Two color-separation positive transparencies of a scene, one projected with "red" light and the other with tungsten lamp light, were superimposed on a screen. The light was polarized so that an observer wearing an appropriate viewer could either see both images in each eye or the "red" image in one eye and the "white" image in the other. These two situations gave different results, not the same results, as some previous investigators have claimed. Land's major results cannot be "obtained stereoscopically." We conclude that the process by which color is formed could possibly be a process of the retina or the lateral geniculate body, and does not necessarily have to be a process of the cerebral cortex as implied by the binocular experiments which purported to give the fuller gamut of color.