Eighth Report of the Si ate Entomologist. 137 



Figure s 2ff and 4, Those of Bucculatrix poTnifoliella are shown, for 

 comparison at a and h in Figure 5. 



The cocoons have only been observed upon the leaves, but as they 

 are seemingly too few for such an abundant larval presence, it is prob- 

 able that many of them are made on leaves that have fallen to the 

 ground, or possibly, on or beneath the bark of the trunk, or even on 

 adjoining trees and shrubs, as some of the species of Bucculatrix are 

 known to have a fondness for traveling. None were found on the 

 small twigs that have been sent me with the leaves. 



The construction of the cocoon is an interesting operation to watch. 

 It differs in method from that with which we are familiar in many of 

 the Bombycid and other moths "where the caterpillar spins its thread 

 around it, and gradually hides itself from view. In its earl}^ life the 

 birch-leaf caterpillar had been an outside builder, as will be noticed 

 hereafter, and its habit is not abandoned in this its mature stage. 



The cocoon is apparently commenced by the caterpillar marking out 

 with a few threads laid upon the surface of the leaf the outline of one 

 end. From this, as the base of operations, the ribbed walls are built 

 up in the following manner: Its thread is fastened at one end of the 

 basal line, and then carried over in the required curve and at a slight 

 elevation to the other side where it is fastened to the base. The 

 thread continues to be carried from one side to the other, but 

 not in single curves made by one sweep of the larval head. If 

 closely noticed through a lens (as these observations were made), 

 the head will be seen moving in a series of concave curves, rest- 

 ing momentarily at the four or five points where the ribs are being 

 formed. Thread after thread is thus rapidly carried over; the base is 

 gradually widened as it approaches the middle part; the short curved 

 sweeps of the larval head build in the ribs and the depressed portion 

 between them, while each thread is applied, and apparently adheres 

 closel}^, to the preceding. This is continued without cessation, and as 

 the building progresses the caterpillar gradually moves backward before 

 its advance. Whether its eyes or the length of its head serves as the 

 only measure for the diameler of the cocoon is not known. Possibly 

 its body may be used as a guide after the manner of the caterpillar of 

 a Limenitis butterll}^ which repeatedly during the construction of its 

 hibernaculutn creeps within it as if to test the dimensions of the silken- 

 lined leaf-tube within which, with entrance left unclosed, it is destined 

 to pass the winter. 



When the cocoon is about three-fourths completed, a new procedure 

 ensues: the caterpillar arrests its work, turns itself about, and moves 

 18 



