VISUAL SENSATION. 1M5 VISUAL SENSATION. to construct a triangle which slioulil represent exact quantities obtaiuoil by experinient. The one here given is that of KUnig; it is llie result of a vast amount of work dune with the aid of the admirable color-mixing apparatus of Hclniholtz. This curve is therefore at onee the expression of a large number of definitely ob- served facts, and ' also, in consequence, ex- tremely serviceable in enabling one to keep all these facts in mind (Ebbinghaus) . The choice of the exact wave-lengtlis whi('li represent the pure red, yellow, green, and blue is based upon ob- servations made on the color-blind, to be men- FlG. 3. THE CONSTITUENTS OF TOE COLOR9 OP THE SPECTRUM. tioncd below. The proportions in which the three physical color constituents must be mixed in order to reproduce all the colors of the spec- trum is shown on a diti'erent plan in Fig. 3. These curves are also the result of actual meas- urement, of the most e.xaet kind; the color-tri- angle, in fact, represents the same measurements in terms of trilinear coordinates. It is unfortunate that nomenclature is in a very backward stage in the suliject of light and color. Both these words are used witirperfect ambiguity to denote now the sensation and now the physical cause of the sensation — viz. certain trains of ether waves, the photogenic radiations. For instance, the fact of color which is exhibited in the color-triangle is that erythrogenic radia- tions mixed with chlorogenie radiations in a cer- tain proportion suffice to produce the yellow color sensation. This has nothing to do with the question whether yellow, in the subjective sense, is a color blend (like blue-green) of green and red — it is evidently not that. So white (the sensa- tion) is not compounded of yellow and blue sen- sations, but it is caused by the working together of the causes (physical and physiological) of the sensations of yellow and of blue. If these tw-o very different things were to be kept distinct, there would be no difficulty in reconciling the different view^s in regard to color ; the facts of three-color mixture would be accepted by the psychologists (as they are not now), and the physicists would not imagine that they constitute a theory of color sensation. Helmholtz called his book Phi/siolof/ical Optics. The psychological facts of color are to-day so much more in evidence that one writing on the subject now would write on psj'chological optics. The color-triangle should therefore be thought of rather as consisting of four contiguous color triangles, or, to make use of a convenient device of the mathematicians, portions of different sheets of four coincident planes, each the locus of the (more or less unsaturated) four possible binary color lilends. If, for instance, a slight- ly bluish green have added to it more and more of a slightly purplish red (a, Fig. 1), there will result four separate series of color-changes: a series of whitish bluish greens {c..d), green- ish yellows (d..h}, yellowish reds (6..c), and reddish blues (c. .u). The characteristics of color-blindness need re- statement in this connection. Obs<'rvatioiis on the color sense of the color-blind by means of colored worsteds serve for the practical purpose of testing railway employees, but they are of no scientific interest. By means of the color-mix- ing apparatus of Helmholtz and KiJnig, it has liecn shown that the cases of congenital color- blindness are either total (black-white vision alone remains) or partial of the type of yellow- blue vision (red-green blindness). So few cases of blue or yellow blindness have been described (none have been accurately examined) that this form of defect must be regarded as insufficiently cstablislied. Either the absence or the rarity of yellow-blue blindness would indicate that blue and yellow were the primitive forms of color- vision, and that red and green, the last to be attained, are the most readily lost. All the color e(piations of the dichromates are accepted by those having normal eyes (tetracliromates), but many distinctions which are made by the lat- ter are not recognized by the former. .The fun- damental colors of the color-triangle (Fig. 2 or Fig. 3) can be chosen (and have been here so chosen) that all the sensations of the red-green blind can be reproduced by mixtures of those elements in the proportions represented by either the pair of curves W and B or the pair'W, and B; in both instances, however, the sensations per- ceived are yellow and blue. Red-green blindness is therefore of two types, according as the dis- tribution throughout the spectrum of the un- differentiated yellow sense follows the curve of red or the curve of green. The sensations of the achromate would be represented on a line at right angles to the plane of the color triangle through the point W. The color scale of the yellow-blue visioned (yellow, whitish yellow, white, bluish white,andblue)]Will be represented on the line YB ( Fig. 2 ) . The degree of saturation of the differ- ent spectral colors will be got by projecting the spectrum upon the yellow-blue line by means of lines drawn through the points R or G accord- ing to the type of blindness. Thus the neutral points of the defectives of the first class are the points in which the line RW cuts the color curve — that of the defectives of the second class the points in which that curve is cut by the line GW. As there is only one particular color-tone, which is a pure green or a pure red with no ad- mixture of blue or yellow, it follows that most of the reds or greens of nature appear to these defectives to be not absolutely colorless, but either blue or yellow (brown)" of " a very un- saturated quality. Hence they learn to u.se the fourfold nomenclature of the normal person with a remarkable degree of good luck; it is no worse for them to say 'scarlet' for the whitish yellow in which they see the soldier's coat than it is for us to coin a fresh term, pink, for w-hat is merely a whitish red. Hence it is not incomprehensible that the existence of such a defect should have remained unknown until about a hundred years ago. and that, even after it was known, a man of striking scientific en- dowment, such as William Pole, should not have
