510 Conclusion. 



biennis and 0. mnricata. 0. alhida would certainly be 

 too weak; and the inconstant and sterile or partly sterile 

 forms would obviously perish sooner or later. 



But direct investigation is necessary before a satis- 

 factory answer to this question can be given. 



Lastly, let us examine my results from a theoretical 

 standpoint. Two points stand out prominently. First, 

 the question as to the origin of the whole mutation pe- 

 riod ; and secondly the analogy between the phenomena 

 observed and the origin of species in general. 



We regard the beginning of the mutation period as 

 the time when the latent potentialities, present in this 

 period, first arose. For obviously the capacity for pro- 

 ducing 0. gig as has not been a property of all the an- 

 cestors of my Lamarckiana; it must have arisen at some 

 time : and similarly with the other species. Are these 

 potentialities as old, or perhaps even older than La- 

 marckiana itself ? Probably not. It is simpler to suppose 

 that they either arose on the spot where they first ap- 

 peared, or a little earlier in the life of the species. 



I call the origin of the latent potentialities. Premuta- 

 tion; the mutations themselves are only the expressions 

 of these. One of the objects of future investigation 

 ought to be to determine the essential conditions of this 

 premutation and if possible to induce it at will. 



Periods of mutation must occur, or at any rate must 

 have occurred, times without number, in nature. For 

 groups of allied species, which, so far as one can judge, 

 are related to one another exactly as my Oenotheras are, 

 occur throughout the animal and vegetable kingdom. 

 Wherever the constancy of the characters of such species 

 has been established in cultures they constitute the es- 

 peces affines of Jordan. I have, already, often men- 



