434 PHYSIOLOGICAL EEGULATIONS 



obtained to warrant generalization. Sometimes the earliest gen- 

 eral statement proves to be correct, within limits ; often not. Since 

 the limits are not stated, the generalization is not an inductive con- 

 clusion but a hypothesis. Hypotheses in their place are useful; 

 those that are "reasonable" are often less reliable than those that 

 are free of wishful thinking. In biology, however, one need not 

 subscribe to a conclusion without knowing how much is opinion, and 

 upon what and how many observations it is based. 



Premature generalizations might ultimately be divided into 

 those that later prove to be widely true, and those that prove to be 

 true only in a minority of instances. Often arises the human ques- 

 tion, which was the discovery? the first vague guess, or a later 

 specific guess, or one of the several tests of whether the formulation 

 was true or not? The first was one guess that was later found to 

 correspond with the facts under limited circumstances, among other 

 guesses that were later discredited since they were either untest- 

 able, or never tested, or found by test to be untenable. The second 

 was also an inference or jump beyond that which had been demon- 

 strated. Of the tests each was a step in mapmaking. "It is much 

 easier to discover than to see when the cover is off," remarked 

 Thoreau (1863, p. 71). 



If physiology is to be more than a body of qualitative facts about 

 which are built up projected or extrapolated aphorisms, physiol- 

 ogists will engage in searching out, by critical induction, conclu- 

 sions that are epitomes of the observed facts and relations. Data 

 will be at a premium, and perhaps no one will be a compleat physiol- 

 ogist who does not furnish quantitative relations with his observa- 

 tions. "Truths are fecund only if bound together" (Poincare, '13, 

 p. 279). Far-reaching procedures of synthesis, combination, con- 

 solidation, will probably take precedence over the search for key- 

 factors and analogies. 



A different opinion is clearly expressed thus : "Of the two types 

 of formulation consciously employed that which is purely statis- 

 tical is unsatisfying, since the constants it provides are dimension- 

 less and arbitrary ; expressions derived from physical theory have 

 the enormous advantage that the character of their contained con- 

 stant immediately gives basis for concrete experimental tests. 

 Without such tests these equations may also remain arbitrary" 

 (Pincus and Crozier, '29). I am not convinced that in any single 

 instance in biology, nor in any other science, the character of the 



