i 3 2 HERMANN VON HELMHOLTZ 



other of the two colours flashes out at the edge of the field, 

 while the remaining area (kept as large as possible) exhibits 

 the complementary colour. If a colour mixture is obtained that 

 can be taken as white, white daylight must be admitted from 

 some other part of the room, and allowed to fall on white 

 paper in order that its colour may be compared with that 

 of the mixed light. Helmholtz then found that the mixed 

 colour altered somewhat according to the position of the image 

 on the retina. When he combined red and greenish-blue, so 

 that the common field of illumination appeared as nearly white 

 as possible, with red predominating slightly, the image appeared 

 distinctly green on fixing a point on the paper lying near 

 the bright area; and the same occurred when the eye was 

 brought so near that the area of mixed colour covered a large 

 enough portion of the field of vision for many elements of 

 the retina, in addition to the yellow spot, to receive the 

 image. 



After defining more precisely his use of the different names 

 of colours he succeeded, under the above conditions, in pro- 

 ducing white from a mixture of indigo-blue and yellow, of 

 cyano-blue and golden-yellow, of violet and greenish-yellow, 

 and of greenish-blue and red. Green alone failed to give 

 any simple complementary colour; in order to produce white 

 it had to be mixed with purple, that is with at least two 

 other colours, red and violet. He then examined the sensi- 

 bility of the eye for the individual elements of the violet 

 end of the spectrum, and found that the human eye could 

 detect all the refrangible rays of this region which were able 

 to pass through the prisms, and he accordingly altered the 

 name ' invisible rays ' to that of ' ultra-violet rays '. The ob- 

 jective intensity of these is by no means a vanishing quantity, 

 as is apparent from the fact that while we perceive nothing 

 of the ultra-violet rays of a spectrum thrown on to a sheet 

 of plain white paper, because they are masked by the ordinary 

 diffuse light, the same parts of a spectrum thrown on to paper 

 soaked in quinine solution will, owing to the less refrangible 

 light from the fiuorescing quinine, affect the retina with suffi- 

 cient energy to be visible. The comparison of colour-tones 

 at different points of the ultra-violet spectrum implied the use 

 of equal light-intensities, since the colour-tone of this band 



