PLEBEIUS. 168 



being all that is left, and, at the central entrance to the blister, the 

 upper membrane only. When older, it feeds on the upper- or under- 

 surface of a leaf, preferring the upper-surface, eating right through the 

 leaf except the skin of the opposite integument, which it does not 

 pierce until nearly full-grown, when Lyman notes (Can. Ent., 1902, 

 p. 127), that it eats the leaf entirely through, making holes some- 

 times away from, and sometimes at, the edges thereof. This is 

 much the habit of the larvae of Plebeius argus (aef/o?i), and P. 

 argyrognomon, on their respective foodplants. The larvae of the genus 

 Plebeius appear, in nature, to be almost entirely restricted to 

 leguminous plants for food. Structurally the larvae appear to be 

 fairly typical of the Lycaenids. They are provided of course with 

 the " honey-gland " on the middle of the 7th abdominal segment, 

 and the lateral caruncles or eversible tubes of the 8th abdominal 

 segment, and they have been, perhaps, rather more frequently 

 noted as being attended by ants than any other Lyeaenid larvae. 

 Indeed, one of the first extended records of commensalism between 

 ants and these caterpillars was recorded (Papilio, iv., pp. 92-93) 

 by Edwards in a western Nearctic species, Plebeius melissa. He 

 observes that, on June 9th, 1883, he introduced a small ant to a larva 

 of this species, and then explains, as we have already quoted (anted, 

 vol. viii., p. 32), in detail, how the ant obtains the fluid it sought. 

 The association of the larvae of Plebeius argus (aegon) with Lasins 

 niger is noted by Aurivillius (Ent. Tids., v., pp. 190,227), of the larvae 

 of P. argyrognomon with Formica cinerea by Thomann (Beobach. 

 Symb. L. argus. etc., 1901, pp. 1-40), whilst Powell mentions (Ent. 

 Bee, xviii., p. 214), that he found ants attending the larvae of Plebeius 

 argus (aegon), and Chapman notes (op. cit., p. 244) that he found the 

 best way of discovering the larvae of P. argyrognomon (argus) was to 

 follow up the ants that were busy with them. 



The Plebeiid pupa is very different from that of the Everids. Our 

 detailed description of P. argus (aegon) pupa (infra) gives the structural 

 characters of the group. Scudder describes (Butts. New Engl., p. 962) 

 the pupa of P. scudderii, long and slender, almost uniform green, with 

 a full plump abdomen, covered with a very delicate reticulation of 

 waved lines. 



The species of this genus usually show very marked sexual colour 

 dimorphism, the $ s violet or purplish -blue, the $ s brown ; the spots 

 beneath usually well marked with a well defined orange or yellow band 

 between the marginal and submarginal row of spots. But one of the 

 characteristic features of the underside is the usual presence of highly- 

 developed, metallic-blue, or metallic-green kernels to the marginal 

 spots of the hindwing ; a feature also found in other allied groups, but 

 rarely so well or extensively developed as in this. Gynandro- 

 morphism appears to be fairly prevalent in Plebeius argus (aegon), 

 many examples of which we are noting in our account of this species. 

 Ruhl records three of these forms in P. argyrognomon as follows: — 



(a) Right side <? , left $ ; taken near Amboise, at an elevation of 6000 feet 

 (Pal. Grosft-Schmett., p. 233). 



(b) Right side £ , left forewing ? , left hindwing brown with two narrow blue 

 streaks, and a little blue dusting at the base and at one point near the border ; whole 

 left underside $ ; body J ; taken near Prague {op. cit., p. 752). 



(c) A similar specimen, left side ? , right side J , partially only in colouring, but 

 much smaller in size, with much smaller fulvous spots than the left side (op. cit., p. 752). 



