14 BULLETIN 1487, U. S. DEPARTMENT OF AGRICULTURE 
of feeding. The interruptions in feeding at the times of molting 
were very brief, usually of only two or three hours’ duration. On an 
average 744 days were spent as a pupa, bringing the total period 
from egg to adult, in the case of the summer generations, to about 
22 days. The artificial method of carrying the parasites through 
the immature stages apparently had no effect upon the rapidity of 
development, for in the case of a large number of individuals which 
were allowed to develop normally within Apanteles cocoons, the total 
period from egg to adult ranged from 15 to 30 days, depending prin- 
cipally upon temperature. 
Hemiteles tenellus is one of the first of the hyperparasites which 
attack Apanteles melanoscelus to appear in the spring, sometimes 
emerging from overwintering cocoons as early as April 20. The 
number of generations annually varies from one to four, with three 
being the most common number and with four being much more 
common than one. There is, however, much irregularity in this, 
even among the progeny of a single parent. In numerous cases 
where several cocoons were attacked by the same individual, some 
produced adult Hemiteles after 18 to 24 days to begin another genera- 
tion, while others did not yield adults until the following spring. 
The species hibernates as a mature larva within the cocoon of its 
host. 
Females are invariably produced in parthenogenetic reproduction. 
In the laboratory several pure lines of females have been obtained 
through 12 generations, over a period of three years. The male 
is unknown. In the course of the rearing of many thousands of 
parasites from field-collected cocoons of A. melanoscelus nothing has 
been obtained that could be the male of this species, although females 
have always been found in large numbers. Males of the European 
Hemiteles areator (Panz.), which is not clearly distinguishable from 
tenellus morphologically, and of which tenellus has sometimes been 
regarded as a subspecies or variety, occur abundantly in museum 
collections. Having the opportunity, during the summer of 1924, 
- to determine the result of parthenogenetic reproduction with ff. are- 
ator, the writers found that unfertilized females produce males. 
It is evident from this disparity in the biology of the two forms that 
they are quite distinct. | 
The writers have obtained H. tenellus with remarkable regularity 
from all their collections of cocoons and usually in considerable 
numbers. The fact that it usually has several generations annually 
and can develop upon a great variety of hosts, its ruggedness, and 
its characteristic of producing females in parthenogenesis combine 
to make this hyperparasite one of exceptional importance. 
HEMITELES FULVIPES GRAVENHORST 
(Fig. 2) 
Like H. tenellus, H. fulvipes attacks a large number of different 
primary parasites. It agrees with that species also in nearly all de- 
tails of its biology. The eggs and the larval instars are indistinguish- 
able; the period of larval development, the number of generations 
annually, and the form in which hibernation occurs all agree. In 
one particular, however, H. fulvipes differs decidedly from H. tenellus: 
In parthenogenetic reproduction males are always produced, whereas 
the progeny of unmated females of H. tenellus are invariably females. 
* 
eae Se eS ee Ue le 
