485 



rather improved. The experiments were made by two colour-blind 

 observers, one of whom, however, did not obtain sunlight at the 

 time of observation. The other obtained results, both with cloud- 

 light and sun-light, in the way already described. It appears from 

 these observations 



I. That any two colours of the spectrum, on opposite sides of the 

 line " F," may be combined in such proportions as to form white. 



II. That all the colours on the more refrangible side of F appear 

 to the colour-blind " blue," and all those on the less refrangible side 

 appear to them of another colour, which they generally speak of as 

 "yellow," though the green at E appears to them as good a repre- 

 sentative of that colour as any other part of the spectrum. 



III. That the parts of the spectrum from A to E differ only in 

 intensity, and not in colour ; the light being too faint for good experi- 

 ments between A and D, but not distinguishable in colour from E 

 reduced to the same intensity. The maximum is about f from D 

 towards E. 



IV. Between E and F the colour appears to vary from the pure 

 " yellow " of E to a " neutral tint " near F, which cannot be distin- 

 guished from white when looked at steadily. 



V. At F the blue and the " yellow " element of colour are in equi- 

 librium, and at this part of the spectrum the same blindness of the 

 central spot of the eye is found in the colour-blind that has been 

 already observed in the normal eye, so that the brightness of the 

 spectrum appears decidedly less at F than on either side of that line ; 

 and when a large portion of the retina is illuminated with the light 

 of this part of the spectrum, the limlus luteus appears as a dark 

 spot, moving with the movements of the eye. The observer has not 

 yet been able to distinguish Haidinger's " brushes " while observing 

 polarized light of this colour, in which they are very conspicuous to 

 the author. 



VI. Between F and a point ^ from F towards G, the colour appears 

 to vary from the neutral tint to pure blue, while the brightness in- 

 creases, and reaches a maximum at from F towards G, and then 

 diminishes towards the more refrangible end of the spectrum, the 

 purity of the colour being apparently the same throughout. 



VII. The theory of colour-blind vision being "dichromic" is con- 

 firmed by these experiments, the results of which agree with those 



