The Mason-Wasps 



pacity of the cell and the slender form of 

 the future insect, the cocoon is rounded at 

 the top and suddenly truncated at the base, 

 which is rendered hard and opaque by the 

 stercoral plug, the by-product of the 

 lacquer-factory. 



The hatching-period varies, of course, ac- 

 cording to the temperature and also accord- 

 ing to certain conditions which I am not yet 

 in a position to specify. One cocoon, 

 woven in July, gives birth to the perfect 

 insect in the course of August, two or three 

 weeks after the larva's period of activity; 

 another, dating from August, opens a 

 month later, in September; a third, no mat- 

 ter what its date of origin during the sum- 

 mer quarter, goes through the winter and 

 does not burst until the end of June. By 

 combining the birth-certificates recorded, I 

 seem to distinguish three generations in the 

 year, generations which are often but not 

 invariably realized. The first appears at 

 the end of June: this is the one whose co- 

 coons have gone through the winter; the 

 second is seen in August and the third in 

 September. So long as the very hot 

 weather lasts, evolution is rapid: three or 

 four weeks suffice to complete the Pelopaeus' 

 cycle. When September arrives, the fall in 

 104 



