This paper is a survey of the incidence in Britain during the present century, and prior to the outbreak of the present war, of some of the disorders tentatively allotted to the psychosomatic category in virtue of their compliance with “the psychosomatic formula” (1943). An examination of prevalence indicates that the diseases studied fall into two main groups, viz., those with a falling incidence and those with a rising incidence. The divergence in secular trend suggests a broad difference in etiology. The distinction may be stated as follows: In the diseases that decreased (hypochromic anaemia, chorea and female peptic ulcer) environment in its physical aspects is recognized to be of high etiological relevance, whereas in the diseases that increased, it is environment in its psychological aspects that is known to be causally significant, the influence of external physical factors being wholly indeterminate. The establishment of this distinction, both etiologically and epidemiologically, helps to clarify “the concept of a psychosomatic affection”. A refinement of its definition would be as follows: a psychosomatic affection is a disease which complies with the psychosomatic formula and whose prevalence rises or falls in accordance with the rise or fall of communal “upsetting events”, i.e., in accordance with the pressure of environment in its psychological aspects. Diseases which comply with this definition during the period of historical time under review comprise a high proportion of illnesses attributed to the following labels: peptic ulcer and gastritis; psychoneurosis, including anxiety states and hysteria; exophthalmic goiter, the hypertensive cardiovascular disorders (hidden under such terms as hypertension, coronary thrombosis, angina pectoris, diseases of the myocardium, cerebral vascular lesions, arterio-sclerosis, and nephritis), and diabetes. Analysis of the prevalence of these diseases reveals that their incidence showed an upward trend during the quarter century 1909–13 to 1934–38, a period corresponding to the reign of George V. In anxiety states, exophthalmic goiter, and diabetes the upward trend involved both sexes. The degree of rise was, however, greater in males in exophthalmic goiter, and in females, in diabetes. In peptic ulcer and also perhaps in the hypertensive disorders the upward trend involved males only. In all affections the rate of increase appears to have been most marked in the younger age groups. The findings as regards changes in sex incidence might be restated as follows: The diseases which preponderated in females during the late nineteenth century (viz., peptic ulcer, exophthalmic goiter—also perhaps hysteria and essential hypertension) became manifested increasingly during the twentieth century in males, whereas the diseases which preponderated in males (viz., diabetes, suicide) became manifested increasingly in females. This finding suggests inter alia that the “personality type” of males was becoming more feminine and that of females more masculine. The incidence of the various separate affections showed considerable variation as between different social groups and classes. Statistical studies of the incidence of psychosomatic affections require taking account of two key phenomena, viz., Age Shift and Sex Shift. These imply that there can be no standards of age and sex incidence applicable to all places (i.e., all groups) at all times. A proposition of age-trend inversion is adduced which, although it may or may not be “true”, provides an experimental syntax to be proved or disproved by later research. It may be stated as follows: A rising trend in the incidence rate (but not necessarily the mortality rate) of psychosomatic affections is accompanied by a retrogression of the age-peak, i.e., the age of maximum frequency, towards the younger age groups and, conversely, a falling trend is accompanied by a progression of the age-peak towards the older age groups. A study of the epidemiology of the various separate psychosomatic affections will require a further communication. In the meantime it may be remarked that this recherche du temps perdu suggests that much of the “clinical material” so often subsumed under the heading of general medicine is not an eternal invariable but a function of the historical era and the social set-up in which persons happen to live and physicians to practise.