RURALIDES. 299 



abdominal appendages of the $ , shared by both the subfamilies, and 

 we have a totality of characters, drawn from all stages, held by these 

 two groups in common, and in distinction from others, which cannot 

 be exceeded by any other combination of subfamilies in a homogeneous 

 whole." Kirby, however (Handbook, etc., p. 7), commenting on this, 

 says that " there are many differences between the two groups thus 

 united by Scudder, and even in pattern and general appearance they 

 are so dissimilar, that there are very few species belonging to either of 

 them which would be likely to be mistaken for the other, even at the 

 first glance, by anyone who was fairly well acquainted with the general 

 appearance of butterflies." He adds that "the Lemoniids and 

 Lycaenids may be taken as, in some measure, representative of separate 

 faunas, for the Lemoniids, with the exception of a few somewhat 

 aberrant Old World genera, are almost entirely confined to tropical 

 America, whilst the Lycaenids are almost entirely an Old World 

 group, if we except the tropical American species allied to Thecla." 



Chapman, dealing with the phylogeny of the Ruralides from the 

 standpoint of pupal structure, conies, however, to exactly the same 

 conclusion as Scudder. He says that, amongst butterfly pupae, above 

 the Urbicolids, there appear to be two great divisions, one containing 

 the Papilionides (with their Pierid and Nymphalid derivatives), the 

 other containing the Ruralides (Ruralidae or Lycaenidae and Erycinidae), 

 There has been so much specialisation in various (and often in parallel) 

 directions in these two groups, that exceptional pupae may seem to 

 travel far from that typical of the group to which they belong. Still the 

 broad differences are nearly always present, the pupa of the Ruralid 

 group is distinguishable from that of the Papilionid by the short, 

 squat, rounded form, and by the head being quite on the ventral 

 aspect of the pupa, as if the round squat form had been gained by 

 pushing both extremities underneath ; they are further distinguishable 

 in that the Papilionids have horns and processes that are part of the 

 pupa itself, but rarely, if ever, any hairs, whilst the Ruralid pupa has 

 always hairs more or less in evidence, though it may occasionally have 

 pupal. processes also, like those of Papilionids. When we come to 

 differentiate the Ruralides into their two great divisions of Ruralidae 

 (Lycaenidae) and Erycinidae, there is no character that seems to be 

 constant. The Erycinid pupa has. the 2nd leg well developed upwards, 

 so as to reach the face (eye) and separate the 1st leg from the antenna, 

 whilst in that of the Lycgenids the 1st leg is best developed upwards, 

 its base stretching right across the face and touching the antenna, so 

 that the 2nd leg does not reach upwards to the face. As a rule, the 

 Lycaenid pupa is high, whilst, in many Erycinids, the pupa is very 

 flattened, making them often show, with exaggerated conspicuousness, 

 the width across the 3rd and 4th abdominal segments, which 

 distinguishes many pupae of this group from those of the Papilionid 

 stirps. The pupa of Hamearis (Xemeobius) lucina, our only British 

 Erycinid, might, however, so far as this character is concerned, be a 

 Lycaenid. All the Lycaenid pupae, so far examined by me, have been 

 completely solid, without movement ; it is probable, however, that in 

 some of the less specialised Lycaenid subfamilies, forms retaining 

 movement occur, since, in the Erycinidae, the Euselasiinae have the 

 normal two movable segments, the Lemoniinae have one, whilst in the 

 Hamearinae (Nemeobiinae) , as exemplified by Hamearis (Nemeobius) 

 lucina, the pupa is solid, as in the Lycaanids. 



