EXPERIMENTAL DIABETES 359 



given to produce the maximal effect. The sugar in the phloridzin 

 cannot therefore be of any account in influencing the urinary 

 sugar. The drug should, however, be frequently injected, other- 

 wise the influence of the first dose will have begun to disappear 

 before the second dose has had time to act. The injections 

 should be made three times a day. Observations similar to the 

 above on six rabbits gave an average D : N ratio of 2" 7 : 1. In 

 general the quotient in starved dogs poisoned by phloridzin 

 is about 3'S. 1 



We see that this ratio is very nearly the same as that found 

 by Minkowski in starved depancreated dogs. This fact would 

 seem to indicate firstly, that in phloridzin diabetes, as in pancreatic, 

 the tissue proteid is the source of the sugar, and secondly, that 

 the two forms of diabetes are identical in their causation. Con- 

 cerning the source of the sugar in starved phloridzin-poisoned 

 dogs, however, Pfltiger has pointed out that in most of the 

 hitherto recorded researches, an approximate estimate of the 

 possible glycogen contents of the animal at the time the phloridzin 

 was given shows that in them all, the sugar excreted in the urine 

 as a result of the action of the drug might quite well have 

 been derived from this source, and not at all from the tissue 

 proteids. More recent work by Kraus ( 3 ) has, however, established 

 beyond doubt that the glycogen stored in the tissues cannot be 

 the only source of the sugar in phloridzin diabetes. A number 

 of cats were similarly fed for some time. The glycogen in five 

 of these was estimated by Pfltiger' s method and the following 

 amounts (expressed as grape-sugar) were found per 100 grm. body 

 weight : (1) 0*2637 ; (2) 0-3773 ; (3) 0-2414; (4) 04700 ; (5) 0-1985. 

 The other cats were kept under phloretin (1-2 grm. per kg. body 

 weight) and starved for from five to eight days ; during this time, 

 the total amount of sugar excreted in the urine was carefully 

 estimated, and at the end of the period the cats were killed and 

 the glycogen determined. The total sugar excretions plus the 

 glycogen (expressed as dextrose) remaining in the body per 100 

 grm. body weight (initial weight taken) were : (1) 0-5887 ; (2) 04356 ; 

 (3) 0-3272 (leucin given in food) ; (4) 0'7724 (alanin given in 

 food) ; (5) 1-2282 (alanin given in food). 



1 In dogs starved for three days, von Mering found that 20 grm. phloridzin 

 given by mouth caused, in the first twenty-four hours, a D : N ratio of about 5:1, 

 which afterwards fell to 3 : 1. 



