PATHOLOGICAL METABOLISM OF DIABETES 271 



in perfused livers. The conversion has been shown to be complete in the 

 case of several carbohydrates which suggests that it would be possible to 

 demonstrate the quantitative conversion of any of these carbohydrates that 

 are capable of complete utilization in the normal body. The tetrose sugars, 

 although utilizable, have been little used. The pentoses are not utilizable 

 in health except in traces and are not glucose formers in diabetes. Gly- 

 cogen yields in the end only glucose when hydrqlyzed in the body or out of 

 it. It is not only true that all utilizable carbohydrates are capable of con- 

 version into glucose in the diabetic organism, but it is highly probable that 

 in health they are so converted prior to oxidation, reduction or storage as 

 glycogen (or analogues), that conversion into glucose is the normal process 

 of assimilation and a necessary prelude to storage or oxidation. This is 

 sometimes questioned but the support for the view is extensive. Glucose is 

 the natural sugar of the blood. It is the one simple sugar that can be 

 obtained by hydrolyzing glycogen. Efforts to produce a second kind of 

 glycogen in the liver by feeding or by perfusing livers with levulose or 

 other sugars (Cremer) have not been successful. There is only one kind 

 of glycogen known and this is built up out of glucose. Glucose, as a rule, 

 may be injected directly into a peripheral vein at the rate of 0.8 gm. per 

 kg. hour in animals and man without abnormal glycosuria (Sansum & 

 Woodyatt). For a 50 kg. man this is at a rate sufficient to supply 96 

 calories per kg. per day. On the other hand, lactose when injected by vein 

 at the same rate reappears quantitatively in the urine. Levulose, reputed 

 to be more easily utilizable than glucose in diabetes, produces definite levu- 

 losuria when injected at the rate of 0.1 to 0.15 gm. per kg. hour and possi- 

 bly less. Even the triose glyceric aldehyde, which oxidizes so easily in the 

 laboratory that it reduces Fehling's solution rapidly at room temperature, 

 produces triosuria when injected intravenously at the low rate of 0.1 gm. 

 per kg. hour. Galactose behaves in the same way. Yet all of these sugars 

 are completely utilizable in large doses when given by mouth, in which 

 case, of course, they pass to the liver first. They are all glycogen formers 

 when perfused through livers and to form glycogen requires glucose. The 

 liver is the prominent glycogen forming organ of the body and the promi- 

 nent site of assimilation in general. All of these facts and others point 

 strongly in the one direction that the body burns (or otherwise utilizes) 

 glucose and cannot burn directly any considerable quantity of any other 

 carbohydrate. In diabetes the power of assimilation is intact, but the 

 attack on the glucose molecule fails. Hence glucose appears in the urine. 

 Protein. The feeding of protein to a case of severe diabetes with gly- 

 cosuria increases the glycosuria. The same holds in completely phlor- 

 hizinized dogs. As the experiments of Lusk have shown both in phlor- 

 hizinized dogs and diabetes mellitus 100 gm. of protein fed may lead to 

 the appearance of 3.65 grams of extra glucose for every gram of extra 



