582 H. S. Jennings 



greater permanence of certain combinations and activities is evi- 

 dent everywhere outside the hmits of organisms, w^hile within the 

 system making up the individual organism there are conditions 

 which require the prevalence of this principle of operation, on a 

 large scale. To selection, or the greater permanence of certain 

 combinations, within the organism, we must look for an under- 

 standing of many of the most important problems of biology, and 

 particularly of those having to do with adaptation. The study 

 of the internal adaptations of organisms might indeed be defined 

 as the search for those combinations of structure and activity 

 that are most lasting. The study of the laws in accordance w^ith 

 vv^hich certain combinations are lasting, while others are fleeting, 

 must become one of the main lines of investigation. The pioneer 

 v^ork of Roux ('81) in this line was most promising, and has been 

 followed up to a certain extent; but thorough experimental investi- 

 gations along such lines are what is needed. In the meantime, 

 the relative permanence of those combinations which we call 

 individuals must remain one of the chief objects of study. As 

 Kellogg ('07) has well noticed, we have few, if any, cases even of 

 this, that are clearly and accurately observed and analyzed. 



All together, it is clear that a study of the processes which result 

 in the complex "adapted" organism must be largely a study of 

 the relative permanence of different combinations — a study of 

 selection. This of course requires a study of the chemico-physical 

 laws in accordance with which the processes are brought about, 

 and in accordance with which some of their products are more 

 lasting than others. It is in many respects unfortunate for an 

 understanding of this line of work that it has received the figura- 

 tive and anthropomorphic name of selection. When we speak 

 merely of the relative permanence of diff'erent combinations 

 (whether these combinations are individuals, processes, or chem- 

 ical compounds), we call up no associations foreign to the matter 

 in hand, and thus run no risk of arousing misconceptions and 

 prejudice due to such associations. 



In studying the racial processes, that have resulted in giving the 

 organism its "hereditary" properties, we meet one great diffi- 

 culty. We cannot reproduce the long series of conditions which 



