1909] Mimicry in the Butterflies of North America 211 



The resemblances above described suggested the investigation 

 and comparison of structural characters in order still further to 

 test the relationship between these Old and New World Danaines, 

 and also the validity of the genera created by Moored Such a 

 comparison had already been partially made by Rothschild and 

 Jordan, who in 1903 published the conclusion that Limnas and 

 Tasitia cannot be generically separated.^ I therefore wrote to 

 my friend Dr. Jordan, asking if he would kindly extend his survey 

 over all the four so-called genera. He found that in Salatiira 

 genittia and Anosia plexippus, having larvae with two pairs of 

 filaments,^ the male genitalia are of the same type; while in 

 Limnas chrysippiis and Tasitia Berenice, hax'ing larvae wdth three 

 pairs of filaments, these genitalia are of a second type. The final 

 opinion of this distinguished authority on the relationships 

 between the Rhopalocera, was given in the following words i^" 



"It appears to be certain that Anosia plexippus does not 

 stand apart from the others. Therefore, if Tasitia berenice, 

 Limnas chrysippiis and Salatiira genutia are placed in one genus," 

 plexippus, also must be included. I do not think you need 

 hesitate thus to simplify the classification of these insects." 



I have no hesitation in accepting this advice, and in fusing all 

 the four genera created by Moore into the single genus Danaida. 

 Within this genus it has been made evident that the group of 

 forms ranged around Danaida plexippus is the New World repre- 

 sentative and close ally of the group of D. genutia; while that of 

 D. berenice is similarly representative of the group of D. chrysippiis 

 It is interesting to note that both the American Danaidas have 

 . become much larger than the corresponding Old World species, 



' Proc. Zool. Soc. Lond., 1883, 201. 



^Nov. Zool. vol. X. Dec, 1903, 502. 



' Dr. Jordan was at first inclined to think that Anosia plexippus should be 

 separated generically, basing his conclusion in part on the larval characters 

 {Trans. Ent. Soc. Lond., 1908, 450). A more extended review of the Tring 

 material pointed in the opposite direction, and Dr. Jordan wrote on December 

 10, 1008, as follows: — 'I find from our specimens (of preserved larvae) that — 



(1) in Euploca (in the wide sense) there are 4 pairs of filaments, or 

 three (the 3rd being absent), or two (the 3rd and 4th being absent). 



(2) In Danaidae, inch of Anosia & Limnas, there are 3 pairs (the 3rd of 

 the 4 pairs of Euploea being absent) , or 2 pairs (the 2nd and 3rd being absent) . 

 I find that, for instance, genutia and purpurata have 2 pairs only, like 

 plexippus. The larva therefore does not furnish any^^argument for sep- 

 arating plexippus as a genus.' 



'" In a letter to the author, dated December 15, 1908. 



'' Dr. Jordan's opinion that these three genera should be united is quoted 

 in Trans. Ent. Soc. Lond., 1908, 450. 



