314 PHYSIOLOGY CHAP. 



while glycogen increases. (c) During hibernation the fat of 

 marmots disappears gradually, while the glycogen, as we have 

 seen, remains almost constant. 



The possibility of the formation of sugar from glycerol was 

 chemically demonstrated by Ernil Fischer. Cremer and A. 

 Liithje further observed that when glycerol was administered to 

 depancreatised dogs, glycosuria increased approximately in pro- 

 portion to the dose. They concluded that a similar conversion 

 took place within the body. 



Other cogent arguments in favour of the theory which derives 

 a part of the sugar normally formed in the body from the 

 decomposition of proteins or of fat can be adduced from the well- 

 known disease of diabetes mellitus, the chief symptom of which is 

 the constant presence of a large amount of sugar in the urine, 

 identical with that formed in the liver from the glycogen (glucose 

 or dextrose}. In its milder forms, the diabetes ceases with the 

 absolute exclusion of carbohydrate from the diet ; in the graver 

 forms, the sugar is decreased, but does not entirely disappear from 

 the urine, even with an exclusively flesh diet. Innumerable 

 investigations have been devoted to the study of this disease, and 

 if all that has been published on the subject were collected it 

 would, as Bunge remarked, furnish a library. 



Our task is to define the fundamental points of this complicated 

 problem. Does diabetes depend on an alteration of the kidneys, 

 by which the epithelia of the urinary canal allow the glucose 

 constantly present in the blood to filter through more readily? 

 No ; for diabetes is constantly associated with hyperglycaemia, i.e. 

 increase of sugar in the blood, rising from 0'05-0'15 per cent to 

 0'22-0'44 per cent. Does it depend on increased hepatic glyco- 

 genesis ? No ; because the liver of persons who have died of diabetes 

 show in not a few cases a definite, sometimes a considerable, amount 

 of glycogen in the liver (Kiilz, v. Mering), although as a rule it 

 contains but little, as shown by Fig. 99, which gives the iodine 

 reaction of the liver cells of a normal person and of one who has 

 died of diabetes. Glycogen has, moreover, been found in the 

 hepatic cells extracted during life in severe diabetes, by puncture 

 of the liver with a trocar (Frerichs). On the other hand, it is 

 interesting to note in diffuse diseases of the liver (hepatic 

 cirrhosis, acute fatty degeneration, phosphorus poisoning) that 

 there is no sugar in the urine, since both glycogen and sugar 

 have entirely disappeared from the liver. In 17 cases of phosphorus 

 poisoning Frerichs found no trace of sugar in the urine ; after 

 administration of 100-200 grms. glucose he only saw it appear in 

 minute amounts in two cases. It is thus impossible to account 

 for diabetes on the assumption that the hepatic cells have become 

 less able to store up sugar. 



All the evidence makes it probable that the hyperglycaemia 



