K) BRITISH BUTTERFLIES. 



been observed. Comstock states that the larvae of the almost sub- 

 tropical species, Eudamus proteus, feeds not only on a variety of 

 leguminous plants, but also does damage in Florida to the crops of 

 turnips and cabbage. The larva of Epargyreus tityrus feeds on so 

 large a number of leguminous plants, that Scudder thinks it would 

 possibly refuse nothing belonging to this natural order. The larva of 

 the allied Achalarus lycidas also chooses Leguminosae for its foodplants 

 (although other plants have been very dubiously suggested as possible 

 foods). The larvae of Thorgbes bathyllus and T. pylades are leguminous 

 feeders, both species appearing to be confined entirely to plants of this 

 natural order. 



Whilst, therefore, the food-habit appears to be fairly well-defined 

 as between the larger groups, the feeding-period seems to be varied 

 and uncertain. We cannot class the larvae as a whole as summer- 

 feeding, nor as over-wintering (feeding in autumn and again in spring). 

 Even a detailed consideration of the feeding-period in our few British 

 species shows considerable variation in this direction, and suggests a 

 division roughly into tribal sections rather than any general super- 

 family character, a division apparently to be noticed in the species of 

 a tribe scattered over a large geographical area, and subjected to great 

 differences in climatic conditions, rather than in the habits of more 

 distantly related species, occupying similar areas under similar climatic 

 conditions, or even the same ground. This generalisation, however, 

 cannot be pushed far, as illustrated by the differences in the allied 

 Adojmea lineola and A. flava, and still more markedly in Urbicola 

 comma and Augiades sylvanus. 



To illustrate the variability existing in this respect it may be 

 noted that the Thymelicines are essentially spring-feeding larvae. Of 

 our three British species, Adopaea lineola hybernates as egg (the fully- 

 formed larva contained within the egg all the winter), Thymelicus 

 acteon almost certainly has the same habit, whilst Adopaea riant, 

 although hatching in the late summer (teste Buckler), does very little 

 feeding before hybernation, and commences its winter rest when not 

 more than 2mm. in length. The actual feeding-period of all these 

 larvae, however, practically lasts from March to June, whilst their 

 habits of living in tubes formed of grass-blades, of resting by 

 day and feeding chiefly by night, are almost identical. Our two 

 Urbicolid species, however, are very different. Urbicola comma, like 

 some of the Thymelicines, hybernates as a fully-formed larva inside 

 the egg, hatches in early spring, feeds like them until June (or early 

 July), makes a tubular dwelling of grass- blades (but of larger size), 

 and pupates and emerges rather later than they. The closely- allied 

 Augiades sylvanus, however, commences its larval life in July and 

 August, feeds on into its fourth instar, and does not hybernate until it 

 is some 13mm. in length. By April 20th, 1903, the larva of Erynnis 

 lavatJierae (a species belonging to the opposite branch of the Urbicolid 

 stirps) was found in the Val Verzasca, in its penultimate skin, feeding 

 in a leafy nest on Stachys recta (Ent. llec, xv., p. 298), so that u 

 would seem to have a similar larval feeding- period to that of A. 

 sylvanus. The knowledge derived from the feeding-periods of Niso- 

 niades tagcs and Hes/wria malrae (on the same side of the Urbicolid 

 stirps as /','. laratherae) made this observation unexpected, for both 

 these species have strictly summer- feeding larvae, the larvae of both 



