MUTATIONS OF OENOTHERA SUAVEOLENS DESF. 259 



self-fertilization might split into three types. If then we assume 

 the presence of two lethal factors, one linked to the normal and 

 the other to the mutated gametes, this would explain the presence 

 of one-half of empty grains among the seeds, the absence of the two 

 expected constant types and the repeated splitting of the third type. 



Crosses of Oe. suaveolens with Oe. Lamarckiana 



From all the results of our experiments as well as from the syste- 

 matical descriptions, we may conclude that Oe. suaveolens and Oe. 

 biennis are near allies, resembling one another in numerous and 

 prominent characters. From this consideration we might further 

 infer the probability, that Oe. suaveolens would behave in the same 

 way in its crosses with Oe. Lamarckiana and split this species into 

 the hybrid types laeta and velutina, if Oe. suaveolens is the female 

 parent, but yield uniform hybrids if its pollen is made use of in the 

 cross. On the other hand, the discussion of the possibility of a mass 

 mutation, just given, would lead to the expectation of the appearance 

 of mutants of the type lutescens, even as in the case of Oe. suaveolens 

 x biennis described above. As a matter of fact this mutability 

 occurred in both the reciprocal crosses, as will be shown below. 

 In the crosses with other species the normal gametes would yield 

 the intermediate hybrids, whereas the lutescens gametes should give 

 lutescens hybrids. But why these should occur so rarely and fail in 

 so many of our experiments, remains to be investigated. In crosses 

 with Oe. Lamarckiana four types of hybrids would thus have to be 

 expected, of which two might be externally identical as in the case 

 of Oe. grandiflora x Oe. Lamarckiana. This would explain the oc- 

 currence of laeta, velutina and lutescens in the proportion of 2 : 1 : 1 . 

 A large degree of likeness in the behavior of Oe. suaveolens to that 

 of Oe. Lamarckiana and Oe. grandiflora would thus be revealed. 

 But since the aim of my experiments was essentially the study 

 of the mutability of Oe. suaveolens and not that of its hybrids, I 

 have postponed these questions to a later opportunity. 



Oe. Lamarckiana x suaveolens. — I made this cross in 1914 and 

 cultivated 59 offspring in 1915, all of which became annual. When 

 they reached a height of half a meter in July and proved uniform 

 with the exception of the lutescens, I threw away one half of them 

 and left the others to flower in August and September. I counted 

 46 normal specimens and 13 lutescens, giving a percentage of 22 

 for the latter. The lutescens resembled the hybrids of the same name, 



