Single cell spectrally opposed responses: opponent colours or complementary colours?

In the 1950s De Valois and colleagues, followed by other researchers, discovered spectrally opposed single cells in the primate LGN. They called them Red-Green and Yellow-Blue opponent colour cells, interpreting them as the biological implementation of Hering's opponent colours theory. By the 1990s, it became increasingly clear the growing data on such cells did not match Hering's unique hues Red-Green, Yellow-Blue. Yet these cells today remain misleadingly described by opponent-colour or similar various terms, with no agreed best term. This paper reviews much of the data to show spectrally opposed responses in primate retina, LGN, and striate cortex are complementary colours of three types: Blue-Yellow, Red-Cyan, and (less often) Green-Magenta. Such cells may be termed 'complementary colour cells' as a generic category, and further specified by the respective hue pair, e.g. Red-Cyan. (In psychophysics, 'complementary colours' refers to a pair of colour stimuli that admix white.) The difference between opponent colours and complementary colours is more than semantic: it concerns colour constancy, which in theory is aided by complementary colour cells.