of certain lepidopterous larva and pupa. 43 



like angles, but are really bends. This effect is produced 

 in the larva of E. orbicidaria by thrusting out the pos- 

 terior pair of true legs, which are situated on the convex 

 side of the bend I have described. The spiral attitude 

 is also chiefly maintained during the long rests before 

 ecdysis, and is usually abandoned when the larva is 

 nearly fuh-grown, although tendencies towards the habit 

 are then occasionally observed. 



When the larvae are disturbed they resort to other 

 methods of protection ; first letting themselves down by 

 a thread, and, if further irritated, falling to the ground 

 and wriggling with great rapidity for some time. This 

 habit is also followed by E. jjendularia and E. omicronaria, 

 and it is common in many larvse. The rapid movements 

 seem obviously directed towards escape from insect- 

 parasites, but it might also be useful in working the 

 larva down among the roots of grass, &c. But the 

 remarkable thing about the two species of larvae men- 

 tioned above is that these bilaterally symmetrical animals 

 should sometimes assume an asymmetrical position when 

 at rest. The best instance of such attitudes that I have 

 seen was pointed out to me by Professor Westwood, i.e., 

 some of the Phasmidce, which, resembling branches with 

 lateral twigs, hold their limbs asymmetrically, thus 

 increasing the protective likeness. But I have not 

 before observed any departure from bilateral symmetry 

 in the various protective attitudes of lepidopterous larvae. 

 It seems very likely that this habit will be found during 

 young stages of other Geometer larvae, unless protected 

 by a green colour, or in some other special way. It is 

 possible, however, that the habit is confined to the 

 Eplnjridce, or specially manifested by them, as this 

 family in some respects stands alone among Geometi-cs. 

 Since writing the above I have been told of other in- 

 stances, and have also found some myself. In all cases 

 the asymmetrical attitude appears in Geometer larvae 

 which feed on plants or parts of plants in which the usual 

 protective attitude would not avail. 



The protective attitude of Notodonta ziczac as an 

 INSTANCE OF SIMULATED ANGULARITY. — Truly angular atti- 

 tudes are entirely impossible during prolonged rest for 

 such soft-bodied, semicylindrical, organisms as caterpillars 

 (with the necessity for the constant circulation of an- and 

 blood through their respective systems). Therefore the 

 appearance of a very angular attitude is a protection, as 



