In Chapter 1 I proposed that pavor noctumus be viewed as a physiological defense against stress resulting from a pre­ ceding nightmare. I demonstrated the compatibility of the alternating phases of inhibition and excitation observed therein with Selye's conception of countershock and sug­ gested that anxiety is associated at first with an inhibition of physiological activity serving a signal function. In Chapter 3 I argued that the repetitive character of such dreams dem­ onstrates a failure of mastery, which is to say, a failure in the psychological mechanism we describe in the concept signal anxiety. This failure was conceived in developmental terms as a failure to attribute meaning to one's own states of tension, and was viewed as operative in the repetitive phenomena of the transference neurosis as well as in pavor nocturnus. These repetitive phenomena were elaborated in

Chapter 4 in terms of the meaningless renunciations of subjectivity represented by reparative mastery of inade­ quately developed ego functions and were laid to experi­ ences of traumatization in the symbiotic phase of develop­ ment. I suggested that depression invariably followed the interpretation of reparative mastery in treatment.