112 



which turns brown. A few days after the larvae left their molting 

 cocoons, I observed a few quite large molting cocoons. Suspect- 

 ing that they might be cocoons in which the larvae molted the 

 second time, as the larvae within them were greenish and not 

 yellow, I placed some of the yellow larvae, which had just left 

 their mines, in a bottle with fresh apple leaves. After making 

 molting coooons and molting, they fed for two and a half or 

 three days and then made a second cocoon which differed from 

 the first only in size, being about T s i n - m diameter. These 

 second cocoons are made and entered the same way as the first 

 ones, and the larva remains on its back inside them for two days, 

 when it forces its way out, leaving its second cast-off skin in the 

 cocoon, and goes on feeding as before. The only subsequent 

 molt is when the larva transforms to a pupa." 



The larval state lasts about three weeks. When full grown the 

 larvae migrate from the leaves to the lower side of the twigs where 

 they spin their cocoons. When the trees are badly infested these 

 cocoons occur in great numbers massed side by side so as to nearly 

 completely cover the lower surface of tli2 twigs. 



The duration of the pupa state of the spring brood is only about 

 a week, the moths beginning to emerge soon after the middle of 

 July. The subsequent history of the species has not yet been 

 traced. Hence we do not know whether this spring brood is fol- 

 lowed by more than one brood or not. The cocoons of the autumn 

 brood were found in the latter part of September. As the spring 

 brood requires only a month to pass through the larval and pupal 

 stages, it seems probable tHat between the middle of July and the 

 latter part of September there would be sufficient time for the de- 

 velopment of two broods, thus making the species three brooded 

 in this latitude. 



While the duration of the pupa state of the spring brood is only 

 a single week, that of the autumn brood is much longer, lasting 

 from the middle or the latter part of September to the following 

 April or May, between seven or eight months. 



Method of Treatment of the Pest. — As this insect is rare at 

 Ithaca, we have been able as yet to \xy only a few experiments in 

 the destruction of it. Infested twigs sent us by several correspon- 

 dents have furnished the material for these experiments. 



The fact that during the winter the cocoons of all of the indi- 

 viduals infesting a tree can be found exposed, chiefly on the 

 lower surface of the twigs, naturally suggests this point of attack. 



.A liberal pruning and burning of the infested twigs would do 



