154 NATURE SKETCHES IN TEMPERATE AMERICA 



which is perhaps more effective than the whip organs when its 

 deadly power is considered. On the underside of the neck of 

 our species there is a sacculated depression, opening externally 

 by a transverse slit which, in a moment of excitement, exposes 

 the opening of two little tubes. These tubes eject a yellowish 

 spray, which is evidently poisonous. This fluid, in species of 

 similar habits found in England, was examined by Poulton and 

 found to be formic acid. It is (juite probable that even birds 

 would not relish such a bombardment and the Hymenoptera 

 attacking the caterpillar appear to have a poor chance of escap- 

 ing such a deadly shower bath. 



I do not recall any record of an observer having seen this 

 device in actual operation against an enemy in our American 

 species. I once saw a small winged i)arasite run the gauntlet, 

 taking the larva unawares, and in spite of this protective device 

 it succeeded in depositing its egg upon Cerura s skin. 



The ])uss larva is light green, the head is dark j)urpli.sh, and 

 a triangular marking aj)pears beiiind the head. About eight 

 or nine small spots decorate the sides of its body. The dorsal 

 area above is jjrovided with a peculiar dark marking of lilac 

 brown, commencing forward on the fourth segment and 

 extending backward to the tail. The widened portion of the 

 forward marking is more or less sprinkled with a lighter color. 

 Underneath the body, just behind the back legs, is a pair of 

 large spots. The lilac-brown patches of color on the greenish 

 yellow body of this caterpillar resemble the leaves perfectly, 

 the former color simulating the seared and withered brown 

 edge of the willow leaves. 



After two days of feeding on the willow leaves indoors the 

 larva became restless and refused to cat, and then a change 

 came over it. Instead of the pure green, the body became 

 somewhat brownish, and the appendages seemed to lose their 

 function, becoming reduced in size to mere .shrivelled filaments. 

 It travelled over the stems and down on the floor of the breeding 

 vivarium, back and forth, when finally it .settled on a piece of 

 newspaper. It then began to .spin a rather loo.se, gauzy web 

 forward around its head, which it attached here and there to 

 the paper on the ground. The weaving continued farther and 

 farther backward to about the middle of the body, when the 



