246 



THE AMERICAN NATURALIST [Vol. XLII 



made to interchange, although they possess no evident 

 morphological differences. This result depends, as I 

 assume and as seems to be pretty well authenticated 

 by all the researches, upon a question of nutrition. It 

 does not depend upon any peculiar structure of Aster, 

 Solidago or Erigeron; but here we have three genera 

 which furnish different kinds of nutritive matter, or in 

 some way differently affect the nutrition of the fungus, 

 so that a fungus which lias adapted itself to live upon 

 one of these generic groups of plants will not live upon, 

 or can not live upon, the other. It is a very nice adjust- 

 ment to a particular kind of nutriment; I think we are 

 quite well assured of that fact. 



Now, the question arises, are these three species? If 

 you choose to look up the literature, you will see that 

 the speaker has described them as three species under 

 three independent names. That does not mean that he 

 believes they are truly three species, but that it was a 

 convenient way of designating the three physiologically 

 different kinds until a time when the matter could be 

 more fully considered. Xow, are these three species? 

 Well, to confess to you, not to say it too loudly, I do not 

 believe they are. They are not three species in the gen- 

 erally accepted sense because they can not be distin- 

 guished morphologically. What then is, or should be, 

 our conception of a species ! Is it something to be dis- 

 tinguished physiologically or morphologically, or in both 

 ways? This is not a hypothetical question, because the 

 literature of the subject that I represent is very full of 

 descriptions of species based entirely or partially upon 

 physiological data, and I am assured that we are not 

 alone in this method of describing species. 



If I understand aright, the ornithologists, at least 

 some of them, are willing to assert that species may 

 differ by characters which can not be described by words 



