Pityriasis rosea.

T HE Medical Research Council's investigation into pityriasis rosea in 1959 has already been published (Proc. roy. Soc. Med., 55, 556). This second investigation, based on a very simple questionnaire, has produced some additional information, but has not altogether solved the question put, namely "What is the cause of this disease?" Much interest was shown. In 1964, 180 observers notified 531 cases, 260 male, 271 female; in 1965, 161 observers notified 514 cases, 243 male, 271 female10 observers reporting in both years. Distribution is probably world wide. Reports have come in from Australia (Campsie, N.S.W.), from New Zealand (Wellington), from Rhodesia (Bulawayo), from Scotland, Wales and Northern Ireland, and from every English county except Bedfordshire, Cornwall, Dorset, Monmouthshire, Northamptonshire, Rutland and Worcestershire. Seasonal notification (figure 1) follows closely the pattern apparently shown in 1959-a high incidence in January, February and March, a steady decline to the summer months of July, August and September, a sharp rise in October and November, and a considerable fall in December. If notifications from the northern and southern hemispheres are separated, there is some suggestion that the seasonal incidence is reversed; but the numbers from the southern hemisphere, 23 in 1964 and 12 in 1965, are very small. As to the interval between onset and diagnosis, the resemblance between 1964 and 1965 is quite remarkable. A fair proportion of cases are reported in the first three days; then comes a very high peak 'one week ago', followed by similar but lesser peaks at 'two weeks' and 'three weeks'. There is even, in both years, an appreciable peak at 'ten days'. In 73 notifications, seven per cent of the total of 1,045, information was too vague to be of use. It is not possible, therefore, to make a firm estimate of the interval between