PERFECT SOCIETIES OF INSECTS. 119 



iiest; ill some species it not being more than fifty or 

 sixty. — They may more easily be studied than either 

 wasps or hive-bees, as they seem not to be disturbed 

 or interrupted in their works by the eye of an ob- 

 server ^ 



I am, &c. 



' This account of the proceedings of humblc-hces is cbicfly taken from 

 Reaumur, vi. Mem. 1. ; and M. P. Huber in Linn. Tran':. vi. 91 !— 



