No. 383.] BASIS FOR A THEORY OF COLOR VISION. 851 
measurements, much more accurate and detailed than any 
heretofore made, of the visual elements in all classes of animals 
(also in different parts of the human retina), accompanied by an 
experimental study of their reactions toward different colors. 
(9) Length of the Retinidial Fibrils.—The variation in 
length of organic fibrils capable of responding to ether waves 
is rather narrow, since there is but slight variation throughout 
the animal kingdom in the radius of visual rods. The length 
of the cross retinidial fibrils, using the radius of the rods as a 
guide, appears to vary roughly from less than 1 to about 4u 
in length (Bdellostoma). The length of the rods may vary 
considerably more than this, but, as already stated, this may be 
taken to indicate the varying number of superimposed fibrils 
they contain. 
The length of the retinidial fibrils in any individual must be 
determined primarily by the physical.and chemical properties 
of the available protoplasm without reference to any possible 
advantage to be derived from a response to ether waves of a 
particular length. For surely there is no obvious reason why 
animals should not utilize for visual purposes ether waves several 
octaves higher or lower than they do now, if only the necessary 
end apparatus could be produced. But since the form of the 
vertebral column, or of a mountain range, is the resultant of its 
Own composition and of the forces acting on it, and since the 
forms actually produced are the only ones possible under those 
conditions, so the form and position of the retinidial fibrils must 
be fixed by the inherent properties of the fibrils themselves, 
modified by the.ether waves acting on them. We may infer, 
therefore, that, on the whole, animals respond to all those ether 
waves which are capable of modeling their available protoplasm 
into resonant parts. 
It might be fairly demanded that since the extreme red waves 
visible to us are about twice as long as the extreme violet ones, 
the longest retinidial fibrils should be approximately twice as 
long as the shortest fibrils. But, judging from the very un- 
Satisfactory data at hand, they appear to be about four times 
as long. For example, the longest retinidial fibrils, situated 
at the base of the human cones, should be about .0025 mm. 
