The Laevifolia-Family. 265 



I must now return to the transplanted seedlings. Sev- 

 eral of them were weak, and died sooner or later ; espe- 

 cially the most easily recognizable of them all, O. albida. 

 In the case of others too many were raised for them all 

 to flower. But I grew the majority of them during the 

 whole summer; some of them flowered; others remained 

 in the rosette stage. I saved seeds from O. albida in 1897 

 (from plants which came up in 1896) from 0. ntbri- 

 nervis, O. nanella and O. scintillans. The first three 

 when self-fertilized bred true, the last did not (see page 

 244). Further details on this point will be given in the 

 second chapter under the heading of the plant in ques- 

 tion. 



§ 6. THE LAEVIFOLIA-FAMILY. 



In 1887, I noticed in the locality for Oenothera La- 

 marckiana near Hilversum, a group of individuals whose 

 peculiar characters showed them to be a distinct form. 

 I therefore gathered some of their seeds and sowed them 

 in the following year in my experimental garden. The}^ 

 gave rise to two forms (as was to be expected from the 

 fact that no attempt was made, or was possible, to insure 

 self-fertilization) of which the one was the ordinary 

 Lamarckiana whilst the others were like the parent plant. 

 T propose to call this subspecies O. laevifoUa on account 

 of its smooth leaves which are not or at any rate scarcely 

 perceptibly crumpled, as is the case in Lamarckiana. 



For the first few years I let the two forms grow and 

 flower together and took no further precaution than iso- 

 lating them from the rest of the cultures. In the year 

 1894 I began the practice of enclosing the 1)lossoms be- 

 fore opening in paraflined paper bags and of fertilizing 



