Apeil 15, 1898.] 



SCIENCE. 



519 



In the construction of the curves by 

 Koenig and Diderici it should be observed 

 that the areas are made the same. Eed, 

 having the greatest extent of different wave- 

 lengths included in the sensations, has the 

 least height. Violet, having the least ex- 

 tent, has the greatest height. The corre- 

 sponding ordinates, therefore, must not be 

 confounded with the ordinates representing 

 relative brightness. Of all the spectrum 

 colors violet has the least brightness. The 

 proper distribution of ratios for brightness 

 is shown in Fig. 3, which is due to Captain 



a jb: 



Abney, and made from very careful obser- 

 vations with his ' color-patch ' apparatus 

 after the publication of the work of Koenig 

 and Diderici. They agree quite well with 

 Rood's results obtained twenty j'ears ago, 

 so far as brightness is concerned (Modern 

 Chromatics, p. 34). These curves addition- 

 ally show maxima quite different from the 

 ones just discussed. Abney 's red maximum 

 is between the C and D lines, corresponding 

 nearlj' to scarlet ; his green maximum in 

 the yellowish green, and his violet maximum 

 in the blue ; all of which seems much more 

 consistent than the result obtained by 

 Koenig and Diderici. 



In the face of these diversities between 

 the results of highly skilled observers, all 

 of whom have assumed the truth of the 

 Young and Helmholtz hypothesis, it may, 

 perhaps, be asked whether physical investi- 

 gation of this difficult subject has settled us 

 upon much firmer ground than that occu- 



pied by the opponents of this hypothesis. 

 The history of the idea of primary colors 

 shows that scientific precision has not been 

 its chief characteristic. We are perfectly sure 

 that there are some hundreds or thousands 

 of different hues represented by diiferent 

 wave-lengths which produce effects upon 

 the retina, no one of which has any better 

 claim than any other to be considered 

 primary. As a matter of convenience there 

 are certainly great advantages to be de- 

 rived from suitable grouping of these wave- 

 lengths, but it may be well questioned 

 whether there is any physiological or 

 psychological basis for such grouping. If it 

 is only a matter of convenience has there not 

 been an enormous amount of labor expended 

 in the attempt to find a foundation that is 

 only imaginary? The phenomena of color- 

 blindness are the ones of most importance 

 in this connection. Red blindness is the 

 most common, green blindness almost 

 equally so, and violet blindness so rare that 

 it is in practice hardly taken into account. 

 A glance at Abney's curves shows that this 

 is what ought to be expected. But defi- 

 ciency of color sense for red is often accom- 

 panied with less marked deficiency for yel- 

 low, green and blue, and there seems no 

 good reason for considering any one of 

 these deficiencies more fundamental than 

 any other. We may continue to use the 

 color-sensation curves, and find that, in- 

 stead of indicating uncertainty about the 

 true hues of primai'y colors, they merely 

 show natural diversity among the different 

 individuals subjected to examination. We 

 may still use the Young and Helmholtz 

 hypothesis as the simplest and most con- 

 venient representation of color phenomena, 

 but with large reservation and prudent 

 silence about primary colors. These we 

 may rightly call prominent colors, while 

 we profess our total ignorance about the 

 way in which they aifect the retina or the 

 brain. This condition of somewhat discon- 



