigations. CHAPTER VI.