A prominent aim in the management of diabetes is the production of increased tolerance for carbohydrate (and presumably also for the other foodstuffs) by the patient. The degree of success in this direction has been measured as a rule by observing the effect of definite intakes of starch or sugars on a single symptom, glycosuria. A moment's consideration suffices to show that the results of this method may be determined by no less than two factors: one, a metabolic factor related to the actual capacity to utilize absorbed carbohydrate; the second, an alteration in renal permeability that might occur quite independently of any more general alteration in the capacity to oxidize or transform sugar.
Both of these factors deserve to be taken into consideration. In an elaborate study of blood sugar at the Medical Clinic of the Johns Hopkins Hospital, Hamman and Hirschman7 have thus presented the problem: Sometimes