EXPERIMENTAL DIABETES 365 



of the condition is, however, indicated by the fact that in guinea- 

 pigs very marked vaso-constriction follows suprarenal extract 

 administration, but no glycosuria, and, further, that the painting 

 of the pancreas with many reducing substances is also followed by 

 glycosuria (Herter). 



Animals may after being repeatedly injected with the drug 

 acquire a certain tolerance towards it (Noel Paton 33 ). The 

 ammonia excretion is considerably increased during the glycosuria. 



Before proceeding further, let us briefly consider whether these 

 experimental forms of diabetes bear any analogy to the disease 

 Diabetes mellitus in man. A description of the disease itself 

 would be out of place here, but a consideration of its possible 

 causes, deduced from the foregoing experimental observations, 

 must be of very great interest and importance. 



We have seen that glycosuria itself may be induced by 



(1) An overproduction of dextrose by the liver piqure, &c.; 



(2) A want of destruction of dextrose in the organism 



pancreatic diabetes ; 



(3) A production of sugar from the blood by the renal cells 



phloridzin diabetes. 



To which of these processes is the glycosuria in Diabetes 

 mellitus due ? If we can but answer this question, then some 

 indication may present itself of the cause of the other symptoms. 

 At the very outset we may delete the third possibility, for were 

 D. mellitus analogous with that produced by phloridzin injection 

 which typifies glycosuria due to this cause we should certainly 

 expect to find the blood of a diabetic patient to contain a more 

 or less subnormal percentage of dextrose, which, however, is not 

 the case ; there always being an excess. 1 Moreover, there is no 

 evidence that the kidney tissue is in any way pathological at an 

 early stage in the disease. These arguments do not, of course, 

 preclude the possibility of some of the mildest transitory forms 

 of glycosuria in man being due to renal trouble; the disease 

 D. mellitus is, however, not due to it. 



Either of the first two conditions enumerated above would 

 cause hyperglycsemia, which always seems to exist. This hyper - 

 glyccemia is the immediate cause of glycosuria. 



1 The analytical data with reference to this question are, however, very 

 meagre. 



