Now, before proceeding further in this exposition, I would point out that if memory images are habitually synthesized with sensations to form a given perception, and if perception is a matter of synthesis, then, theoretically, it ought to be possible to dissociate these images. Further, in that case, the perception as such ought to disappear. That this theoretical assumption correctly represents the facts I have been able to demonstrate by the following experiment which I have repeated many times. I should first explain that it has been shown by Janet that by certain technical procedures some hysterics can be distracted in such a way that the experimenter’s voice is not consciously heard by them, but is heard and understood subconsciously. The ordinary procedure is to whisper to the subject while his attention is focused on something else. The whisper undoubtedly acts as a suggestion that the subject will not consciously hear what is whispered. The whispered word-images are accordingly dissociated, but are perceived coconsciously, and whatever coconsciousness exists can be in this way surreptitiously communicated with and responses obtained without the knowledge of the personal consciousness. In this way I have been able to make numerous observations showing the presence of dissociated coconscious complexes which otherwise would not have been suspected. Now the experiment which I am about to cite was made for the purpose of determining whether certain experiences for which the subject had amnesia were coconsciously remembered, but the results obtained, besides giving affirmative evidence on this point, furnished certain instructive facts indicative of the dissociation of secondary images.