If a diabetic patient is kept on a constant diet somewhat above his carbohydrate tolerance, with other conditions of life, particularly exercise, also constant, the amount of sugar excreted nevertheless will vary from day to day. If frequent samples of urine are examined, the excretion of sugar will be found to vary somewhat, independently of the meals. Some meals are followed by an immediate glycosuria, while others are followed by a slight increase or none. Closer investigation reveals these variations as expressions of a regular daily periodicity. Often in diabetic patients there is a minimum excretion of sugar around noon, even with a constant diet. In severe diabetes during starvation, Hatlehol1found a continuous decrease in the sugar content during the day and an increase during the night, often with the greatest excretion in the early morning hours. When first encountered, these phenomena were puzzling and inexplicable. Now light