364 J. T. PATTERSON. 



of the Autographra caterpillar. By the time the caterpillar 

 reaches the stage for pupation, the larval parasites consume its 

 entire contents, leaving only the skin of the host which later 

 hardens to form the mummified carcass containing the parasitic 

 pupae. Since the parasites readily emerge in the laboratory the 

 species is very favorable for experimental studies. 



Whenever the infection is heavy an abundance of material 

 can be secured from the field. But in order to avoid the possi- 

 bility of the host egg or larva becoming infected by a species 

 other than Paracopidosomopsis, I have used almost exclusively 

 material reared in the laboratory under experimental control. 

 The only exception in this paper is Fig. 12, Plate II. 



In all of the experiments the moth eggs have been protected 

 from parasites by having the moths lay their eggs under a bell- 

 jar. The fresh eggs thus obtained are then parasitized. The 

 female parasite has been permitted to make but a single ovi- 

 position in each moth egg. The parasitized eggs and the larvae 

 developing from them have been protected from subsequent 

 infection by keeping them in a closed vessel from which all 

 parasites were scrupulously excluded. Under these conditions 

 of protection, I find that the parasitic egg produces asexual as 

 well as sexual larvae. These curious larvae develop irrespective 

 of whether or not the female parasite laying the egg is virgin or 

 impregnated. 



From the standpoint of experimental work, the material 

 presents one difficulty. In a previous paper, I have shown that 

 the female, in about two cases out of three, deposits two instead 

 of one egg at a single oviposition. It is frequently difficult, 

 especially in late stages, to determine in a given parasitized 

 caterpillar whether one is dealing with a case involving the pro- 

 duct of one or of two parasitic eggs. In early stages this difficulty 

 seldom presents itself, because if two polygerms are present they 

 nearly always lie some distance apart in the tissues of the host. 



The age of any stage, referred to in this paper, is based upon 

 the time of oviposition. There is, however, considerable varia- 

 tion in the rate of development of different eggs. Furthermore, 

 temperature greatly influences the rate of development. In 

 August and September about 28 to 30 days elapses between the 



