CHOOSING PHYSIOLOGICAL VAKIABLES 429 



change have been studied intensively (Henderson, '28). Though 

 often not regarded as such, this study is stated by its author to be 

 description. The special conditions in blood (physical hetero- 

 geneity, static equilibrium, chemical components) have been sup- 

 posed by many to constitute a special case ; however, while peculiar 

 features are to be emphasized in any single investigation, no one 

 need suppose that the methods there employed are by any magic 

 limited to blood. 



The multitudinous measurements of rates of oxygen consump- 

 tion are now commonly recognized as descriptive. In the past 

 certain hypotheses concerning the seats (locations) of regulators 

 were set up. Many are the notions that have been vigorously 

 flouted (Lusk, '33). I gather that these guesses have largely been 

 forgotten now (Benedict, '38), and that the search for uniformities 

 and variabilities and comparisons in the light of hypotheses of rela- 

 tionships, sufficiently satisfies those who give their efforts to this 

 field of quantitative physiology. 



Unforunately for human endeavors, there is no way of proving 

 that one procedure is *' better" than another. No one will know 

 except by arbitrarily weighed results whether quantitative descrip- 

 tion was more economical of or enthusing to investigative effort 

 than the search for causal relations or any other procedure. And, 

 of course, each person will weigh differently the results of any 

 search to which his predilections have brought him. 



It is amusing to contemplate what advance in description of 

 water balance would have materialized if all the effort that has been 

 expended in gathering data concerning water in animals had some- 

 how been coordinated. It is coordinated, but in small bits only. 

 To coordinate it on a large scale would require that numerous 

 investigators agree on the precise procedure by which water rela- 

 tions are to be varied (treating content as an independent variable 

 in experimental studies). It might be supposed that no two per- 

 sons would be likely to have a sufficient intensity of interest in the 

 one physiological system. But it might equally be supposed that 

 the mutual gain in correlative material to be realized would even 

 enhance the interests. It is impossible to foresee what net modi- 

 fication in imaginative or incentive possibilities would ensue. In 

 any such effort a preliminary period of exploratory indirection 

 might prevail ; as soon as rough data are available, a detailed tenta- 

 tive plan matures. Perhaps the freedom of which all investigators 



