Predictions about chromatic receptive fields assuming random cone connections.

We hypothesize that color vision depends on random connections between cones containing different pigments and neurons at higher levels in the macaque visual system. This hypothesis predicts the same types and proportions of chromatic receptive fields reported in the physiological literature at least up through the lateral geniculate nucleus. The results suggest that the specificity of connections demanded by the labelled-line model of color coding are unnecessary to account for current physiological data.