Sir J. F. W. Herschel on Colour-Blindness. 157 
monly met with, sufficed to give a very great extension of the green, 
almost to the extinction of the yellow, while, when the papers used 
were pale-yellow or clay-coloured, answering to the tints called 
“buff” or “ maize” (nearly approximating to Chevreul’s orangé 
4 and 3), and which might naturally have been expected to transmit 
yellow rays more abundautly at all events than the blue, the spectra 
(viewed at the back of the papers) were particularly full and abun- 
dant in green, occupying the whole of the debateable ground. In 
the case of the former, a narrow yellow space was seen, and the blue 
was very much enfeebled, and separated from the green by a very 
perceptible suddenness of transition. With the latter the green was 
finely exhibited, and the yellow confined to a narrow orange-yellow 
border: the blue and violet much enfeebled. 
On further considering these facts, there seemed to be hut three 
ways of accounting for them :—Ist, by the effect of contrast. This 
I consider to be disposed of by the suppression of the adjacent 
colours, as recorded above. 2ndly, by extinction of a yellow element 
of colour over the space DE, allowing a substratum of green to sur- 
vive ; or, which comes to the same, by the extinction of the red ele- 
ment over the same space, which, by its combination with (an as- 
sumed elementary) green, produced the original brilliant straw-yellow. 
And 3rdly, by admitting as a principle, that our judgment of colours 
absolutely, zz se, and independent of contrast, is influenced by the 
intensity of the light by which they affect the eye, and that very 
vivid illumination enfeebles or even destroys the perception of 
colour. As the apparent change of colour from pale-yellow to green 
in the cases above related was always accompanied with a great 
diminution of general intensity, it occurred to me to produce such 
diminution by optical means, which should operate equally on all the 
coloured rays, and diminish all their intensities in the same ratio. 
This was accomplished by viewing the spectrum (as projected on 
purely white paper) by reflexion on black glass, or by two successive 
reflexions in different planes, and I found the very same effect to 
take place. That portion DE of the spectrum which in the unre- 
flected state appeared dazzlingly bright and nearly colourless, was 
seen by one such reflexion, and still more so by two, green. The 
extension of the green region was greater, and the limitation of the 
yellow portion more complete, according to the amount of illumina- 
tion destroyed by varying the angles of incidence on the glasses. 
When much enfeebled by two cross reflexions, the aspect of the 
spectrum was that represented in Chevreul’s coloured picture of it 
from the line A to H. When enfeebled by other means, as by view- 
ing the spectrum thrown on a blackened surface, the effect was 
exactly the same. 
The last of our three alternatives, then, would appear to be esta- 
blished as the true explanation ; and in respect of the second, it is 
eliminated by the consideration that neither the slight degree of 
coloration in the bluish papers, or the tint of the pale-yellow ones 
which effected the change, would give rise to so great a preferential 
extinction of yellow or red rays as an explanation founded on that 
alternative would require, ‘The phenomenon is certainly a very 
