EXPERIMENTAL EVOLUTION AMONGST PLANTS. 9 



In what, then, does an accidental horticultural origin differ from 

 any other origin ? Simply in the fact that one takes place under 

 the eye of man, and the other occurs somewhere else ! It is 

 impossible, at the present day, to make a definition of a species 

 which shall exclude many horticultural types, unless an arbitrary ex- 

 ception is made of them. The old definitions assumed that species 

 are special creative acts, and the method of origin is therefore stated 

 or implied in all of them. The definition itself, therefore, was 

 essentially a statement of the impossibility of evolution. We have 

 now revised our definitions so as to exclude the matter of origin, 

 and thereby allow free course to evolution studies ; and yet here 

 is a great class of natural objects which is practically eliminated 

 from our consideration because, unhappily, we know whence the 

 forms came ! Or, to state the case differently, these types cannot 

 be accepted as proofs of the transformation of species because we 

 know certainly that they are the result of transformation ! 



Now, just this state of things would be sure to occur if De 

 Varigny were to transform one species into another. People would 

 say that the new form is not really a species, because it is the 

 result of cultivation, domestication, and definite breeding by man. 

 He could never hope to secure more remarkable transformations 

 than have occurred a thousand times in the garden ; and his 

 scheme — so far as it applies to plants —is essentially that followed 

 by all good gardeners. Or, if the prejudices of critics respecting 

 the so-called artificial production of species could be overcome, 

 he could just as well draw his proofs of evolution from what has 

 already been done with cultivated plants and domesticated animals, 

 as from similar results which might arise in the future from his 

 independent efforts. I am not arguing against the scheme to 

 create a species before our eyes, but I am simply stating what has 

 been and is the insurmountable difficulty in just this line of 

 endeavour — the inability of the experimentor to satisfy some scien- 

 tific men that he has really produced a species; for it is a singular 

 thing that, whilst all biologists now agree in defining a species upon 

 its tangible and present characters, many of them nevertheless act 

 upon the old notion that a species must have its origin somewhere 

 beyond the domain of exact history. 



This notion that a species, to be a species, must have origi- 



