On the problem of retinal directional sensitivity.

When two physically similar pencils of light enter the eye, one through the pupillary centre and the other near the edge, the former appears brighter and, when monochromatic light is used, generally as thought it were of a shorter wavelength. This is usually attributed to the photoreceptors, notably the cones, being more sensitive to light travelling along them axially than to light with a radial component. However, allowance has also to be made for pre-retinal factors. One of these is the angle of incidence on the cornea; the Fresnel formulae show that, at perimetric angles less than 15 degrees, the ensuing asymmetry in the fractions of light lost by the two pencils needs compensating. Furthermore, a correction is required for differential light absorbance along different paths in the crystalline lens. Finally, the narrowness of the pencils necessitates the application of diffraction theory: the harmonic content of a light beam varies inversely with the distance between its point of pupillary entry and the optic axis, the effect varying across the spectrum. The above corrections suggest that there is little difference between the directional sensitivities of human rod and cone mechanisms; the above chromatic effect can be explained partly on the basis of pre-retinal factors, and earlier theoretical explanations are rendered more plausible.