498 ON TRIPLE HYBRIDS. 



With Lamarckiana it produces twins identical in both reciprocal 

 crosses and both of them as large-flowered as the parent species. 

 The /fl^/fl is bright green, with smooth leaves and slender flower buds. 

 The velutina is hairy and reddish, with furrow-shaped leaves and thick 

 buds. It flowers more profusely and resembles 0. Hookeri almost 

 exactly, but is stouter, with dense and richly flowered spikes. 



The twins derived by 0. Hookeri from 0. lata cannot be distin- 

 guished externally from those derived from 0. Lamarckiana. 



The American subspecies of 0. biennis used for my experiments 

 was collected by me in 1904 at Chicago, near Jackson Park. It seems 

 to be the same form as that which is most common in the eastern 

 states. Other subspecies I have collected in Pennsylvania, Kansas, 

 Missouri, and elsewhere. All of them are quite different from the 

 European form, which is the one used in my Mutations-Theorie and 

 in my article on twin hybrids {Opera VI, p. 472). A systematic 

 treatment of these numerous forms is still wanting, and therefore I 

 will provisionally designate my strain by the name of its source and 

 call it Chicago. Its most curious quality is that of producing twins 

 when combined as a male parent with 0. Lamarckiana and not when 

 crossed with the pollen of this species. In this respect it is exactly 

 opposite to the European 0. biennis, and therefore very useful. 



The 0. biennis from Chicago is a taller plant, more richly branched 

 and darker green than the European form. Its flowers are much 

 smaller, sometimes reaching the size of those of 0. muricata, but 

 deeper yellow. The stigma is surrounded by the anthers which 

 open early in the bud. The velutina and laeta which it produces 

 from Lamarckiana can hardly be distinguished from those produced 

 by the European 0. biennis, although the leaves are narrower and 

 darker green. I cultivated this strain during three succeeding 

 generations. 



0. scintillans and 0. lata are the mutant species described in my 

 Mutations-Theorie. The scintillans used originated from Lamarck- 

 iana seed of 1889 which was sown in 1895. In 1906 1 had the fourth 

 generation from continued pure self-fertilizations and used it for my 

 cross. From this same strain the lata used in the experiments of this 

 article arose as a mutant in 1901. It was artificially fertilized. My 

 crosses were made with specimens of the second and third generation 

 of its progeny. 



I will now describe the crosses from which the triplets resulted. 



Oenothera scintillans x strigosa. — This cross was made in July, 

 1907, between two biennial specimens. It yielded a small lot of seed 



