PARASITISM TN INSECT CONTROL. 105 



wandering characteristics of those of the gipsy moth, but rather the 

 opposite, find themselves crowded in excessive numbers upon a 

 limited variety of shrubs and trees; complete defoliation of these 

 comparatively few host plants quickly follows, and weather condi- 

 tions being favorable to the development of disease, wholesale 

 destruction is all that intervenes between an unnatural migration or 

 starvation. Such reduction is followed by a period of years during 

 which the parasites check but do not overcome the tendency to 

 increase, and it is only a little while before the process is repeated. 



There were, in certain localities in eastern Massachusetts in the 

 summer of 1910, continuous strips of roadside grown up to a variety of 

 trees and shrubs, the most of which were defoliated by tent cater- 

 pillars, all of which had hatched from eggs deposited upon the 

 occasional wild-cherry tree which was present. Several such strips 

 were visited at about the time when the caterpillars elsewhere were 

 beginning to pupate, and not a single living caterpillar or pupa 

 could be found amongst the thousands of dead and decomposing 

 remains of the victims of overpopulation. These were but a repeti- 

 tion of conditions as observed a few miles north in New Hampshire 

 12 years ago. How frequently similar conditions occurred during 

 the intervening period is not known. 



In addition to those species mentioned in the preceding pages, 

 quite a number of other leaf-feeding Lepidoptera have been more or 

 less casually studied in a less comprehensive but at the same time a 

 careful manner. 



PARASITISM AS A FACTOR IN INSECT CONTROL. 



In reviewing the results of these studies, the fact is strikingly 

 evident that parasitism plays a very different part in the economy 

 of different hosts. Some habitually support a parasitic fauna both 

 abundant and varied, while others are subjected to attack by only 

 a limited number of parasites, the most abundant of which is rela- 

 tively uncommon. No two of the lepidopterous hosts studied, 

 unless they chanced to be congeneric and practically identical in 

 habit and life history, were found to be victimized by exactly the 

 same species of parasites. Neither are the same species apt to occur 

 in connection with the same host in the same relative abundance, 

 one to another, year after year in the same locality, nor in twq 

 different localities the same year. 



At the same time there are certain features in the parasitism of 

 each species which are common to each of the others, whether these 

 be arctiid, liparid, lasiocampid, tortricid, saturniid, or tineid, one of 

 the most common of which is that each host supports a variety of 

 parasites, oftentimes differing among themselves to a remarkable 

 degree in habit, natural affinities, and methods of attack. Depart- 



