The diagnosis of diabetes mellitus, in a frank instance of the disease, is one of the easiest that can be made reliably. Even in patients who are symptomless, diagnosis is possible on the basis of a dependable determination of the level of sugar in blood taken from the patient after twelve or fourteen hours of abstinence from food. With the Folin and Wu technic for blood sugar analysis, the finding of something more than 120 mg. per hundred cubic centimeters of reducing substance expressed as dextrose establishes the fact of diabetes with an error which is smaller, I suspect, than that involved in the diagnosis of any other disease. A degree of overstatement is involved in this assertion and certain reservations will be mentioned later.
On the other hand, to be able to affirm with confidence that glycosuria in a given person is significant of diabetes when the fasting blood