THE COLORATION OF BUTTERFLY LARVAE. 63 



towards making up the general appearance of the creature, and, by 

 their aid, combined with their colours and patterns, the separation of 



species may probably in all cases be tolerably sure The vast 



majority of butterfly caterpillars are green, though but exceedingly 

 few of them, if indeed any, are uniformly green throughout. Most of 

 them are longitudinally striped, either with lighter and darker shades 

 of green, or with yellow or various shades of brown. Many of them 

 have the additional adornment of points of brighter or darker colours, 

 which are almost invariably confined to the little papillae with which 

 the body is almost always studded. Such are the vast majority of the 

 Satyrids, the Pierids, the Hesperiids, and the Libytheids. These 

 longitudinal stripes are by far more common than elsewhere in the 

 middle of the back, where they mark the course of the dorsal vessel, 

 on the lower portion of the sides, where they mark the alignment 

 of the spiracles, and midway, or about midway, between these two ; 

 when most variegated, the stripes are multiplied, especially upon the 

 upper half of the body, and often show a greater degree of intensity at 

 the extreme anterior, or extreme posterior, end of each segment. 

 Other green caterpillars are marked with oblique stripes, which 

 generally part from the darker mediodorsal line at about such an 

 angle, as Lubbock remarks, as the ribs of a leaf part from the main 

 stem. These oblique stripes almost invariably run down the sides 

 from in front backward, generally cross two or three segments, and 

 may or may not join a stigmatal line below, or the dorsal line above." 

 In a broad way these remarks of Scudder are true, and, even in the 

 Palaearctic Satyrids, where the larvae of allied species sometimes run 

 very close indeed, the number, direction, and general character of the 

 lines will usually distinguish the species. The character of the 

 markings of the Satyrid, Pierid, Hesperiid, and Libytheid larvae, 

 although consisting of longitudinal lines of a different shade from that 

 of the ground-colour, is essentially different in each of these groups, and 

 the position and arrangement of the lines, apart from the mediodorsal 

 line, are generally quite sui generis within the limits of each separate 

 family. This is particularly the case in the Pierid and Libytheid 

 larvae. Those of the Satyrids and Urbicolids (Hesperiids), being 

 largely grass-feeders, present much more marked general similarities 

 than do the others to these, or to each other. The grass-feeding 

 larvae, too, also present a markedly large percentage of dimorphic 

 forms — green and pale brown — whilst many are only brown (green 

 being an unknown or exceedingly rare colour aberration in the larvae 

 of these species). Almost all these groups present excellent forms of 

 cryptic coloration. Lubbock long since pointed out that longitudinal 

 stripes are very common markings, and are most common, and, indeed, 

 almost universal, upon such larvae as feed upon grasses and other 

 elongated forms of vegetation, while they are comparatively rare upon 

 such as feed upon broad-leaved plants, a general statement, particu- 

 larly verified in the larvae of the Satyrids and Urbicolids (sens, rest.), 

 and, to a less extent, in those of certain Pierids, e.g., Euchlo'e, Anthocharis, 

 etc., where the resemblance is to the long seedpods of the foodplants, 

 rather than to the leaves, at least in the later stages. Of this, 

 Scudder notes that " the green colour of all the North American 

 Rhodocerids and Pierids, and notably of Eurema lisa, which feed upon 

 broad-leaved plants and lie exposed upon the surface beside the midrib 



