Four cone pigments in women heterozygous for color deficiency.

It is generally accepted that human color matching is trichromatic (requires three primaries) because there are three types of cone, each containing one visual pigment. A color match between two stimulus fields is thought to be determined at the receptor level where the rate of quantum absorption by each of the three cone pigments must be equated to obtain a color match. Since there are three cone pigments, a set of three primaries is required to match any given light. A fundamental law of color matching, which follows directly from this view, is the law of additivity (known also as Grassmann's third law). Additivity means that, if the same light is added to two stimulus fields that match in color appearance, the resulting mixtures will match in color appearance. The color match is not upset because the addition of the same light to both fields does not upset the equality of quantum absorptions for each of the visual pigments. 1 The additivity law holds quite well for small visual fields presented to the rod-free foveal area, but it typically fails in the parafovea, in which rods as well as cones may see the stimuli. Recent evidence suggests that the breakdown is due to the involvement of signals from rods (as well as from the usual three types of cone) in the match. 3 In support of this view we find that additivity holds in the parafovea of most observers if rods are prevented from seeing the stimuli. However, the law does not hold for some women who are heterozygous for red-green color deficiency. Although their color vision is relatively normal, these women carry abnormal, as well as normal, genes for color vision. It is now well established that there are two loci on the X chromosome involved in red-green color vision. Each locus is thought to determine the nature of the visual pigment contained in one of the two red-green sensitive cones. Since women have two X chromosomes, they may be heterozygous and carry two different alleles at either of these loci. Recent developments in genetics suggest that when this happens, each cell is influenced predominantly by one allele, with the predominant allele varying from cell to cell, resulting in the production of more than three types of cone photoreceptor cells with different visual pigments. The failure of additivity for some heterozygous women means that, for them, subjectively similar colors may differ in their effects on the cones, and this in turn means that their eyes must contain more than three types of cone. However, because these women accept trichromatic matches, the cones must feed signals into only three neural channels. In our experiments we asked observers to make Rayleigh matches 6 between two halves of a large annular field presented to the parafovea. A uniform mixture of red (660-nm) and green (546-nm) lights in the upper half of the field was matched to a yellow light (588 nm) in the lower half. The wavelengths were chosen to avoid any significant stimulation of the short-wavelength-sensitive cones. Therefore only two primaries were required to match the 588-nm light. We used the large annular test configuration to obtain matches that were as precise as possible. 7 In order to prevent the rods from seeing the stimuli, matches were made during the cone plateau period after exposure to a bright bleaching light. During this period cones have nearly completely recovered from the bleach, but rods are still too insensitive to detect the test stimuli.8 Observers made matches under three experimental conditions: (1) with no background light present, (2) with a concentric blue background light (455 nm) superimposed over both halves of the test field, and (3) with a red background light (670 nm) superimposed over the test field.9 Under each experimental condition the experimenter determined the set of 660-546nm-mixture ratios that was judged to be a perfect color match to the 588-nm light.10 The additivity law predicts that the 660-546-nm mixture that matches the 588-nm light should not be altered by superposition of the backgrounds. If some of the settings that were acceptable as a match with no background present were also acceptable with a background present, we concluded that additivity held for that observer within the limits of his color discrimination. If none of the mixtures that were acceptable with no background present was accepted on the background, we concluded that additivity failed for that observer. Nineteen men and twenty-one women served as observers. The Rayleigh matches of all 40 in the no-background condition fell within the normal range, but 19 of the 21 women were recruited because they reported some incidence of red-green color deficiency in their families, suggesting that they might