An afterimage colour following prior adaptation to visual stimuli is thought to be due to bleaching of photochemical pigments, or to a mechanism of neural adaptation on the retina. It is obviously difficult to measure these opponent colours exactly, however it is also obvious that an accurately measure of afterimage colour is fruitful to know colour vision mechanism. In this paper, afterimage colours were measured by the blank rotation method, a series of very short-term blank of colour stimuli causes afterimage clearly, which contributed to know exact opponent colour. The lines on the chromaticity diagram between a stimulus and an induced afterimage colour were not crossed at optical white point, but at white point of colour constancy. The results reported here shows that some afterimages are affected by perceptual colour constancy, and this finding indicates that some afterimages yield not only by photopigments' bleaching but also by some higher mechanisms. This suggestion is agree with recent papers in which, using neon-color spreading configuration, it has shown that some afterimages are not merely by-products of local afterimages, but involve adaptation at a cortical representation of surface.

Shimojo S, Kamitani Y, Nishida S. (2001). Afterimage of perceptually filled-in surface. Science,Aug 31, 293(5535), 1677-1680.