Numerous recent electroencephalographic studies have revealed the cyclical alternation of two distinct types of sleep (Dement and Kleitman, 1957a), which have been termed “fore-brain” and “hind-brain” sleep by Jouvet (1962). The former is accompanied by slow waves and spindles in the electroencephalogram (EEG) and the latter by a low-voltage, fairly fast EEG pattern; spasmodic, conjugate rapid eye movements (REMs); relaxation of neck and throat muscles in cat (Jouvet, 1962) and human (Berger, 1961); and recall of dreams following awakening (Dement and Kleitman, 1957b).

Insomnia is widely accepted as a leading feature of that illness designated by the number 301.1 in the International Classification of Diseases. The name given varies—“endogenous depression”, “manic-depressive psychosis/depressive type”, “melancholia” or “depressive illness”. The present generation of laymen (not to say some psychiatrists) use the word “depression” so freely in order to describe an unpleasant, unhappy mood, whether lasting or transient, that one may prefer the term melancholia, or at least “depressive illness”, to signify that condition which justifies the number 301.1. The adjective “endogenous” implies for many that environmental stress plays little or no part in the onset of the illness; we cannot accept that this is always so. The term “reactive” is used in several ways, generally to indicate that the state of unhappiness arises out of some external circumstance and that it would end when circumstances improved or the individual accepted the realities of practical life. The patients used in the present study were suffering from an illness of a kind which we believe may sometimes develop in the absence of severe environmental stress, while in others it may be clearly provoked by circumstances, but the illness, as it develops, may take on a form which becomes largely independent of the environmental circumstances and may continue even when the provoking factors are past; it has become an autonomous melancholia. It will be apparent that by autonomous melancholia we mean an illness shown by clinical experience to respond especially well to electroplexy.

In the course of the past few years, a series of related studies (Aserinsky and Kleitman, 1955; Goodenough et al., 1959; Wolpert and Trosman, 1958) has demonstrated beyond doubt the association of normal dreaming with the appearance of rapid, binocularly synchronous eye-movements. It has also been claimed that the rapid eye-movements (REMs) represent scanning movements made by the dreamer as he “watches” the visual events of the dream (Dement and Kleitman, 1957a; Dement and Wolpert, 1958). The REMs are absent during dreaming among those with life-long blindness, but are retained for some years by those whose blindness arises later than childhood (Berger et al., 1962a). In a study of undisturbed nocturnal sleep by Dement and Kleitman (1957b) periods of eye-movements were observed to occur fairly regularly at about 90-minute intervals throughout the night in association with the lightest phases of cyclic variation in depth of sleep, as indicated by the electroencephalogram (EEG). These REM periods had a mean duration of about 20 minutes, and 4–6 occurred per night.