ON TWIN HYBRIDS. 



In the group Onagra of the evening primroses the hybrids between 

 different species are, as a rule, constant and uniform through the 

 succeeding generations. In this respect they comply with the com- 

 mon rule for the majority of the characters of specific hybrids. Only, 

 on account of the scarcity of regressive marks in this group, the 

 phenomenon of constancy is allowed to show itself pure and complete. 

 An exception is afforded by Oenothera brevistylis, the character of 

 which splits up according to the formulae of Mendel^). 



As an example of a constant hybrid race I quote the 0. muricata 

 X biennis. It has been found uniform through at least four gene- 

 rations. In the second, the one in which Mendelian marks are seen 

 to split, I cultivated over 80 flowering plants and over 100 rosettes, 

 but no differences could be detected. Last summer (1907) I cultivated, 

 on neighbouring beds, 35 plants of the first and 1 8 of the fourth gene- 

 ration. Both groups produced a number of flowering stems, but, 

 apart from the ordinary fluctuating variability, the characters were 

 exactly the same in all the specimens. They had the flowers almost 

 like those of 0. biennis, but dense and richly flowered spikes like the 

 0. muricata. 



The types of the species used for this experiment were the forms 

 which occur everywhere on waste places throughout Europe, having 

 been introduced, the first from Virginia and the second from Canada, 

 about three centuries and one century ago respectively. They are 

 probably the types on which Linnaeus based his descriptions of the 

 species. In the United States, however, these Linnean species con- 

 sist of quite a number of subspecies, of which I collected over a 

 dozen in Kansas, Missouri, Illinois, and elsewhere, during the summer 

 of 1904. On sowing them in my experiment garden, I observed them 

 to be fairly distinct, each constituting a sharply defined type. The 

 form which is the most common throughout the United States is not 

 the same as the one introduced into Europe, neither for 0. biennis 

 nor for 0. muricata, as has also been pointed out by MacDougal 

 for 0. biennis. 



i) Die Mutationstheorie 2 : 4.29. 



