158 
Proceedings of the Royal Society 
appears orange; at 40° to 50° it looks of a somewhat 
meagre yellow, beyond this a pale yellow. As seen at 
the outside of the eye, the orange only appears wdien the 
obliquity reaches 80°, and the yellow at 90°. 
b. Some orange colours show the change very markedly to 
yellow, and to nearly white. 
c. Emerald green. —This, at 40°, becomes nearly white, gene¬ 
rally yellowish. 
d. Ultramarine. —This is very persistent, visible at 40° as a 
blue. 
e. Pink , of a purplish cast.—This in day light, when placed 
on a white ground, appears—even at a very moderate 
obliquity—a purplish blue ; if placed on a black ground, 
it assumes a lavender blue colour. 
This change of purples and pinks to blue is one of the 
most striking; perhaps the best way of witnessing it is to 
use two thicknesses of cobalt blue glass, fortified with a 
pink or purple one, so as to allow both extremities of the 
spectrum to pass freely. This screen, held before a gas 
light, appears by direct vision of a fine pink colour, but 
by a moderate obliquity it is reduced to a bright blue. 
/. A bluish-green glass, held in front of a gas light, appears 
to become blue by oblique vision. 
g. A yellowish-gree?i glass becomes by oblique vision more 
decidedly yellow. 
Remarks and Speculations on the Phenomena. 
Under oblique vision the purples or pinks become blue, and the 
extreme red becomes dull. It would appear, therefore, that towards 
the margins of the retina the sensation of blue is less reduced in in¬ 
tensity than that of red , and a step in the explanation of the results 
is this: the red in the purple or pink becomes a dull orange or 
yellow under oblique vision ; this gives rise to the sensation of white 
light when combined with a part of the blue, and reduces the re¬ 
maining part of the blue to a paler cast. The same explanation 
applies to a blue-green becoming blue—the green becomes white or 
pale yellow under oblique vision, and so dilutes the blue ingredient- 
to a paler shade. 
