288 TWIN HYBRIDS OF OENOTHERA HOOKERI T. AND G. 



almost linear leaves. They appeared in larger numbers than in the 

 corresponding cross of Oe. Hooker i. I made the cross more than once, 

 on one plant in 1913 and on two in 1915. The first gave 126 offspring 

 with 9 percent of linear leaved specimens, the others gave 14 and 

 23 percent of this type among 70 and 170 plants. The linear leaved 

 specimens were all exactly of the same type, with gray hairy foliage 

 and thickly conical flower buds. Moreover their progeny was also 

 uniform. I cultivated 60 specimens of the second and 63 of the third 

 generation, but did not find the least deviation. 



The typical hybrids of both crosses resembled one another fully. I 

 had a second and a third generation in 1915 and 1916. The second 

 was uniform, without narrow-leaved plants, but with one lutescens 

 among 32 specimens. Two of the typical ones were self-fertilized and 

 yielded each in 1913 a progeny of 70 flowering plants, among which 

 no linear leaved ones were seen. But 9 and 4 lutescens were counted. 

 In order to make wholly sure that no linear-leaved forms are pro- 

 duced by self-fertilized seeds of the hybrids, I repeated the first 

 generation in 1915, self-fertilized five typical specimens and sowed 

 their seeds in 1916. The cultures embraced 40—100, together 382 

 plants, but only one specimen of the linear-leaved type appeared. 

 It was a weak and low individual, but resembled in all respects the 

 progeny of the linear-leaved individuals of the first generation. 

 The cultures embraced 28 specimens of lutescens and four of the type 

 described as mut. jaculatrix (de Vries 1916 a). 



Thus we see, that our experiments coincide with the cross between 

 Oe. suaveolens and Oe. Hookeri in regard to the appearance of types 

 which are more or less evidently due to mutations of the former 

 species. But the splitting into two almost equal groups, which is 

 so characteristic of the crosses of Oe. Hookeri, is absent here, even as 

 it is absent in all the other crosses of Oe. Cockerelli. 



Summary 



1. Oenothera Hookeri T. and G. produces a splitting laeta and a 

 uniform velutina in its crosses with Oe. grandiflora, exactly as it does 

 in those with Oe. Lamar ckiana. 



2. This production of twin hybrids in the first generation is due to 

 the mass mutation of Oe. grandiflora into Oe. mut. ochracea. The 

 typical gametes produce the velutina, whereas the mutated ones give 

 rise to the laeta. 



3. The percentage figures for these splittings were 41 for velutina 



