Why do we not perceive photons?

Abstract

VERY few photons need to be captured by the receptors of the human eye to excite a sensation of light. The number required is certainly no greater than 100 and may be much less^1,2. A single rod requires only one photon to signal light³. A question which therefore arises is why do we not see punctate photon absorptions when we look at a very dim patch of light? The retina catches incident photons only sporadically in time, and therefore in space, when a source is very dim. If very few photons are sufficient to detect light, it might be expected that we should see isolated flashes of light, so that a patch of light appears mottled or speckled. Indeed, if a dim object is viewed on the screen of a cascaded image intensifier, so that each quantum caught by the cathode is amplified, we do see quanta as individual flashes or specks on the display phosphor². This apparent contradiction creates a physiological problem ignored in most standard textbooks on vision although the issue has been raised^4,5. We report here a possible solution; we have observed point sources of light set up one at a time very rapidly and at random by computer within a square patch on the face of an oscilloscope and find that as the luminance of the point sources is decreased the number of specks seen increases markedly, and that the appearance of motion, due presumably in the first place to the effect of the chance sequences of specks on motion-sensitive neurones, becomes weaker. We propose that at or near the threshold, the number of specks seen increases to infinity so that the specks crowd together to appear uniform and so the motion ceases.