o/ Coloi'i- i>n><l<'ri'<l //// X/./// i'/faneous Contrast. 227 



on the left-hand glass. The black wafers being the only spots common 

 to both fields are easily fixed binocularly, and the white wafers, each 

 seen with a different eye against a different colour, appear on either side 

 of the combined black spots. 



Under these circumstances, although the blue and red fields combine 

 more or less to produce a purple sensation, each white spot retains the 

 contrast colour due to that constituent of the coloured background 

 which alone affects the eye in which its image is formed. 



It was only necessary to find some method of substituting for the 

 white spots two small spectra in order to demonstrate the cause of the 

 greenish-blue appearance of the white spot on the red glass, and of the 

 orange hue of the white spot on the blue glass. To do this, I place 

 over each eye-lens one of Thorp's replicas of Rowland's gratings having 

 15,000 lines to the inch. Two slits are held in a frame in front of the 

 aperture by which light is visually admitted when using the stereoscope 

 for opaque photographs. The spectra of the first order of these slits 

 appear in the middle of the two glasses. In order to prevent direct 

 admixture of the colours of each spectrum with those of the opposite 

 backgrounds, two opaque squares of black material are cemented to 

 each of the coloured glasses, so shaped as to appear of the exact size 

 and position of the spectra. On looking through the stereoscope, two 

 spectra are seen side by side on a field, the colour of which continually 

 oscillates from red through purplish-grey to blue. That connected 

 with the red glass shows little or no red, but a splendid green and an 

 equally splendid violet ; while that belonging to the blue glass has the 

 red well developed, the green pale and dingy, and the blue almost 

 absent. The effect of varying the nature of the blue screen is very 

 instructive. With cobalt glass the red is not very bright, owing 

 probably to the transmission of some red rays by the cobalt glass, but 

 the addition- of a film stained with Prussian blue, by which these rays 

 are absorbed, greatly improves the red. On the other hand, a pale 

 yellow film which cuts off the violet causes the violet of the spectrum 

 on the blue ground to stand out brightly, while a purple film brings 

 out the green, which, owing to the green light transmitted by 

 ordinary cobalt glass, is generally a good deal enfeebled. In each 

 case the contrast of the two spectra seen by different eyes is so well 

 marked that the experiment seems likely to be of service in teaching. 

 It should not, however, be forgotten that the conditions are not quite 

 so simple as in the ordinary production of artificial colour blindness, 

 and that the results are also somewhat more complex. 



Bering's contention, that contrast phenomena originate in the eye 

 rather than in the mind, is substantiated, but the complementary 

 colour to red is shown to consist not of one simple colour-sensation 

 but of two at least, namely, green and violet, and in my own case of 

 blue also. Against a magenta background the complementary colour 



s 2 



