Stiles and Crawford showed over 75 years ago that the sensitivity of the retina varies with the place of entry of the light into the pupil, and thus with the direction of incidence of light upon the retina. The effect is a property of the cones. Rods have a very limited directional sensitivity with a common centre of symmetry with the photopic SC-function. One reason, amongst others, for the existence of the Stiles-Crawford effect is supposed to be the suppression of stray light in the ocular media. The light from, for example, the sun or the bright sky, will be scattered in the ocular media, and cause a haze over the visual scene, thereby decreasing contrast. Thanks to the Stiles-Crawford effect, this scattered light will not be perceived. In order to judge the weight, with respect to other reasons, of this stray light theory, a quantitative study of the amount of stray light is required. This suggestion is based on indications that the scattering in the ocular media is mainly forward scattering, and that stray light coming from the retina seems not to contribute much either. Indirect evidence in favour of the stray light hypothesis is that in nighttime conditions, the visual surround is dark, disturbing stray light does not exist, and the rods therefore do not need directional sensitivity. This is indeed the case, and therefore the pupil size is fully used for the few incoming light quanta. At mesopic levels, the disturbing effects of a large pupil, such as spherical and chromatic aberrations, will be limited for the cones by their directional sensitivity, in the same way as a small pupil does at high luminance levels. It is therefore proposed to consider the directional sensitivity of cones as a virtual pupil.