Abstract
1. Protanopes and deuteranopes show one maximum of wavelength discrimination which occurs near their neutral point in the region of 500 mmicro (blue-green for color-normal). The value of the just discriminable wavelength interval Deltalambda is about 1 mmicro at this point and is much like the normal. To either side of this, Deltalambda rises. It increases rapidly on the short-wave side, and slowly on the long-wave side, rising to about 50 mmicro at the two ends of the spectrum. 2. The brightness distribution in the spectrum for dichromats falls only partly outside the range established for color-normals. The protanope curve is narrower than normal, and its maximum lies nearly 15 mmicro to the left of it. The deuteranope curves are about the same width as the normal, and their maxima lie slightly but definitely to the right of it. The main difference between protanope and deuteranope spectrum sensitivity lies on the red side of brightness curves, where the deuteranope is strikingly higher. This difference furnishes the only reliable diagnostic sign which may be applied to an individual dichromat for separating the two types. 3. The average position for the neutral point of twenty-one protanopes is 496.5 mmicro; of twenty-five deuteranopes 504.3 mmicro. The range of variation in the position of neutral point is twice as great for the deuteranope as for the protanope. 4. Dichromatic gauging of the spectrum cannot yield unique mixture values for any wavelength because of the large stretches of poor wavelength discrimination. Data have therefore been secured which locate the spectral ranges that can match specific mixtures of two primaries when brightness differences are eliminated. The form of the data is much the same for a protanope and for a deuteranope; the only difference is in the relative brightness of the primaries. 5. Previously accepted anomalies in the spectral matching of dichromats which have led to the rejection of the third law of color mixture for them, have been eliminated. They are shown to have been due to the non-uniqueness of color matches and the usually disparate brightnesses of the primaries. Color mixture matches for dichromats are valid at all brightnesses.