No. 538] 



MUTATION IX (KXOTHKRA 



599 



simple hybrid. 8 These facts that, so far as now known, 0. 

 biennis, 0. Lamarckiana and 0. grandiflora all behave 

 alike in crosses with each other (that is, they give blends), 

 together with the fact that none of the mutants revert 

 towards either 0. biennis or 0. grandiflora, have always 

 seemed unfavorable to the view that 0. Lamarckiana has 

 originated suddenly from such a single cross. On the 

 other hand, it by no means follows that 0. Lamarckiana 

 or various other species of (Enothera could not be pro- 

 duced by sufficient crossing of races under suitable condi- 

 tions. My studies of a colony of Oenotheras from the Eng- 

 lish sand dunes, show that numerous races which breed 

 true and have numerous entirely unanticipated characters 

 (often differing from each other conspicuously in every 

 organ) can be obtained as the result of indiscriminate 

 crossing. Some of these races when crossed give blends 

 and have thus far defied Meudelian analysis. I have been 

 able to synthesize at least one of these true-breeding 

 forms by crossing of races. A number of the other pecu- 

 liar constant races have doubtless had a similar origin, 

 though too complex to be easily repeated. The character- 

 differences of these races do not appear to be Meudelian 

 and the races come true except for varying percentages of 

 1 ' mutants." 



The mutants of 0. Lamarckiana themselves, when 

 crossed with their parent form, show apparently a differ- 

 ent type of behavior from the species, but that is a matter 

 I will not enter into in the present paper. 



How do the Mutants of 0. Lamarckiana Originate?— 

 Tower's experience ('10, p. 315) in crossing species of 

 Leptinotarsa and thus synthesizing a new race which 

 breeds true except for sporadic "mutants," appears to 



8 In my cultures of Oenotheras from the sand dunes of the coast of 

 England, I have obtained occasional mutants having many of the flower- 

 characters of O. biennis. But in this case there was abundant opportunity 

 for recent crossing with 0. biennis to have taken place, and this would seem 

 to be the most probable explanation of their occurrence. I have more re- 

 cently seen similar occurrences in forms from botanical gardens which have 

 obviously undergone much crossing. 



