8 Some Solitary Wasps of Texas. 



first Qgg laid was on August 30, and the total period of develop- 

 ment would be thirty-nine days. The wasps emerged (Fig. 11) by 

 a small round opening gnawed in the cell wall. This hole was 

 in three cases near the last sealed end, in one case at the opposite 

 end. 



One cell, the last one stored, I broke open to examine the con- 

 dition of the caterpillars. I was especially anxious to see this, 

 as I had once observed a specimen hang from a twig by one of 

 the hind legs and chew the head of a cotton-worm, holding the 

 caterpillar with her front and middle legs. I had often seen 

 Microbembex do this in exactly the same way, and as this species 

 feeds her larva on any insect, dry or fresh, or on any part of an 

 insect in almost any condition, I concluded that she sucks as much 

 of the insects' juices as she needs for her own sustenance and 

 feeds the rest to the larva. Charles Janet has likewise observed 

 this with Vespa crabro, where the workers, wlien the colony is 

 threatened with over-population, kill some of the larvse and pupae^ 

 suck their juices and feed the remainder, rolled up in balls, to 

 the surviving larvae. In the case of Odynenis, I am pretty certain 

 that she takes a caterpillar occasionally for her own delectation. 

 I could draw quite near to the chewing individual and could see 

 every movement, even of the mandibles, during the process. The 

 condition of the caterpillars in the cell I opened, moreover, pointed 

 to the fact that in chewing the caterpillar she was not preparing 

 it for her offspring, but was satisfying her own hunger. Of the 

 seven caterpillars found in the one cell^, all were in good condition, 

 and four responded very perceptibly to stimulation, one of the 

 latter moving spontaneously. All remained in practically the same 

 condition until attacked by the growing larva. (Fig. 10.) 



I watched the growth of the larva until it was ready to spin 

 its cocoon (September 26th). But as I had torn away a goodly 

 part of the wall of the cell, it could not spin its cocoon all around 

 and died. But the absence of the wall was not the chief cause of 

 the pre-pupa's death. When I returned to her on October 7tli, 

 I found the pupal skin completely covered with egg cases of a 

 parasite. On close examination, these proved to contain myriads 

 of mites in various stages of development. Such egg-cases I have 

 often noticed on dead digger-wasp larvs, and they usually appear 

 on the articulating membranes between the segments. 



From cell No. 4 the imago failed to emerge. In November, when 

 I broke open the cell, I found the nearly mature wasplet dead. 



