154 THE PROBLEMS OF EVOLUTION 



shown any perceptible approach to the local type. 

 They are still obviously of the sonoriensis race J ^ ^ 

 Sumner's results in this and in other papers ^ on 

 the same subject show great care in the evaluation 

 of results and a clear appreciation of their true 

 significance. His published results may be taken 

 as fairly sound evidence that geographic races, or 

 subspecies, may be germinally stable. The only 

 doubtful factor is time; it is conceivable that a 

 long succession might show responses which are 

 not attained in a few generations. But this, after 

 all, is only a possibility which must be taken into 

 account for the sake of accuracy. The experiments 

 cited covered a period of eight years. 



What do these differences indicate .^^ First they 

 show that characters may exist in nature genera- 

 tion after generation without being permanent to 

 the same degree as strictly heritable characters. 

 They illustrate very clearly the point that char- 

 acters may be definitely a product of a particu- 

 lar environment without fitting the individual to 

 meet the conditions which produce them, since in 

 the species mentioned, the distinctive characters 

 are superficial. They suggest two possibilities as 

 to the origin of permanent characters, one that 

 they are based on mutations as ordinarily inter- 

 preted and the other that they are the ontoge- 

 netic characters, or individual responses, developed 



8 Op. cit., p. 697. 8 ^^. ]S!ai.^ Vol. LVIII, pp. 481-505, 1924. 



