230 



THE AMERICAN NATURALIST [Vol. XLIV 



tion to the five possible hybrid forms, were obtained also 

 the four possible stable forms. So that I now have fer- 

 tile plants, free of all hybridity, of four sorts, viz., 18 

 with pubescent capsules and dark seeds, 9 with pubescent 

 capsules and buff seeds, 18 with glabrous capsules and 

 dark seeds, 44 with glabrous capsules and buff seeds. 



Surely, what I have done in the garden, nature might 

 do in the wild, thus evolving a distinct species with three 

 varieties. 



In justice to my subject let me say that I am far from 

 maintaining that hybridism is furnishing in the genus 

 Viola all the new forms for natural selection to work 

 upon. Hybridism adds no new character to a group of 

 species; it simply recombines in multifarious ways char- 

 acters already existing. However numerous the patterns 

 that appear in the revolving kaleidoscope, their number 

 is limited; and if we looked long enough we should find 

 them substantially recurring from time to time. To get 

 a strictly new pattern, we should need to insert in the 

 apparatus a new fragment of colored glass. In the evolu- 

 tion of living organisms the new piece of colored glass is 

 what the biologists are considering under the name of 

 mutation. 



