834 THE AMERICAN NATURALIST. (VoL. XXXII. 
series of fibres, or hairs, over which sweeps the whole gamut of 
sound waves, each wave length probably affecting one element 
only as it passes along the keyboard. In the eye, on the con- 
trary, every point in the most sensitive part — that is, every 
cone — appears to possess within itself a mechanism similar 
to that in the whole cochlea, for it responds at the same time 
to almost any wave length from red to violet, or to many com- 
binations of them. 
It is, therefore, to the rods and cones that we must look for 
the solution of our problem. They are very small transparent 
bodies and apparently structureless. It is true that in Sir 
Isaac Newton’s time there were supposed to’ be three sets of 
fibres in the retina, the stimulation of one set initiating 
mainly the sensation of red, another that of green, and the 
third that of blue. By combining these three sensations in 
various ways, it was thought that all the remaining color sensa- 
tions could be produced, just as we can produce all the other 
spectral colors by mixing red, green, and blue light. Subse- 
quently, various imaginary chemical compounds were substi- 
tuted for the fibrils. But no evidence has been produced to 
show that these things exist. The physiologist «makes believe” 
these substances are present in the retina, but he is very careful 
not to say just how they should be distributed in order to pro- 
duce the effects that actually are produced; or why one sub- 
stance responds to one wave length more than to another; or 
how the single or combined responses are transmitted without 
confusion and loss of individuality to the cerebral centres. 
I believe that a reasonable answer can be given to these ques- 
tions if we can only clear our minds of old traditions and look 
at the problem from a new standpoint. 
A comparative study of the visual elements gives us this 
new point of departure, and we venture to offer a theory of 
color vision based on the fact that the rods and cones, or the 
parts corresponding to them in the lower animals, are not 
homogeneous, but fibrillated, and that in a number of inverte- 
brates the fibrils are arranged according to their length in 
accurately graded series, and in such a position that they 
always stand at right angles to the rays of light that fall on 
