178 PROCEEDINGS OF THE ACADEMY OF [Jail., 



9. The Evolution of Limenitis (Basilarchia) archippus from 

 AN Ancestor with a Pattern like that of 



L. (B.) ARTHEMIS. 



The origin of archippus, suggested in the title of this section, is 

 due to Scudder (6, 277-8, 714). All I have done is to support the 

 published views of this distinguished naturalist by making a careful 

 analysis of the markings of archippus and arthemis, by this means 

 demonstrating that the details of the mimetic pattern are accounted 

 for on his hypothesis! I am sorry to find that neither Scudder's 

 hypothesis nor the results of my analysis carry any conviction to 

 Dr. Skinner, who uses the following words: ^'Arthemis and weide- 

 meyeri [with a very similar pattern] have flourished prosperously in 

 the .struggle for existence, and it is difficult to understand why 

 archippus should be so specially favored. The statements attempt- 

 ing to prove the evolution of archippus from an ancestral form 

 (arthemis) seem to me very inconclusive" (33, p. 127). Dr. Skinner 

 makes no alternative suggestion as to the origin of the mimetic species. 

 The doctrine of evolution — for it is hardly necessary to discuss the 

 ancient belief which would assume that archippus was originally 

 created in its present form — leaves us only two hypotheses. Either 

 archippus was evolved from some form of Limenitis which has 

 entirely disappeared or from one which is more or less closely repre- 

 sented by a species still in existence. The former alternative aban- 

 dons the problem as insoluble, and abandoned it must be if there is 

 no sufficient evidence that the ancestor can be reconstructed from any 

 existing form. I agree with Scudder in preferring the counsel of 

 hope to the counsel of despair. L. (B.) arthemis and weidemeyeri 

 present us with an ancestral pattern wide-spread in the genus and 

 found not only in North America, but also with little change in the 

 Old World section of the temperate circumpolar zone. Archippus 

 is so closely related to arthemis that the larval and pupal stages are 

 almost identical, and although the imaginal patterns are so different, 

 Scudder indicated, and I have attempted to trace in detail, the 

 manner in which one pattern may be derived from the other. I 

 really think that if Dr. •Skinner, with specimens of archippus and 

 arthemis before him, will verify the details of the account in my 

 earlier paper (31, pp. 456-459), he will find that many minute 

 features on the wings of the mimetic species are interpreted and 

 correlated in a satisfactory manner. And a hypothesis that interprets 

 stands, until replaced by another that interprets better. 



With regard to Skinner's inference that inasmuch as arthemis 



