No. o3Sj 



MUTATK) X IX (EX OTI1 E H . 1 



601 



tioii" is due to a more general disturbance of the germ 

 plasm than would occur in hybrid splitting. 



The mutation process, therefore, while probably a re- 

 sult of previous crossing in the ancestry of 0. Lamarcki- 

 ana, is not a simple case of the splitting off or reappear- 

 ance of types which entered into that ancestry. It is 

 probable that much of the hybridization-behavior of the 

 genus (Enothera, including particularly crosses which 

 involve the mutants of 0. Lamarckiana as one of the pa- 

 rents, will ultimately be harmonized with Mendelian 

 categories. But it is also probable that the appearance 

 of the "mutants" of 0. Lamarckiana is not a case of 

 Mendelian splitting as we understand that process at 

 present. The origin at least of such mutants as (). ,>'ni<i< 

 and 0. rubricalyx can not be explained on this basis, and 

 at present can only be ascribed to a general disturbance 

 or condition of instabilitv, which probably resulted from 

 previous crossing. The" change in climate to which 0. 

 Lamarckiana has been subjected may also very well have 

 had something to do with this disturbance, although this 

 is less probable since this plant when brought hack To 

 America continues to exhibit the same mutation phe- 

 nomena. 



I have tried to show (1) that 0. Lamarckiana, like 

 other allogamous forms, has undoubtedly undergone 

 crossing in its ancestry, and (2) that, whatever may have 

 been the relation between these crosses and the appear- 

 ance of mutants, the important matter to decide from the 

 evolutionary standpoint is, Will these forms survive in 

 nature and become the starting points for new races? 

 As regards 0. gigas, I have pointed out (Gates. '09ft) 

 many cases among plants, of species which have probably 

 originated in an analogous manner. 



There is one further phase of the mutation process 



