1. The annual number of notified cases of erysipelas is approximately 17,000, and assuming complete notification of the disease the general case fatality is approximately 6 per cent.

2. The death-rate, deaths in terms of the population, varies according to age, being highest at the beginning and end of life and at a minimum between the age of 5 and 10 years. The mortality of males is identical with that of females up to age 25, but is afterwards in excess.

3. The disease has in recent years a well-marked seasonal incidence—a winter and spring excess with a summer defect. In this respect it resembles scarlet and puerperal fevers and, although its seasonal incidence has changed in the course of time, the alteration has not been nearly so pronounced as that for scarlet fever.

4. The incidence is highly correlated with overcrowded conditions, the correlation coefficient being in some periods as high as + 0·83 in the divisions of London and + 0·70 in the sanitary districts of Glasgow.

5. The morbidity from erysipelas is fairly well correlated in time with that from scarlet fever and erysipelas, but in London, in the urban and in the rural districts of England and Wales, the spacial correlation is very small. On the other hand, in Glasgow the spacial correlation between the incidence of erysipelas and that of scarlet fever is highly negative, −0·718 ±0·109, whereas between erysipelas and puerperal fever the association is positive, + 0·689±0·109 during the period 1903–8, but these values were much reduced when the partial coefficients were calculated.