OF WASHINGTON, VOLUME XV, 1913 159 



of hosts the parasites of both sexes emerged, the males issuing 

 from one to two days ahead of the females. The following table 

 gives the proportion and distribution of the sexes so far as this 

 was determined. 



Table showing proportion of males an dfemales of Sphceropyx bicolor Cr. in 

 specimens reared from individual hosts. 



When the cocoons, which are of white silk, are spun under 

 natural conditions within the host cocoon they are packed paral- 

 lel in a close mass with their axes nearly perpendicular to that of 

 the host cocoon. The parasite hibernates as a larva within its 

 cocoon. 



Very frequently Sphceropyx bicolor is attacked in its cocoon 

 by Dibrachys boucheanus Ratz., an omnivorous secondary para- 

 site, although it is but seldom that all the cocoons in a mass are 

 parasitized. 



Two cases of double primary parasitism in which this species 

 took part were observed. The other species in each case was a 

 tachinid fly. In the first case the tachinid eggs were found on the 

 skin of a caterpillar, from which a few days later the larvae of 

 ^l>hceropyx emerged. In the other instance, an adult tachinid 

 was reared followed three days later by nine females of Sphceropyx. 



The discovery that certain members of the subfamily ( helo- 

 niii(i\ to which X/i/Kt'ropyx belongs, parasitize the eggs of their host, 

 their larvae issuing later from the host larvae, led to the conduct- 

 ing of experiments on a small scale to determine the oviposition 

 habits of the species under discussion. Some eggs of two species 

 < f . 1 /mtela, one of which feeds on wild cherry and the other on pear, 

 were secured and exposed to the attack of S. bicolor. It wa.- 

 noted that while the parasites took no notice of the eggs they 

 showed great excitement when on leaves bearing eggs, running 

 rapidly about dragging their ovipositors over the surface of tin- 

 leaf and searching minutely with their antennae. No such excite- 

 ment was shown when uninfested leaves were supplied. This sug- 

 gests that the search for hosts may possibly lie guided somewhat 



