THE METABOLISM OF THE CARBOHYDRATES 661 



combines with glucose to form it is believed to be related to the internal 

 secretion of the pancreas (see page 676). 



The difficulty in explaining why the glucose of the blood does not con- 

 stantly leak into the kidney is, however, the only evidence upon which the 

 hypothesis of a blood-sugar compound rests. No chemical evidence can 

 be offered in support of such a view. On the contrary, all experimental 

 work indicates that the sugar exists in a free state; but unfortunately even 

 this evidence is not convincing. Thus, it has been found that, when speci- 

 mens of perfectly fresh blood are placed in a series of dialyzer sacs sus- 

 pended in isotonic saline solutions, each solution containing a slightly dif- 

 ferent percentage of glucose, diffusion of glucose, in one or other direction, 

 occurs in all of them save one namely, that in which the percentage of 

 glucose in the fluid outside the dialyzer is exactly equal to the total sugar 

 content of the blood. Such a result can be explained only by assuming that 

 all of the sugar in the blood exists in a freely diffusible state. In its general 

 nature this experiment is^analogous to that by which the tension or partial 

 pressure of C0 2 is determined in blood (see page 338). 



It has been assumed by many clinicians that glycosuria may sometimes 

 become developed because the kidney fails to hold back the blood sugar 

 even when the percentage is not above the normal so-called renal dia- 

 betes. For the diagnosis of this condition a comparison must be made be- 

 tween the sugar concentration of the blood and that of the urine. In order 

 to do this at least two samples of blood must be taken, one of them, at the 

 beginning and the other at the end of a period during which urine is being 

 collected. Merely to find that one sample of blood collected before or after 

 or during the period of urine collection contains a normal percentage of 

 sugar, does not necessarily indicate that at some other period while the 

 urine was being produced a temporary hyperglycemia may not have ex- 

 isted. 



