The observations which follow are drawn from my experience with 800 cases of erysipelas which have been under my care in the erysipelas wards of Bellevue Hospital during seven periods of service from February, 1909, to April, 1913, an aggregate of forty-two weeks. DURATION So many claims have been advanced for the efficacy of various forms of treatment in limiting the duration or extent of this disease, that I have had especially in mind to determine the average febrile period of various forms of erysipelas (drawn from a large series of cases practically untreated), in order that there might be a background on which the effects of treatment might be contrasted. In reckoning the febrile period, I have counted the days from the onset of the disease until the rectal temperature has remained at or below 100. Uncomplicated facial erysipelas cases (500) ran an average duration of 6.77 days.