554 DOUBLED CHROMOSOMES OF OENOTHERA 



allowing a clear grouping. Among these, the forms of velutina were 

 the most reliable. 



For this reason we have chosen for the present investigation the 

 pollen of the twin hybrid 0. (biennis x Lamar ckiana) velutina. There 

 are many mutants, especially among the sesquiplex types, which 

 yield sufficiently pure velutina pollen (6), and some of them have 

 been used for control experiments, as will be shown elsewhere. 

 The twin hybrids of 0. Lamarckiana are old forms described since 

 1907, however, and studied by many investigators. Moreover, they 

 easily may be reproduced by anyone interested in the present 

 discussion. 



Oenothera (lata x Lamarckiana) setnigigas 



In 1915 a mutant lata originated from the pure race of 0. La- 

 marckiana, and was chosen for the production of a new line. Every 

 generation of this line has been fertilized by the pollen of 0. La- 

 marckiana, since the lata plants had no good pollen, as is usual in 

 our climate. In the third generation, in 1921, two rosettes of the 

 type of semigigas were observed. The chromosomes in the root- 

 tips of both of them were counted and found to be 21. The plants 

 remained in the condition of rosettes of leaves until the winter, 

 and flowered in 1922, when they were used for this experiment. 

 They were cultivated in the glass covered part of the experiment 

 garden at Lunteren, under very favorable conditions, and flowered 

 richly. On one of them thirty flowers of the side branches were 

 pollinated, after a castration, with the pollen of 0. (biennisx La- 

 marckiana) velutina. They yielded 30 apparently good fruits, with 

 3.3 cc. of seeds, about one quarter of which contained viable germs, 

 as was proved in a test-tube at 30° C. The other seeds had imperfect 

 germs of different sizes. 



The harvest was sown at the end of January, 1923, in a glass- 

 house, and 81 seeds germinated. The young seedlings showed a 

 motley assembly of types of leaves, among which some were more 

 frequent and could soon be recognized. After two months they 

 were transplanted singly into small pots, and from these they were 

 planted in the beds, in May and June, as soon as their root tips 

 had been fixed for the counting of chromosomes. At that time the 

 mutant characters could be recognized clearly in most of the rosettes 

 of root leaves, and the plants were grouped accordingly before 

 being planted. This method has the advantage that, at the time 



