The apparent colour of a target depends to some extent on the colour of the background. Jennes and Shevell have recently shown that a red background scattered with white or green dots has a different influence on the apparent colour of a target than a uniform background with the same space-averaged chromaticity and luminance (1995 Vision Research 35 797 – 805). This could have important implications for our understanding of colour vision. However, it appears to be a very fragile effect. We were only successful in reproducing it when we used the same long dark-adaptation and pre-adaptation times as those used by Jennes and Shevell. One could argue that with shorter adaptation the effect was masked by gradual changes such as shifts in receptor sensitivity. However, when we interleaved trials with and without white dots for well over 20 min there was still no systematic difference between the two kinds of background. Moreover, when we repeated their experiment with a green rather than a red background, adding white dots had no effect. The long adaptation times that are required to produce the effect and the fact that we only found it for a long-wavelength background suggests that the effect may have to do with differences in the extent of rod adaptation. Although it remains an interesting phenomenon, this would probably make it irrelevant for normal colour vision.
