424 PHYSIOLOGICAL REGULATIONS 



ating it meticulously, as for example SW50iN4P3/Ato. I felt that the 

 distinctions did not at present warrant that ideal exactitude. 



Among the variables little distinction was made between inde- 

 pendent and dependent variables, or between stimuli and responses. 

 Yet in any specific experimental arrangement it was usually possi- 

 ble to designate (provisionally) certain variables as subject to con- 

 ditions that the investigator can dictate more easily. I note further 

 that breathing frequencies and drinking frequencies were as useful 

 in the investigation as heart frequencies, in spite of the fact that 

 the former are more often susceptible of ''voluntary" control in 

 man. States of recovery were emphasized, they being relatively 

 free of the experimenter's influences. Such emphasis represents 

 the fact that I had hypothesized how regulations could be studied ; 

 it guided my search for explanatory relationships. 



Among components, various physical and chemical classifica- 

 tions were used. Among paths, and tissues, various anatomical 

 names were also employed. In so far as physiology is a compari- 

 son of phenomena and magnitudes, it appears destined to have little 

 peculiar and specific terminology of its own for single components, 

 but to reserve its language for relations and combinations. 



I believe the above outline of this inquiry is typical of any com- 

 prehensive research. Some one variable (deviation of content) is 

 to be followed through a variety of physiological facts. Some one 

 (component) is set up as a major division. Some group (the four 

 variables) defines the data to be correlated first. And the process 

 of definition and delimitation is followed down to the last detail. 

 Another arrangement might reveal just as much about regulations, 

 and perhaps something different about them; any other arrange- 

 ment would shuffle the same and similar classes into different rela- 

 tive ranks. All this is done either implicitly or explicitly in any 

 design of experiments. 



§ 153. General features 



Certain modes of progressing, from data to correlative data, 

 having been employed, the sort of scheme represented in this pro- 

 gression now appears ' ' logical ' ' in certain respects. Some of these 

 respects were unforeseen. 



(1) More relations are established by keeping one factor present 

 through many correlations. Thus Aa:B, AaC, Ao^D, etc. But 

 since A is common to all, BocC, Bc^D, C°cD can also be correlated. 



