SOME ASPECTS OF THE BIOLOGY OF HUMAN POPULATIONS 539 



at the end of this chapter Pearl, 1925, 19276 and other 

 papers cited in the sources referred to. 



In brief what is found is that both of the fundamental 

 first-order population variables, natahty and mortahty, 

 are directly and markedly influenced by density of popula- 

 tion above a certain magnitude. The sense of this influence 

 is that fecundity rates are markedly lowered by small 

 increases in density at relatively low densities, while after a 

 certain density is passed further increases produce only 

 shght decreases in birth rates down to an asymptotic limit; 

 and, second, that death rates are insignificantly aff"ected by 

 increasing density at relatively low densities, while after a 

 certain density is passed death rates markedly increase 

 with increasing density up to an asympototic limit. 



In short it is possible to account for all the main features 

 of the growth of experimental populations of the fruit-fly, 

 by a simple hypothesis as to the correlated behavior of 

 three variables, natality mortahty, and density. There is 

 some evidence that the situation is similar in human popula- 

 tions, in its fundamental biology, although there the influence 

 of other second-order variables comphcates the situation. 



THE COMPOSITION OF POPULATIONS AND DIFFERENTIAL 



FERTILITY 



The total size of a population is of course only one of its 

 attributes. There are others of great biological interest, 

 notably the composition of a population in respect of 

 diff"erent kinds of individuals. All societies, whether of ants 

 or men, tend to difl^erentiate into castes, each performing 

 a different function in the whole integrated group. Wheeler 

 has particularly discussed this phenomenon. 



The writer has recently studied one aspect of this matter in 

 the population of the United States (1927a). In a recent 

 report on natahty from the United States Census Bureau 

 (Birth Statistics, 1924) there is a table (numbered 10, pp. 

 1 71-18 1) which makes available some new and welcome data 

 regarding difl"erential fertihty in this country. 



The data apply to the United States birth registration 

 area exclusive of Delaware, Maine, Massachusetts, New 

 Hampshire, Rhode Island and Indiana. 



