708 Mutations 



more simple lines. But the changes have by no 

 means removed the doubts, nor have they been 

 able to supplant the general impression of dis- 

 tinct groups, combined by broad lines. This 

 feature is very essential, and it is easily seen to 

 correspond with the conception of swarms, as 

 we have deduced it from the study of the lesser 

 groups. 



Genealogic trees are the result of comparative 

 studies; they are far removed from the results 

 of experimental inquiry concerning the origin of 

 species. What are the links which bind them 

 together? Obviously they must be sought in 

 the mutative periods, which have immediately 

 preceded the present one. In the case of the 

 evening-primrose the systematic arrangement 

 of the allied species readily guides us in the de- 

 limitations of such periods. For manifestly 

 the species of the large genus of Oenothera are 

 grouped in swarms, the youngest or most recent 

 of which we have under observation. Its imme- 

 diate predecessor must have been the subgenus 

 Onagra, which is considered by some authors as 

 consisting of a single systematic species, Oenoth- 

 era, biennis. Its multifarious forms point to a 

 common origin, not only morphologically but 

 also historically. Following this line backward 

 or downward we reach another apparent muta- 

 tion-period, which includes the origin of 



