Adaptation, afterimages and cone saturation.

At low background intensities fovea1 increment thresholds obtained against briefly presented backgrounds are mediated by the initial appearance of the background and increment fields. and at high intensities by a short-term hom~hromati~ afterimage. Masking the afterimages shows that the thresholds mediated by the initial image are a ~ntinuousIy accelerating function of background intensity, whereas afterimage thresholds follow a power function with an exponent of around 0.9. This pattern of results is obtained whether the eye is dark-adapted or light-adapted to a fixed level; however, on a log-log plot. the increment-threshold functions obtained in the light-adapted eye are shifted along a 45 " line with respect to those obtained in the dark-adapted eye. This suggests that adaptation reduces neural response to lights as if multiplying their intensities by a factor between 0 and 1. A final experiment allows rejection of the hypothesis that decision factors (rather than adaptation processes) are responsible for the transient changes in thresholds observed immediately after the onset of intense backgrounds. If the dark-adapted eye is suddenly exposed to a bright environment the discrimination of intensities within the visual field is relatively poor, but within several minutes intensity discrimination greatly improves. A global picture of this adaptation process can be obtained by measuring intensity discrimination at its end points-in other words, by measuring increment thresholds on background fields flashed in the dark-adapted eye and on continuous backgrounds to which the eye has become completely adapted. For fovea1 stimuli the results of such an experiment can be pieced together from previous studies. Classical measurements of intensity discrimination have shown that increment thresholds obtained on continuous backgrounds are approximately described, except at low intensities, by Weber's law (e.g. see Brown and Mueller, 1965). On the other hand, there have been several studies that have measured increment thresholds when the background and increment fields were briefly flashed together in the relatively dark-adapted eye (Bartlett, 1942; Brindley, 1959; Sperling, 196.5), and several more studies that obtained increment thresholds on both steady and pulsed backgrounds, although not at high The increment thresholds obtained in all these studies are also approximately described by Weber's law. In the first experiment described below, we obtain increment thresholds for pulsed and steady backgrounds over a wide range of background intensities , and con&-m the findings of these earlier studies. The consistency and the apparent simplicity of all these results are deceptive, however. The condition that is …