Huppert, Pettenkofer, and Voit advocate the following theory: Sugar, like urea, is a normal product of the decomposition of albuminous bodies. In health the sugar is oxidised; in diabetes less oxygen than normal is absorbed, owing to the destruction of the red blood-corpuscles occasioned by malnutrition; therefore sugar accumulates in the blood. Sugar is formed from the albuminous constituents of the body which undergo rapid chemical change. This fact they regard as proved by the increase in the quantity of urea eliminated. Von Mehring found sugar in the urine of a diabetic patient after a twenty-six hours' fast.

Porter believes that the renal epithelial cells, which he claims are frequently enlarged in diabetic patients, take an active part in the manufacture of glucose because the blood of diabetic patients never contains enough sugar at any one time to account for all which is found in the urine. The epithelial cells are supposed to manufacture the sugar out of carbon dioxide and water. This theory lacks confirmation, however, and it should be observed that a very small amount of sugar, which furnishes a mere trace in the blood at any one time, but which is constantly eliminated from the large quantity of blood continually passing through the kidneys, may amount in the course of twenty-four hours to a number of grammes. In view of the established facts in regard to the glycogenic function of the liver, it seems unnecessary to believe that the renal epithelium exercises any special metabolic power in diabetes; moreover, the kidneys may appear quite normal in severe cases of diabetes.

Brunton reports several cases due to the presence of a tapeworm. He thinks that the increased appetite caused by the presence of the worm may have been instrumental in causing the glycosuria from overeating, but it is possible that the peripheral irritation of sympathetic nerve fibres may have been conveyed to the diabetic centre in the medulla, and thence reflected to the vasomotor system of the liver.

Some recent experiments and clinical observations justify the belief that the skeletal muscles play a more important role in the production of diabetes than has heretofore been supposed.

During their activity they normally consume glycogen in considerable quantity. If they fail to perform this function properly, it accumulates in the system.

Kiilz has shown that muscular activity favours the consumption of sugar in the organism of the diabetic, and that much less sugar is eliminated while such patients are taking vigorous exercise.

It must be admitted that there are many hepatic diseases and lesions in which a large part of the secreting surface of the organ is destroyed, and in which glycosuria may never be present, but in these conditions it is possible that while a part of the liver is totally destroyed there may be some remaining cells which are still endowed with normal functional activity, whereas in the disease under discussion it is probable that none of the parenchyma of the liver maintains its normal control over metabolic processes, and hence, whatever sugar is brought to the organ by the portal vein passes into the general circulation unaltered. In the graver forms of diabetes, in addition to the functional disturbances of the liver, there is believed to be present also a condition of malnutrition in which sugar either fails to be consumed or, as suggested by Yeo, "we may suppose that in these cases a morbid ferment is formed in the system, possibly in connection with some radical fault of stomach or intestinal digestion, and that this determines the rapid reconversion of glycogen into sugar".