Analysts have assumed with Freud that everyone dreams every night whether he remembers it or not. In earlier experiments Dement and Kleitman have demonstrated this to be a fact. Now that the objective method of detecting dreams (through particular EEG patterns along with scanning eye movements) is well established, the author has undertaken an ingeniously simple, year-long experiment which he modestly reports as tending to indicate 'that a certain amount of dreaming each night is a necessity'.

Eight subjects underwent several nights of their normal sleep to establish the frequency and duration of their dreams. Then for several nights each subject was awakened as soon as he began to dream and then allowed to return to sleep, this pattern being repeated as often as necessary. In every case the frequency of dreams increased significantly (in two cases over threefold). Following these nights each subject was then allowed to sleep undisturbed for several nights and it was discovered that the total dream-time during these recovery nights was