360 PHYSIOLOGICAL REGULATIONS 



relations between exchanges and loads similar to those already 

 indicated, leads to the tentative conclusion that the methods of 

 study and the relations are general ones. Instead, therefore, of 

 regarding the study of water exchanges as an isolated field of 

 physiology, I see in it and each other ''field" a special case of very 

 widespread relations having many features in common. He who 

 learns how to deal with data for heat exchanges discovers forms of 

 relations that indicate directly what data will be suitable for 

 arterial pressures or glucose. Where analogies were sought, com- 

 parisons according to dimensions become feasible. 



What selection may have gone into the inclusions and exclusions 

 required to report these materials? So far as I am aware, the only 

 stipulation was that data exist correlating net rates of exchange of 

 the component with the relative content of the same component. 

 I have not rejected any data that seemed adequate in number to 

 yield a decisive correlation. There are plenty of zero correlations, 

 meaning that the component is not regulated or recovered. It 

 seems probable that data constituting conclusive exceptions (nega- 

 tive correlations of net rate and load) do not exist; yet whether or 

 not they will later be found cannot be predicted. 



Further, I gain the impression that no organism would exist in- 

 definitely in which recoveries did not occur of most of the com- 

 ponents that are regularly represented in it. There are com- 

 ponents for which absence or slowness of recovery is compatible 

 with survival, such as lead in man, uric acid in many kinds of insect 

 pupae, low temperatures in poikilothermic animals, or amputated 

 leg in man. The generalizations made might be alternatively 

 phrased so as to exclude these and similar instances from expecta- 

 tion of recovery, though there is no general rule for knowing what 

 those instances will be before they are tested. 



The general induction is that recovery from deficits occurs by 

 increase of gain over loss, and recovery from excesses by increase 

 of loss over gain. Nothing is said about whether recovery of an- 

 other component regularly occurs or not. If the present investiga- 

 tion were limited to extracting qualitative statements of that sort 

 from biological observations, the results might seem inconse- 

 quential indeed. But since gains and losses are capable of quanti- 

 tative comparisons, it is evident that a pattern has been found by 

 which many components, and perhaps all organisms and their parts, 

 may be studied. 



