156 Royal Society :-— » 
and richness of the tints, or insensible to the brilliancy of the nume- 
rous shades of red and scarlet. It may be, however, that the per- 
centage is on the increase—certainly we hear of more cases than 
formerly ; but this probably arises from the fact of this, like many 
other subjects, being made more generally matter of conversation. 
In further reference to the question of the superposition of colours 
in the spectrum, or of the intrinsic compositeness of rays of definite 
refrangibilities, I may mention a phenomenon which [ have been 
led to notice in the prosecution of some experiments on the photo- 
graphic impressions of the spectrum on papers variously prepared, 
which appeared to me, when first noticed, quite incompatible with 
the simplicity of those rays at least which occupy the more luminous 
portion of the spectrum, extending between the lines marked D and E 
by Fraunhofer, and clearly to demonstrate the presence of green 
light over nearly the whole of that interval. In these experiments 
the spectrum formed by two Fraunhofer flint prisms, arranged so as 
to increase the dispersion, and adjusted to the position of least de- 
viation for the yellow rays, was concentrated by an achromatic lens, 
and received on the paper placed in its focus, which could be viewed 
from behind. A series of white papers impregnated with washes of 
various colourless or very slightly coloured chemical preparations, 
and dried, were exposed; and the spectrum being received on them, 
and the centre of the extreme red image as viewed through a stand- 
ard glass, adjusted to a fiducial pinhole ; a sensitizing wash of nitrate 
of silver, or any other fitting preparation, was copiously applied to 
the exposed surface while under the action of the light. Now, under 
these circumstances, I uniformly found that whereas the spectrum 
viewed from behind through the paper exhibited all over the space 
in question a dazzling very pale straw-yellow, hardly distinguishable 
from white, yet as the photographic action proceeded, and the 
translucency of the paper began to be somewhat diminished also by 
incipient drying, very nearly the whole of that space became occu- 
pied by a full and undeniable green colour, so as to give the idea of 
a distinctly four-coloured spectrum—red, green, blue, and violet ; 
the yellow being in some instances almost undiscernible, and in 
others limited to a mere narrow transitional interval rather orange 
than yellow. It was at the same time evident that a great extinction 
of light (illumination independent of colour) had also been operated, 
the vivid glare of the part of the spectrum in question being reduced 
to a degree of illumination considerably inferior to the red part, or, 
at all events, not much superior. The change of colour was far 
greater than could be attributed to any effect of contrast, and was 
proved decisively not to be due to that cause by hiding the adjacent 
red and blue when the green remained unaffected in apparent tint. 
When, for the photographic preparations wetted as described, 
ordinary, dry, coloured papers were substituted, the change of colour 
in question was always produced whenever the thickness of the paper 
and its absorptive power were not such as to destroy or very much 
enfeeble the more refrangible light. ‘Taking, as a term of compa- 
rison, a putely white, wove, writing-paper, I found that the substi- 
tution of writing-paper, tinted with the ordinary cobalt blue com- 
