BLOOD CHEMISTRY 183 



that the hemoconcentration was due to loss of fluid through 

 the capillary endothelium. This view which has recently 

 been revived by Swingle and his collaborators 614 is entirely un- 

 tenable, for such endothelial leakage would necessarily result 

 in an obvious anasarca and edema of the tissues. The reverse 

 of these conditions is actually found. The tissues of animals 

 dying of insufficiency manifest a dehydration but never an 

 abnormal hydration as was claimed by Gradinescu. 



In adrenalectomized cats, Baumann and Kurland 44 found 

 the plasma to constitute 57 per cent of the total blood volume 

 as compared to 65 per cent in normal animals. This decrease 

 in the plasma volume was attended by an increase of 8 to 10 

 per cent in the plasma solids. Similar changes have been 

 found in dogs 268 and other experimental animals. There is 

 thus a decrease in the relative water content of the blood 

 plasma which may be brought about in one of three ways: by 

 a loss of an excessive amount of water through the kidneys, by 

 a shift of the water from the plasma into the blood corpuscles 

 and tissue cells, or by an increase in the content of solids of 

 the blood. From the obvious dehydration and negative water 

 balance which occur in adrenal insufficiency, it may be 

 concluded that the observed hemoconcentration is a result pri- 

 marily of the loss of water from the body. The state of dehy- 

 dration of the tissues speaks against the view that any con- 

 siderable amount of water could have been lost from the 

 blood to the tissues. Nor is there any reason for supposing 

 that the observed anhydremia is only relative and the result 

 of an increase in the absolute solid content of the blood. That 

 there is an actual loss of fluid from the circulation and not 

 merely a transfer from the plasma to the red corpuscles is in- 

 dicated by the increase in the number of red corpuscles which 

 roughly parallels the hematocrite readings. Thus in an experi- 

 ment of Harrop et alii, 2 ™ in which the plasma volume decreased 

 from 72 to 57.5 per cent, the red blood corpuscles increased 

 from 4.2 to 6.1 millions per cu. nun. There is, however, no in- 



