494 On the Latent Capacity for Mutation. 



not affected by the struggle for existence in which the 

 organisms bearing them are engaged ; they do not weigh 

 in the scale on the one side or on the other. They multiply 

 or perish precisely as their bearers do. But one Oenothera- 

 f ruit may bear between one and two hundred seeds ; and a 

 vigorous plant bears hundreds of fruits. So that it prac- 

 tically never happens that all the seeds produced become 

 flowering plants. It is very largely a matter of chance 

 which of these survive, and therefore a matter of chance 

 which latent mutations continue and which perish. 



I conclude from this that the number of mutations 

 actually observed is no measure at all of the number 

 which presumably arose during the premutational period. 



We must think of the origin of groups of closely re- 

 lated species in other genera and families as essentially 

 analogous to the mutational period in Oenothera La- 

 mar ckiana. Examples of such groups are the long se- 

 ries of elementary species, our knowledge of which we 

 owe to Jordan and his pupils ; and the well-known nebu- 

 lous groups of the old systematists such as Fries, Nageli 

 and others. It is reasonable to suppose that the numer- 

 ous elementary species of Draba verna all arose in a 

 single period in a small locality and that they have spread 

 thence, hither and thither over the whole of Europe.^ 

 Or perhaps they have arisen during the period of distribu- 

 tion. The same may be said of Viola tricolor, Helian- 

 themitm vulgare, etc. The whole appearance of Draba 

 verna, at present, points to a period of mutation of exactly 

 the same kind as that of Oenothera Lamarckiana. 



^ A few species of Draba verna ought to be cultivated side by 

 side in every botanical garden. Their differences and constancy are 

 perfectly obvious and such as to strike the eye of every visitor. I 

 have only two species in cultivation; but even those have been of 

 general interest already. 



