AND THE MECHANIC ARTS. 521 



ous, but not at all conspicuous. In walking through, the woods 

 but few, if any, larvae would be noticed. However, the ground 

 was covered several inches deep in places with partly eaten 

 leaves and excrement, which might be plainly heard constantly 

 falling to the ground.. 



There seems no adequate explanation for this mode of distri- 

 bution, although it was essentially the same in every locality 

 studied. Possibly the central area may have been more or less 

 seriously affected in past years, the outer area representing the 

 extent of migration of the adults for one season. 



Concerning the food plant, with one or two exceptions, but 

 little preference is- shown between various hardwood trees. 

 Practically everything is taken with the exception of the "moose- 

 wood." There is, however, a preference for maple and beech, 

 while the white oak and yellow birch are left until the last. 



In feeding the larvae cling to the under side of the leaf, and, 

 beginning at the outer edge, eat out large crescent-shaped 

 areas. Frequently, two of these areas uniting will cause the 

 outer portion of the leaf to fall. Whether or not the younger 

 larvae feed in colonies is not known, but the more mature forms 

 do not appear to be gregarious in their habits. 



When about ready to pupate the larvae drop to the ground or 

 let themselves down by silken threads, then work their way un- 

 der the leaves to a depth of from two to six inches. It seems 

 quite probable that the larvae molt after leaving the trees, the 

 form usually found beneath the leaves being quite different 

 from the ones on the trees. 



Pupae and Pupation. After the larvae haVe- worked their way 

 well beneath the leaves preparations are begun to enter the pupa 

 stage. This is a very simple process as compared with the elab- 

 orate cocoons made by some of our moths. In this instance the 

 larva works out a simple, small cell, but little if any silk enter- 

 ing into the composition. However, several larvae kept in the 

 laboratory worked their way beneath the dirt and spun quite 

 firm cocoons, which were covered with particles of dirt held firm- 

 ly together by silk threads, while the interiors were lined with 

 silk and perfectly smooth. The length of time consumed in this 

 preparation is not known, but the transformation takes place 

 within a week or ten days after the larvae leave the trees. 



The pupa may be described as follows: Body stout and thick, 

 head somewhat pointed, mesoscutum firmly corrugated, or nearly 

 smooth, abdomen covered with slight depressions. Beside the 

 mesoscutum are from six to ten quadrangular, flattened, unpol- 



