ACCIDENTS BY KAIL AND SEA. 177 



But if he employ a single name, it might in a consequent manner be, 

 according to the case,purpl€ (improperly called red), green, or gray. Sup- 

 posing that he uses "green," according to the denominating method, his 

 chromatic sense will be judged as imperfect for purple and for gray, but 

 as correct for green,'. 



From our point of view, therefore, Dr. Favre's method does not seem 

 satisfactory. Besides, not appearing to us certain, and not supplying 

 us with any basis for a useful classification, it requires more time than 

 is expedient ; nevertheless, this principle seems usually applied in France 

 and England. 



Dr. Stillings's method is also founded upon the designation of colors, 

 and, if for no other reason, it should, we think, be condemned. This 

 method is based upon the principle of colored shadows. Before a bril- 

 liant light in a dark room a colored glass is held, so that the light, pass- 

 ing through the glass, and in consequence colored by absorption, strikes 

 a white surface, a sheet of paper for instance, at right angles. In the 

 neighborhood of this sheet, and between it and the glass, a slender 

 and opaque object, a pencil, say, is held in such a manner that its 

 shadow distinctly falls upon the paper. This shadow then seems tinted 

 with the complementary color of the glass, that is to say, it shows the 

 different shades of purple or red if the glass is green ; green or blue- 

 green if the glass is red, etc., in accordance with the Young-Helmholtz 

 theory. According to Dr. Stilling, the color-blind will be recognized by 

 the fact that the shadow in question appears to him uncolored, black 

 or gray, while to the normal observer it assumes the contrasted color, and 

 the diagnosis is established according to the riames applied to the colors 

 of the shadows by the color-blind. It must be evident that his method 

 deserves very little confidence, and that it simply depends upon a chance, 

 whether after such a jiroof a normal observer may not be declared 

 color-blind. Besides, as the judgment is based upon the name given to 

 the colored shadow by the subject examined, it may readily happen 

 also that a color-blind person may be declared to have normal sight, if, 

 as is often the case, he guess the true name of the color. Briefly, then, 

 this method is not sure under this form. This judgment is not founded 

 merely upon theoretic reasons, but also upon a large number of direct 

 experiments, and is also confirmed by the examples cited by Dr. Stil- 

 ling himself. 



It does not, however, follow that colored shadows may not be used in 

 the examination of the chromatic sense, if so arranged that the exam- 

 iner can i)erfectly regulate the light according to his pleasure. Accord- 

 ing to my method, with mirrors and two lights, a comparison between 

 two colors may be established. The green-blind here finds, as elsewhere, 

 a resemblance between a certain shade of green and purple, etc. As 

 by this method the intensity of the light may be exactly regulated, the 

 feeble perception may also be relatively determined. The experiment 



we have made about this declares in favor of the Young-Helraholtz 

 VI s 



