276 Heredity and Environment 



stimulation of hybridization or "heterosis," as it has-been called, 

 offers an extremely important means of quickly producing very 

 vigorous and fruitful individuals, but not lines or races which 

 breed true. 



2. Mutations. Mendelian association and dissociation of 

 characters produces new forms of adult animals and plants but 

 not new hereditary characters. Permutations of Mendelian 

 characters we may have almost without number, of new 

 combinations of these there may be no end, but no new unit char- 

 acters are formed by such temporary combinations, no new in- 

 heritance factors are created or evolved. New combinations of 

 factors may be cornpared to new combinations of chemical ele- 

 ments ; you can always get out of the combination what went into 

 it, whereas new factors are comparable to the changes which take 

 place in certain atoms, for example radium, by which the element 

 itself is changed in an irreversible manner. The discoveries of 

 Mendel show us how to follow inheritance factors through many 

 combinations and through many generations, but they do not show 

 us how new factors arise. These discoveries have given us an 

 invaluable method of sorting and combining hereditary qualities, 

 but Mendelian inheritance as such does not furnish the materials 

 for evolution. 



In 1901 Hugo deVries startled the scientific world by the pub- 

 lication of his great work on the "Mutation Theory" of evolu- 

 tion in which he proved that the evening primrose, Oenothera 

 lamarckiana occasionally produced "sports" or "mutations" which 

 differed so much from the parent form that they deserved to be 

 called new species (Fig. 100). He discovered and studied a 

 large number of these mutations in Oenothera as well as in some 

 other plants and concluded that evolution takes place by steps 

 or jumps rather than by "creeping on from point to point" as 

 Darwin believed. 



Several geneticists have expressed doubt as to whether there 

 are any such 'things as mutations in the sense of deVries, 



