FAMILY HABITS IN BUTTERFLY LARV.E THE LIMENITIDS. 35 



leaves used have been eaten away, and only the sides need shaping, 

 but, if the residue be insufficient, or for any reason does not answer 

 the purpose, the caterpillar moves to another leaf, and begins the 

 necessary cutting to construct a suitable hybernaculnm. 



Probably still more peculiar are the feeding- and resting-habits of 

 the larvae when young. Gartner observes (Stett. Ent. Ztg., xxi., pp. 

 296-7) that the larva of Neptis aceris, as soon as it leaves the egg, 

 begins to eat through a leaf of Orobus vermis, laterally, as far as the 

 midrib, which, together with the tip of the leaf, it leaves untouched ; 

 at the next meal it attacks the opposite side of the leaf, and so on 

 alternately, gradually approaching the base ; after each meal it crawls 

 over the untouched midrib, to the very tip of the leaf, where it takes 

 a siesta, its head directed outwards, and thus it continues through 

 three stages, when, after some modification in its habits, hybernation 

 takes place. This may now be compared with Scudder's description of 

 the larval habits of Basiiarchia arthemis in its earlier stages. He 

 writes (Butts. New Engl., i., p. 257), " The egg of B. arthemis is 

 placed near the tip of a leaf, and, as soon as hatched, the larva eats 

 away the leaf on both sides of the midrib, and it then rests on the 

 stripped portion of the rib." He then quotes Edwards as saying (Pap., ii.) 

 that the larva of Basiiarchia archippus (disippus) has, in all respects, 

 habits similar to those of that of B. arthemis. The larva of the former 

 no sooner lays bare the rib than it is coated and wound with silk, and 

 to the extremity are fixed grains of larval excrement, only two or three 

 at first, placed one after the other in line ; these are bound together 

 and to the rib, and, being small as grains of rifle-powder, they form a 

 continuation of about the same dimensions as the rest of the perch, 

 and seem effectual to prevent curling as the rib dries. As the larva 

 grows the process is continued, until the artificial portion measures 

 •5in. or -6in., and makes a stout irregular cylinder, the entire perch 

 measuring about l-5ins. It is constantly strengthened by additions of 

 silk, the larva almost invariably, as it goes back and forth, to and 

 from its feeding-ground, adding threads and patching the weak places. 

 On this perch the young larva, i.e., before hybernation, always rests, 

 going therefrom to the leaf for food at short intervals. It occupies the 

 middle of the perch, when resting, and its usual attitude is a twist, 

 the ventral legs clasping, but the anterior half of the body bent down 

 by the side and somewhat under, the perch. If two larvae of B. 

 arthemis be born on one leaf, one always takes possession of the mid- 

 rib, whilst the other constructs an artificial perch by the side of the 

 leaf. This perch is irregularly cylindrical, and composed of frass and 

 small bits of leaf woven together and covered with greyish silk, and is, 

 at first, nearly -25ins. long and about -02ins. in diameter. As the larva 

 increases in size so this special resting-place is made larger. Edwards 

 also notes (Pap., ii.) that, if two larvae of B. archippus (disippus) are placed 

 on the same leaf, one always takes possession of the extremity of the 

 midrib, often with something of a contest and knocking of heads 

 together, but the other will presently be found on one edge, excavating 

 the leaf on either side of a narrow slip, which is to constitute the base 

 of its perch. This is bound with silk, lengthened with frass in exactly 

 the same way as when the midrib is used, and serves every purpose 

 that the latter does to the larva in possession thereof. Scudder says 

 (Butts. New Engl., p. 274) that the larva of B. archippus neve? remains 



