PERFECT SOCIETIES OF INSECTS. 31 



the larvfB ; tlie neuters, unless these should prove to be 

 the larvas of juales, being the soldiers of the community. 



From this circumstance perfect societies may be di- 

 vided into two classes ; the first including those whose 

 w orkers are larva', and the second those whose workers 

 are neuters'^. The white ants belong to the former of 

 these classes, and the social Ili/menopiera tq the latter. 



Before I begin with the history of the societies of 

 white ants, I must notice a remark that has been made 

 applying to societies in general — that numbers are es- 

 sential to the full development of the instinct of social 

 animals. This has been observed by Bonnet with re- 

 spect to the beaver "^ ; by Reaumur of the hive-bee ; and 

 by M. P. Huber of the humble-bee °. Amongst hymeno- 

 pterous social insects, however, the observation seems 

 not universally applicable, but only under particular 

 circumstances ; for in incipient societies of ants, humble- 

 bees, and wasps, one female lays the foundations of them 

 at first by herself; and the first brood of neuters that is 

 hatched is very small. 



I have on a former occasion given you some account 

 of the devastation produced by the white ants, or Ter- 

 miteSj the species of which constitute the first class 

 of perfect societies''; I shall now relate to you some 



* I employ occasionally the term neuters, though it is not perfectly pro- 

 per, for the sake of convenience ; — strictly speaking, they may rather be 

 regarded as imperfectbr sterile females. Yet certainly, as the imperfec- 

 tion of their organization unfits them for sexual purposes, the term neuter 

 is not absolutely improper. 



* (Euv. ix. 163. " M. P. Ruber in L4nn. Trans, vi. 256. Reaum. v. 

 •* Vol. I. 2cl Ed. 242. 



