848 THE AMERICAN NATURALIST. [Vow XXXII. 
medium length fibrils should appear more luminous than those 
on either side of them. For this reason the middle part of the 
spectrum appears the most luminous, shading off on either 
side towards the absolute blackness of the ultra-red and 
ultra-violet. 
(5) If we passed through the cone a ray of unpolarized white 
light, it should appear brighter than a polarized one, because it 
would stimulate all the fibrils in the same transverse plane. 
But in other respects the fibrils at different levels would present 
the same color effects as with polarized light. 
(6) The sensation of whiteness is apparently the result of 
stimulating all the fibrils in one or more sectors of the cone to 
a nearly equal degree. Obviously, this may be done by passing 
through the cone all wave lengths from red to violet, or by 
selecting any two or more wave lengths at such a distance 
apart in the spectrum that through summation effects all the 
fibrils will be stimulated to a nearly equal degree. But if, for 
example, red and green light is selected, the place of maximum 
stimulation, owing to the summation of the two effects, will be 
at a point midway between the red and green fibrils, that is, at 
the level of the yellow fibres ; hence the sensation of yellow 
will predominate, but it will be less saturated than the pure 
spectral yellow. 
(7) The colors visible to a given animal should depend 
primarily on the various diameters of the visual elements, and 
the range of colors visible should depend on the difference 
between the maximum and minimum diameters. Hence any 
variation in the form or dimensions of the visual elements 
should be accompanied by corresponding variations in color 
vision. For example, increasing the length of the cones should 
increasé the total number of fibrils of all lengths in them, and 
hence should be accompanied by increased powers of discrimi- 
nation in all parts of the visible spectrum. We may therefore 
attribute the increased sensitiveness of the retina at the fovea 
to the greater length of the cones there, for each cone is thus 
provided with a greater number of fibrils to respond to any 
wave length within the range of vision. 
Again, increasing the diameter of the base of the cones 
