Four expts. are designed to test Koffka''s hypothesis that the memory trace for color progressively changes in retention in the direction of a more saturated color. College women are used as subjects. The results of Expt. I 1843 [1946] ANIMAL PHYSIOLOGY[long dash]SENSE ORGANS 17279-17288 are inconclusive, due, it is believed, to the inadequacy of the colored paper used in the recognition test. In the remaining three expts., the color wheel is used to vary the saturation. Three different methods are employed: a modified method of limits, a modified method of constant stimuli, and a comparison method. Different standards of saturation are used, and time intervals varying from 28" to 24 hr. The results, as concerns the theory, are negative. There is some indication of a negative time error which is a deviation in the direction opposite from that which is expected from the theory.