IN THE GENUS BASILARCHIA 29 



eaten edge. It should be noted that it is so loosely 

 attached, th^ bits of leaf at all possible angles, that 

 it is moved by the least breath. Meanwhile the cat- 

 erpillar has been growing larger and more conspic- 

 uous and is thus in greater peril from its enemies. 

 There are two possible services that this odd packet 

 may render. A spider wandering over the leaf and 

 observing its motion may seize it and thinking it 

 has a prize, hurry away with it and leave its archi- 

 tect unharmed. This seems to me rather a strained 

 suggestion, for a wandering spider would probably 

 proceed to investigate it on the spot. Another 

 explanation seems more probable. It should be 

 remembered that the leaves preferred by these 

 creatures as food are mostly such as are easily 

 shaken by the wind, and as the caterpillar moves 

 with the leaf and with all the surrounding leaves 

 (in a continual fluttering in the case of the trem- 

 bling aspen, and to a less degree in the other food- 

 plants), this of itseK is a protection to it, as it 

 would more readily escape observation as an object 

 distinct from the leaves, all being in motion to- 

 gether ; but on the more stable leaves, like the 

 willow and especially the Rosaceae and the oaks, 

 the motion in a feeble wind would not be sufficient 

 to be serviceable, and here at least the packet 



