46 GENETICS 



3. Inconstant species, (scintillans and lamarckiana 



itself), that do not always breed true but pro- 

 duce mutants, and 



4. Degressive species, (lata, albida), which are de- 



fective in some way and are incapable of main- 

 taining themselves in nature. 



DeVries' experiments and observations have been 

 repeated on a large scale and extended, notably by 

 MacDougal in the New York Botanical Gardens, by 

 Shull at the Carnegie Institution for Experimental 

 Evolution, Cold Spring Harbor, Long Island, and by 

 Gates in England, and his conclusions have been con- 

 firmed in all essential points. The mutability of 0. 

 lamarckiana is as unmistakable and as diverse in 

 America and England as it is in Holland. 



The critics of deVries, however, regard (Enothera 

 lamarckiana as a hybrid to begin with, from which 

 different strains have simply been bred out. Both 

 Bateson and Lotsy have called attention to the pres- 

 ence of deformed or defective germ cells in (Enothera 

 lamarckiana as evidence of its hybridity, and Davis, 

 by crossing 0. franciscana and O. biennis, has pro- 

 duced a hybrid (Enothera, which he has christened 

 (Enothera neo-lamarckiana because it not only resem- 

 bles 0. lamarckiana but behaves like it in producing 

 mutations. He consequently proposes "dissolution of 

 hybrids" as a substitution for mutation in explaining 

 the phenomena that deVries has described. 



It is somewhat questionable whether this classical 

 plant, which has added at least a five-foot shelf to 



