TWIN HYBRIDS OF OENOTHERA HOOKERI T. AND G. 271 



1 found: 



Table 5 



The average percentage of 27 coincides with the splitting of 

 Mendelian monohybrids and indicates the small flowers as the 

 recessive character. But none of the grandparents had small flowers. 

 The most interesting point is, however, that the size of the flowers 

 shows a splitting, whereas all the vegetative characters, which 

 cause the velutina to resemble Oe. Hookeri so very much, are trans- 

 mitted as a whole. 



Crosses of Oe. grandiflora with Oe. franciscana Bartlett 



The high degree of affinity between Oe. franciscana (Bartlett 

 1914) and Oe. Hookeri in external marks led me to the expectation 

 that the behavior in crosses might also be the same. Both are Cali- 

 fornian species with bright large flowers and long narrow leaves. 

 1 received artificially self-fertilized seeds through the kindness of 

 Mr. Bartlett, sowed them in 1915 and conducted my cultures so 

 as to have annual plants. I fertilized a typical specimen of Oe. 

 franciscana with the pollen of a plant of my race of Oe. grandiflora 

 and cultivated in 1916 the fiist generation in 60 individuals, all 

 of which flowered in July and August. They showed two types, 

 analogous to the laeta and velutina of the previously described 

 crosses. The foliage of the first was broad and pure green, that of 

 the latter narrow and brown green. These differences slowly in- 

 creased during the time of flowering, but remained small. There 

 were 43 percent laeta and 57 percent velutina. The next year the 

 second generation of the latter were uniform in respect to their 

 typical marks, although there was some variation in minor characters, 

 as e.g., in the size of the flowers, the rapid production of stems, 



