Numerous studies have reported that suicide rates tend to be affected by temporal variables (month, day of the week, day of the month, holidays). Few authors have provided convincing explanations of these relations, & no one except Durkheim has suspected that they display similarities which indicate that they might have some causal origins in common. This article attempts to look at several temporal correlations as examples of the same underlying dynamics & to offer a theory which helps to organize a traditionally enigmatic body of research. (28 refs)