In contrast to Land's Mondrian experiment which demonstrated lights with identical spectral composition may be perceived as different colors if viewed with different surrounding colors, the asymmetric color matching experiment shows that colors perceived as the same (matching colors) with different backgrounds may have very different spectral compositions.
In an asymmetric color matching experiment, a subject views simultaneously two scenes each presented to only one eye. The background of the scenes are the same surface but illuminated by different light sources, say, the daylight lamp for one eye and a tungsten lamp for the other. Given a test object of some color different from that of the background placed on one scene, the subject is to select an object from many objects of different colors to place in the other scene so that this object matches the color of the test object seen by the previous eye. The spectral distributions of the reflected lights from the testing and matching objects are then compared. Moreover, given the sensitivities of the three types of cones, their responses to the two objects can be obtained. As can be seen from the figure, the responses to the two matching colors are drastically different.