Foveal cone thresholds.

The method of constant stimuli was used to estimate the psychometric functions for detection of one or two flashes when two light pulses were presented. The test stimulus consisted of two simultaneous 0.5 msec, 1' pulses separated by 17'. Observers reported seeing 0, 1 or 2 flashes. A computer-controlled direct-view apparatus allowed sampling of slightly different foveal locations on each trial. The data were analyzed assuming a binomial probability for sampling of L and M cones and Poisson distributed quantal fluctuation. Under these assumptions, the measurements imply that detection requires a minimum of 5-7 quanta absorbed per cone, and that the effective number of cones illuminated by the 1', 0.5 msec pulse is two. The estimated L/M cone ratio was 1.6 for one observer and 4.0 for the other; each observer's ratio was in general agreement with the value estimated independently by heterochromatic flicker photometry.