299 



the increasing paralyzation. Tliey came to rest with the 

 abdomen directed toward the head of the larva and not quite 

 touching it. From time to time the larva seemed to relax 

 so as to touch the Sclerodermus which would stir slightly l)ut 

 without trying to sting, but the larva would jerk back each 

 time. , 



After about ten minutes of this, there was a moment of 

 fencing again, only two or three movements of the larva and 

 corresponding stinging movements of the Sclerodermus but 

 apparently no actual stinging and the conditions were resumed. 

 After about three minutes the caterpillar's head sank down 

 but very soon the conflict was renewed, the Sclerodermus at- 

 tacking along almost the entire length of the larva, which was 

 aggressive and changed its position entirely. After this con- 

 flict the Sclerodertniis retired to shelter again some distance 

 away. 



Definite efforts were apparently made in this conflict 

 to sting on the side of the anal extremity and the side of the 

 thorax as well as at the head. 



After this conflict both insects remained quiescent for 

 some time but after about 20 minutes, the caterpillar began 

 to move slowly and in about half an hour it was able to move 

 about rather freely but more feebly. In the meantime the 

 Sclerodermus female had joined the others, doubtless more or 

 less confused by the presence of the quiet larva paralyzed by 

 the other two females. 



Ovipositioth. The females have not been observed to 

 oviposit in unj case short of seven days after emergence, 

 giving them ample time for emergence* consumption of the 

 cocoon, dispersal, finding the prey, mastering the proy, 

 feeding upon it and a resting period afterward. 



Careful observation shows that the female does not prepare 

 a place for oviposition by piercing the skin of the prey with 

 her mandibles or sting. Tn oviposition the body is curved 

 transversely around the body of the larva. The sting and 

 sheaths aie extended and the hypopygium was widely sepa- 



