564 



THE AMERICAN NATURALIST [Vol. XLIV 



tion), have no connection with evolution except that they may 

 impede the process. If all organisms came from one original 

 intergenerating stock, deriving its food from the inorganic 

 world, and from the dead individuals of its own kind, how far 

 could evolution have progressed without any formation of sepa- 

 rate species ? What would now become of the organic world if 

 isolation and selection ceased and all the separate species were 

 merged in one.' When I say that there may he evolution with- 

 out isolation I mean without a<I<lit innnl isolation. I do not mean 

 that the undoing of all the effects produced by vital forces 

 making isolation complete, though the different genera occupy 

 the same district, would he an advance step in evolution. On 

 the contrary, I think that such an undoing would mean the 

 crumbling of the whole fabric of the organic world. 



Is not racial evolution a term that we can rightly and wisely 

 apply to all the processes of change in organisms affecting 

 characters that are subject to the laws of heredity and variation ? 

 May it not he applied to all changes in races and species result- 

 ing, not only from the action and reaction of members of the 

 same species upon each other, but also from the action and re- 

 action between individuals or groups and their environments? 

 May not evolution be either divergent, convergent or parallel? 

 either progressive, or retrogressive ? May it not take place with- 

 out any change in the environment, and in that sense be spon- 

 taneous; and may it not be due to vital action stimulated, 

 guided, and controlled hy external conditions? 



Honolulu, Hawaii. John T. Gulick. 



RETROACTIVE SELECTION 1 



In his contribution to The American Naturalist of July. 

 Professor Marshall makes some statements which I wish to cor- 

 rect. Among other things he says : 



