EXPERIMENTAL DIABETES 377 



and hence dextrose, can be derived only from carbohydrates, and 

 he explains the evident sugar formation which occurs in the 

 animal body in diabetes as being derived from the carbohydrate 

 groups in the proteid molecule. He asserts that if we allow 

 10 per cent, of carbohydrate as the average amount in all 

 proteids, then the whole of the sugar which is excreted in 

 diabetes can be accounted for as of proteid origin. We shall 

 presently see, however, that even if we do admit this average 

 it is impossible thus to explain the sugar production in many 

 cases of diabetes ; and, moreover, the premise cannot be granted, 

 for, as we have just seen, none of the proteids which we take 

 as food contains anything like this amount of sugar : it would 

 be more correct, in fact, to state that the average amount is 

 1 per cent. 



Besides the evidence of Liithje and others that Pfliiger's 

 hypothesis cannot possibly hold in the case of depancreated 

 dogs, there is a considerable amount of clinical evidence that in 

 Diabetes mellitus too the hypothesis is untenable. The strongest 

 evidence in this connection has been furnished by F. Kraus 

 and by Mohr. 



F. Kraus ( 34 ) kept a patient for several weeks on a daily diet of 

 milk, eggs, flesh, some white bread, and 20 grm. butter, and found 

 that for every 1 grm. of sugar which was obtainable by hydrolysis 

 from the food, 1-655 grm. appeared in the urine. Mohr ( 35 ) 

 estimated the total sugar excreted by two diabetic patients for 

 four to five weeks, and found that over 1300 grm. carbohydrate 

 appeared in the urine in excess of what could have been derived 

 from the diet, supposing even that all the proteids of this contained 

 10 per cent, of sugar. 



Granted, then, that the sugar in diabetes cannot be derived 

 from carbohydrates, either free or bound to the proteid molecule, 

 we are driven to the conclusion that proteid itself or fat must 

 be its source. 



Let us consider whether fat may not be the source of the 

 sugar. Neutral fat, it will be remembered, is a compound of 

 glycerine and fat acid. We have seen that glycerine feeding 

 probably does not cause glycogen to be deposited in the liver 

 (see p. 331). On the other hand, it has been found by Emil 

 Fischer ( 2 ) that, in the chemical laboratory, glycerine can be 

 readily converted into glycerose by mild oxidation with a mixture 



