No. 561] GENETIC AL STUDIES ON JEN OTHER A 565 



I am perfectly willing to admit the complexity of my 

 stock material of grandiflora and biennis, and also the 

 possibility that the forms may not have been strictly 

 homozygous at the time the crosses were made. It was 

 in no wise necessary for the purposes of my experi- 

 ments that they should be strictly homozygous. My only 

 concern was that the material should be American types 

 of (Enothera without the possibility of contamination 

 through crosses with Lamarckiana. That my forms of 

 biennis and grandiflora had these qualifications there 

 can, I think, be no doubt. They have, as a matter of fact, 

 bred true in the small cultures which have been carried 

 through two generations for biennis A and biennis D 

 and four generations for grandiflora B. 



An abstract of my argument is as follows: (1) Since 

 hybrids of biennis and grandiflora show points of strong 

 resemblance to Lamarckiana and, (2) since the behavior 

 of these hybrids in the F 2 and F 3 parallel closely the be- 

 havior of Lamarckiana when it gives rise to "mutants," 

 (3) therefore, there are strong reasons for believing that 

 the "mutations" of Lamarckiana are due to instability 

 of its germinal constitution resulting from a hybrid 

 origin. The fact that (Enothera Lamarckiana is not 

 known as a component of any native (Enothera flora and 

 the fact that its known history has been entirely as a 

 cultivated plant or as a garden escape naturally greatly 

 strengthen the force of the above argument. 



It does not seem to me that these arguments are 

 answered by a supposition that the behavior of my hy- 

 brids involves a habit of mutation inherited from the 

 parental types. On the contrary, are we not justified in 

 asking of the mutationists evidence from material the 

 status of which, as representative of a wild species, is 

 beyond question? Stomps ('12b) has apparently en- 

 deavored to meet the situation by a study of a cross be- 

 tween the biennis and cruciata of the sand dunes of Hol- 

 land. From the cross he obtained in the second genera- 

 tion a biennis nanella and a biennis semi-gig as. Both 



