COERELATIVES OF WATER CONTENT IN OTHER SPECIES 241 



out for many lesions from changes of blood dilution following the 

 ''water tests" ; but the evidence now indicates that most differences 

 hold only in means, and hence (as usual) repeated and multiple 

 tests are required clearly to demonstrate diseased states. 



In body fluids other than blood, little information is available 

 concerning concentrations, except in urine. It would be possible to 

 go into detail for some dozen constituents of urine as excreted dur- 

 ing ± AW. When urinary rates (and + AW) increase, total sub- 

 stances (Kriiger, '37) and those individually measured all tend to 

 be reciprocally diluted. Or, their rate of excretion changes only 1 to 

 4 fold while the water excreted is modified 25 fold. In steady states 

 of water excretion the reciprocal relation is less dependent on time, 

 and can be described precisely for chloride (Farkas, '32, Kingsley). 



In general, the dilutions undergone by many constituents of the 

 blood and plasma confirm the measurements of volumes in showing 

 that the plasma takes a large share of the body's water increment, 

 a conclusion that may or may not conflict with conceptions based on 

 the ''plausible" assumption that osmotic pressure of fixed con- 

 stituents controls the distribution of added or subtracted water. 

 I imagine that some of the data represented are not reproducible 

 enough to justify the effort expended in correlating them ; but there 

 seems to be almost no way of foretelling which they are. So long 

 as statistically significant numbers of measurements, and their 

 probable errors, were not the fashion in physiology, adequate data 

 were not often provided. Today they serve at least to indicate the 

 sorts of quantitative information that are needed for the descrip- 

 tion in which I am interested. I am not certain that any of the 

 plasma dilutions are significantly different in man from their homo- 

 logs in dog. 



(3) Compositions. Changes of composition of the body as a 

 whole may be measured in metabolic experiments in which water 

 contents are varied while the diet is otherwise constant. The re- 

 producibility of such retentions and depletions, and the role of time 

 in them, may be again questioned, since large numbers have not 

 been done. Among three electrolytes measured by Wiley and 

 Wiley ('33), water privation leads to increased sodium content, 

 either increased or unmodified chloride content, and decreased 

 potassium content. At no time does the deficit of any one (in per 

 cent of content) equal the deficit of water. Each returns to nearly 

 the original value when water balance is restored. The relative 



