ATT AC IDES. 279 



larva of Samia cynthia, whilst Packard considers Callosamia to be, in its 

 larval characters, the last and most specialised genus of a series beginning 

 with Saturnia (carpi ni) and including Platy samia and Samia." Packard 

 thinks that the absence of the tubercles in Callosamia may be a case 

 of arboreal selection, the tulip-tree, its only known foodplant, having but 

 few species of larvae feeding on it. 



Mention may be made that certain Attacid larvae, Philosomia 

 lunula, &c, become covered with a white waxy-powder when half- 

 grown. No suggestion has yet been made as to the use of this powder. 



The Citheroniid larvae, Citheronia, Eacles, &c, are said to change 

 their hue when fullfed, and wander restlessly when seeking for a 

 pupation-place, peculiarities which they share in common w r ith certain 

 Sphingid, Dimorphid, and other larvae. As bearing on this point, 

 it may be observed further that the green, freshly-formed pupa of 

 Aglia tau exhibits very distinctly all the markings of the larva, in- 

 cluding the subspiracular line, a prominent larval feature, the line 

 being continued along each side of the anal flap to its extreme 

 apex. Poulton utilised these markings for the purpose of homologis- 

 ing the pupal segments and structures, the positions of the markings 

 affording valuable confirmatory evidence of the identification of the 

 anal flap of the larva with the terminal spine of the pupa, a subject 

 which is treated by this author at considerable length (Morph. Lep. 

 Pupa, p. 196). 



The process of exuviation of the Attacid larvae is very similar 

 to that obtaining in other larvae. Most, if not all, of the larvae spin 

 a silken cushion, into which they fix the crochets of the prolegs 

 before moulting begins. That spun by Telea polyphemus has been 

 described as a thin carpet of silk, and that of Platysamia cecropia is 

 somewhat similar. The larva usually rests immovably for several 

 hours before exuviation begins ; the membrane of the prothorax 

 swollen, the head appearing very small in comparison. In moulting, 

 the larva leaves behind it only a small mass of crumpled skin, the 

 whole of the cast skin being collected together into a very small space. 

 In many Attacids the cast skin is eaten by the larva. See also anted, 

 vol ii., p. 6. 



The cocoons of the Attacids (sens, strict.) are possibly among 

 the most interesting of those made by any lepidopterous larvae. 

 Many of them are used commercially for their silk, and for this 

 purpose the larvae are kept in a domesticated or semidomesticated 

 state, but the silk is almost always coarser than that of Bombyx 

 mori, and cannot be wound, so that it is usually carded or spun like 

 cotton. Gentry notes a variation in the colour of the cocoons spun 

 by larvae of S. cecropia that have fed upon various trees, e.g. (i) 

 on red currant, the silk is said to be reddish-brown, (2) light-brown 

 tending to grey on cherry, plum and rose, (3) grey-brown on Spiraea, 

 Symphoricarpos and Primus serotina. We doubt whether this would stand 

 the test of careful experiment. The cocoons of the Hemileucids are 

 usually spun on or under the surface of the ground, and the 

 Citheroniids are satisfied with an earthen cocoon. Barrett notes 

 (Can. Ent., xxxii., p. 235) that the cocoons of several species 

 of Automeris have a quasi-hinged grating in the front portion which 

 only opens outwardly; the front is closed, but with very weak 

 (brittle) silk. This strongly suggests Saturnia. Grote writes (Can. Ent: } 



