Luminous efficiency in the red-green region of the spectrum

Monochromatic-test stimuli were produced with a double monochromator with a neutral wedge linked to the mechanism for changing wavelength. Only minor adjustments of the width of the entrance slit were required to produce monochromatic stimuli of equal luminance. For any given wavelength, the radiance of the stimulus was proportional to the width of the entrance slit. These test stimuli were adjusted to equal luminance, using various criteria: (1) flicker, (2) direct comparison, (3) distimulus matching, and (4) step-by-step. For two out of three subjects, measurements made by the distimulus method agree with flicker measurements when the primaries are adjusted to be equal by flicker photometry. They provide a test for additivity. Red and green, which are equal in luminance to yellow by flicker photometry, are brighter than yellow by direct comparison and also by the step-by-step method of assessment. The step-by-step data prove that the discrepancy between flicker and direct comparison involves more than confusion between saturation and brightness and is consistent with the notion of summation of responses in the chromatic and achromatic channels, as proposed by Ives and others.