SELECTION OF MIMETIC LIKENESS 201 



tainly wrong. The only probable hypothesis is 

 that sharpsighted enemies, without analysing the 

 markings, recognize differences in degrees of 

 likeness, and that the selective pressure exercised 

 by them is influenced by the recognition. 



A great deal of attention is rightly directed at 

 the present day to the value of experiment, and 

 indeed it is impossible to over-estimate its impor- 

 tance. But while human performance is of the 

 deepest interest for the solution of mysteries 

 innumerable, of more profound significance still, 

 for the comprehension of the method of evolution, 

 is the vast performance of Nature herself. ^ Be- 

 cause of the bright promise it holds for the under- 

 standing of Nature's experiments, I have brought 

 before you the subject of Mimicry in North 

 American butterflies. 



In the introductory words I spoke of the relation- 

 ship of my subject to the teachings of Darwin, and 

 now I am anxious to connect this address by 

 a closer link to the personahty of the illustrious 

 naturaUst. With the kind consent of Mr. Francis 

 Darwin, I am able to achieve this object by print- 

 ing, for the first time, a letter, recently discovered 

 in the archives of the Hope Department at Oxford, 

 written by Darwin to the Founder in 1837. It is 

 concerned with the insect material collected on 



' See Carl H. Eigenmann in Fifty Years of Darwinism,, New 

 York (1909), 208. 



