Prof. W. B. Rogers on Binocular Vision. 389 
1€ conditions of the experiments are obviously such as to 
exclude all opportunity of a shifting of the image on the retina, 
Such a shifting however is essential to that successive combina- 
tion of pairs of points in the two images which on the theory of 
Brewster is required for the production of the single perspective 
resultant. Hence according to this theory the resultant spectrum 
in these experiments, instead of being a single line in a ae 
tive position, ought to present the form of two lines inclined or 
Rrorsing, situated in the plane of the wall without projection or 
relief. 
In reference to the first two experiments it might perhaps be 
maintained that as the perspectiveness of the original line or 
resultant on which the eyes were converged formed part of the 
lirect perception in the first stage of the experiment, it would 
be likely through association to be included also in the spectral 
or subjective perception. But this consideration, which at best 
does not impress me as of much weight, is ansitly inapplicable 
to the conditions of the last experiment, where the eyes are in 
the first place impressed in succession with their respective ima- 
ges, and where yet when they are together directed to the wall, 
the perspective single resultant at once springs into view. pea? 
4, Without resorting to these troublesome efforts of subjective 
vision the following is a simple proof that pictures swccessively 
impressed on the respective eyes are sufficient for the stereoscopic 
effect. Let a screen be made to vibrate or revolve somewhat 
rapidly between the eyes and the twin pictures of a stereoscope, 
over each, completely excluding 
a8 apparent in these conditions as when the moving screen is 
withdrawn. 
one eye, and the retained impression in 
Jements of the perspective resultant which 
5. The i i xperiments described by Prof. Dove many 
years v8 pene ea a stereoscopic effect was obtained by the 
