CHOOSING PHYSIOLOGICAL. VAKIABLES 427 



distinguishing among increments supposed to be similar, ascer- 

 taining uniformities and diversities of relations, demonstrating 

 which forms of relations combine frequently, schematizing relations 

 into comprehensive concepts, and avoiding the conclusion that 

 correlations known are more crucial to the organism than those 

 unknown. 



§ 154. DESGRiPTioisr 



The general investigation here pursued serves to illustrate the 

 fact that concrete results are gained by this particular sort of study 

 of quantitative physiological variables. It may be recalled, paren- 

 thetically, that none of the procedures are novel. What do I see 

 in the use of the procedures mentioned? Is this book an exercise 

 in arithmetic, or is it a part of physiology to which no other 

 approach is known? 



By one definition, all biology is the representation of whatever 

 orders and relations are to be found in vital phenomena. Observa- 

 tions and measurements require coordination, for facts are prob- 

 ably meaningless to man except as they are related with other facts. 

 To comprehend the relations requires some scheme; the difficulty 

 is to have a scheme without confusing the procedure with the find- 

 ings ultimately represented. Of course, ' ' nature ' ' is not limited by 

 one or many schemes, it may be guessed, nor bounded by the num- 

 ber of dimensions portrayable in human record or thought. 



The study and representation of such order and relations may 

 be included in the term quantitative descriptive physiology. By its 

 scrupulous definition of variables it indicates plans, and maybe the 

 only one, by which relations are found, comprehended, and con- 

 ceived. Pareto ('34, p. 178) speaks of mathematics as ''the one 

 method so far discovered for dealing with interdependencies." 



It may be said that science constructs maps of one portion or 

 another of phenomena. All maps are inaccurate, corrections are 

 continually required. ''Empirical" observation stores instances; 

 description makes maps. Where no map exists an explorer has 

 completely blind choice of where to go, and in an early stage of a 

 science it seems almost fortuitous what observations are made 

 next. But later, in proceeding from one observation to another, 

 various mental gymnastics ("logics") are introduced to suggest 

 connections. I believe quantitative description is the known means 

 of coordinating those particular bodies of data that show how an 



