282 



TWIN HYBRIDS OF OENOTHERA HOOKERI T. AND G. 



The beds were inspected also during August and September, when 

 in full flower and with ripening fruits, but the result was the same. 



As was to be expected from the behavior of the crosses of Oe. 

 Hookeri, the reciprocal cross did not give this splitting in the second 

 generation. 



If we compare these figures with those given in my book on Grup- 

 penweise Artbildung (p. 103) for the second generation of the hybrids 

 of Oe. Hookeri, we find an almost exact parallelism, since the figures 

 for rubiennis were in that case 15, 20 and 22 percent. 



Crosses of Oe. Lamarckiana mut. velutina with Oe. biennis 



From the considerations, given above, it may be deduced that 

 Oe. Lamarckiana mut. velutina (syn.: Oe. blandina) must behave in 

 an analogous manner in its crosses with the pollen of Oe. biennis. 

 Therefore I examined this case, but only to such an extent as seemed 

 necessary to verify the conclusions. 1 made the cross in 1915, sowed 

 the seeds in 1916 and had a uniform lot of hybrids, which bore the 

 type of Oe. Lamarckiana x biennis {Opera VII, p. 317). Among 

 these I self-fertilized two specimens and found among their off- 

 spring, 24 and 31 percent rubiennis and 76 and 69 percent blandina- 

 like hybrids. The former exactly resembled the first hybrid generation, 

 the latter duplicated the pistil parent of the original cross. 



Summarizing our results, we have for the crosses with the pollen 

 of Oe. biennis the following mean constitution of the second generation. 



Table 12 

 Hybrids of Oe. biennis in the second generation. 



As in so many analogous cases the figures for the velutina-Wke 

 offspring are higher than 50 percent. This shows that the splitting 

 does not go into equal groups. 



Crosses of Oe. Hookeri with Oe. suaveolens 



Crossed with different species, Oe. suaveolens Desf. gives, as a rule, 

 uniform and constant hybrids. So, e.g., Oe. biennisxOe. suaveolens 



