NATURAL SELECTION 133 



characters" broke down on examination. The char- 

 acters, if acquired, were due at least in part to the 

 hereditary constitution, and hence would be reproduced 

 from the germ cells. Even if the inheritance of ac- 

 quired characters sometimes occurred, it was certainly 

 too rare to be important. 



The process of natural selection of course knows noth- 

 ing of these matters. The creature is selected on ac- 

 count of what it is, no matter how it became so. Thus 

 a highly educated person of mediocre ability would have 

 an advantage over an uneducated one who might be 

 markedly superior from the standpoint of inheritance. 

 If the principal characters of organisms were such as are 

 not inherited, natural selection could do nothing for 

 evolution ; there would be no relation between fitness 

 to survive and ability to leave fit offspring. Obviously, 

 this is not true ; but we can no longer assume that all 

 sorts of variations tend to be inherited. 



7. The researches of Mendel, greatly extended and Limitations 

 supplemented in later years, have shown that many ^ selection * 

 individuals are heterozygous or cross-bred. These, 

 though "selected," will not come "true." They break 

 up into all sorts of new combinations. This is why 

 eminent men do not usually have sons equal to them- 

 selves. Some types, such as the "blue" Andalusian 

 fowl, are incapable of existing except as heterozygotes, 

 and no process of selection will cause them to have more 

 than half their offspring "blue." The other half will be 

 blacks and speckled whites. Furthermore, selection 

 cannot eliminate the recessives, - - those determiners 

 which may be present in the germ plasm of cross-bred 

 individuals without producing any effect. Morgan, in 

 working with flies, has found a number of "lethal" 

 factors, which when received from both parents are 



