NOTE ON EUJNLEUS. HI 



and the chrysalis a girth around the middle, there is no need of further discussing any 

 supposed Satyrine affinities. 



Their reference to the Lycaeninae rather than to the Lemoniinae is easily defended, 

 though naturally not so conspicuous; for the fore legs of the male are amply provided 

 beneath with spines, and the end of the single exarticulate joint is armed with a pair 

 of stout curved spines and densely fringed with a pile of short hairs, appearing to 

 represent the paronychia ; while the subcostal nervure of the fore wings has only two 

 superior nervules. Our knowledge of the early stages of Lemoniina3 is exceedingly 

 meagre, so that we can here draw no decided conclusions. There is, indeed, no greater 

 desideratum in the study of butterflies than a knowledge of the transformations of the 

 principal genera of this subfamily. But from what little we do know, we are aware 

 that the onisciform character of the larva, almost universal in LycseninaB* is hardly 

 noticeable in the caterpillars of Lemoniinae ; and in this respect the Eumaeidi, especially 

 previous to the last stage, certainly show Lemoniid affinities, both in the form of the 

 body, the comparatively large head, and the nature of the clothing ; the last, not alto- 

 gether dissimilar to that of Feniseca, consists of fleshy thorny tubercles in longitudinal 

 rows. The chrysalis, speaking again with the same limitations, appears to accord 

 with the Lycaeninae ; for in the Lycaeninae, just as the caterpillar is strikingly different 

 from those of all other butterflies by its onisciform habit, so the chrysalis differs from 

 all in having the head invisible from above, being bent over to form a part of the 

 ventral surface ; generally the same is also true of the terminal segment of the abdo- 

 men. Though I have been able to examine autoptically only one or two chrysalids of 

 Lemoniinae, yet the published illustrations enable one to judge that in these points they 

 probably do not differ, as those I have seen do not, from the great body of other 

 butterflies. Now the chrysalis of Eumwus atala agrees in these points with the 

 Lycaeninae, although the terminal segment of the abdomen is more exposed than usual 

 with them. 



While, then, there would seem to be little doubt that the Eumaeidi should fall 

 between the Lemoniinae and the Lycaeninae, there appears to be small reason to place 

 it as a separate group equivalent to either of them, and still less to place it within the 

 limits of the Lemoniinae. It should rather be placed, at least until our knowledge of 

 the transformations of Lemoniinae is vastly improved, as a group of LycaeninEe, in 

 nearest position to the Lemoniinae. 



We may further consider the question, What are its relations to the other members 

 of the subfamily Lycaeninae ] At least as found in temperate regions, this subfamily is 



* Mr. W. H. Edwards has recently shown that the earliest stages of Feniseca lack the onisciform character 

 of later stages. 



