I will describe my discovery of the colour centre, area V4, in the primate brain, damage to which in the human brain leads to the syndrome of cerebral achromatopsia, or colourless vision. V4 was full of surprises, which challenged what we thought we knew about colour theory. As we explored it, it became increasingly difficult to deal with that discovery exclusively in terms of the classical colour theories of trichromacy and opponency, neither of which accounted adequately for the perceptual constancy of colour categories.    To explore V4 further we needed a new theoretical and methodological approach, using a more natural experimental paradigm than the traditional reduction screen, in which the colour of a ‘point’, or patch, isolated from its surrounds is manipulated by varying its constituent amounts of long, middle and short wave light. This was especially so since many V4 cells are selective for hues rather than to lights of specific wavelength. There thus appeared to be a major electrophysiological disconnect between the perceived colour and the stimulating wavelengths that was not evident when using old methods which relied on the colour of isolated ‘points’. What was required was a clear electrophysiological demarcation, not made before, between colour and wavelength. The Land Colour Mondrian, a relatively simple abstract multi-coloured composition with no recognisable shapes, provided the methodological platform needed to explore this physiological distinction. The use of Land’s paradigm allowed me to demonstrate that wavelength selective cells are not necessarily colour selective and that hue selective cells respond to a patch of their preferred hue regardless of the precise wavelength- energy composition of the light reflected from it alone, thus demonstrating a physiological, single-cell, counterpart to perceptual colour categorisation constancy. Moreover, I found that wavelength opponent cells can be made to give an “ON” or “OFF“ response to a patch of any hue of the Mondrian display placed in their receptive fields, by simply adjusting the wavelength composition of the light reflected from the patch, without changing its colour category.   More recent use of the Colour Mondrian in perceptual experiments has shown that the colour of the after-image produced by viewing a patch in a Mondrian context is, like the colour of the patch itself, independent of the precise-wavelength energy composition of the light reflected from it (e.g. the after image of a green surface that is reflecting more red than green light is magenta, just like the after-image of a green surface that is reflecting more green than red light). The use of theoretical and experimental approaches that are better adapted to studies of cortical involvement in colour vision, has thus revealed new and unsuspected ways of how the brain constructs colours.   In short, one of the most venerable and foundational fields of natural philosophy and modern science remains a pioneering field, as open-ended and as important as before for our understanding of how we obtain knowledge of our world through colour.