of Colour-disease. 8 7 



corresponded to a blindness for red. The shortening of the 

 spectrum increased during the state of narcosis, but it must 

 be left an open question whether the narcosis or the condition 

 of the sun was more the cause ; for the spectrum had in fact 

 become somewhat shorter, in consequence of the haze about 

 the sun. 



In order to be able to control the length of the spectrum in 

 each case, it appeared to me better to continue the experiments 

 with the objective spectrum. My arrangement consisted in 

 allowing a solar ray from the mirror of a solar microscope to fall 

 through a slit into a perfectly dark room, where it first fell on 

 a lens with a focal length of 3 feet, and then on a flint-glass 

 prism. The spectrum was received on a paper screen, or, better, 

 on a second, after it had passed through a slit in the first, in 

 order to cut off the dazzling parts, especially the yellow. While 

 I displaced the first screen the person experimented upon always 

 closed his eyes. The first screen was so placed that the larger 

 Fraunhofer's lines were distinctly seen. They were used as a 

 scale, and a place was designated by estimation as -JH, when it 

 was midway between G and H. 



After 5 grains of santonic acid the end of the spectrum in one 

 case (XXXIII.)* was successively 



at JH instead of H (as it was frequently in other cases), 

 subsequently at G instead of ^H, 

 finally at G instead of H. 



Of course, in artificial colour-blindness, I always compared my 

 eye both before and afterwards with that of the person expe- 

 rimented upon. 



Although the spectrum terminated so beautifully, it had the 

 great defect that, from the fluorescence of the paper beyond the 

 violet, there was often a foggy colourless zone with Fraunhofer's or 

 rather Stokes's lines, which often rendered uncertain an accurate 

 determination of the limit. Fluorescence explains why to some 

 during narcosis the violet end did not apDear shortened and black, 

 but whitish (cases XXXIII., XXXV., XXXVI., XXXIX.), since, 

 even within the violet and blue in the spectrum, fluorescent sub- 

 stances (to which paper and the skin belong) effect a lowering 

 of the refrangibility. By means of the portion of diminished 

 refrangibility the violet-blind sees the violet part of the spectrum 

 like a dull zone, just as a healthy person sees the ultra-violet. 



As, further, in all these experiments there was only an altera- 

 tion at the blue end, in opposition to the first experiment t, this 

 defect was peculiarly important. 



* Vii'chow's Archiv, vol. xviii.p. 20. 



t In case 41 also, red-blindness occurred, after santonate of soda. 

 H 2 



