The well-known fall in sensitivity of the eye when exposed to light and the subsequent rise in the dark has long been attributed to the bleaching and subsequent regeneration of rhodopsin and of cone visual pigments. This conjecture, held for thirty years without evidence, was shown to be on the right lines when Dowling (1960) demonstrated in rats that, after full bleaching, there was at each stage of regeneration a linear relation between the amount of rhodopsin in the retina and the log e.r.g. threshold at that moment. In man a quite similar relation was found between the log visual threshold measured in the usual way and the rhodopsin concentration measured directly upon the same area of retina by the technique of retinal densitometry. Though this result could be inferred with some confidence from work on normal eyes (Rushton, 1961a) it was displayed more clearly in the eye of a special subject (Rushton, 1961 b) who lacked nearly all cone vision but whose rhodopsin and rod vision appeared quite normal. If the threshold is expressed in units of the fully dark-adapted value, the relation found was that log threshold is proportional to the amount of rhodopsin bleached. But, as was pointed out in the brief Discussion of the previous paper (Rushton, 1961 b), there are two respects in which this simple relation fails.(a) If the brief adapting light bleaches only a small fraction of rhodopsin, visual sensitivity recovers rapidly, and all agree that the process is in some way different from that linked to the slow regeneration of rhodopsin after substantial bleaching. In the present paper, as in the former one, all bleaches are substantial; thus the rapid recovery phenomenon lies outside the conditions here examined.(b) The simple relation found that log threshold was proportional to the amount of free opsin could not be quite right. To be sure, something changes uniquely with the rhodopsin concentration and may affect the threshold in the way described, but it is not the threshold itself that stands in unique relation to pigment concentration. For several workers have shown that threshold depends also upon the kind of test flash used to measure it. In DARK ADAPTATION IN A ROD MONOCHROMAT 613 his last paper Lythgoe (1940) Crawford (1947) and Arden & Weale (1954). Wolf & Zigler (1950) placed an acuity grating in the test flash and found that the shape of the resulting dark adaptation curve depended upon whether the threshold task was to see the flash or to resolve the grating. These results were confirmed and extended by Brown, Graham, Leibowitz & Ranken (1953). There would of course be little difficulty in reconiciling these observations with the unique dependence of dark adaptation upon the regeneration of rhodopsin, if a change in the nature of the test flash did nothing more than shift the log threshold curve bodily up and down. But all investigations have shown that not only the position but the shape of the dark adaptation curve alters and that the range of the log threshold change for a given bleach depends u...