SOCIAL TRANSITIONS 265 



book presents the evidence which finally caused him 

 to decide that with ants the whole matter of caste 

 formation is primarily controlled by heredity. 



This is a question which will undoubtedly occupy 

 students of ants for years to come. The evidence is 

 not all in, and the fact that at present it tends to 

 indicate that ant castes are determined by heredity 

 makes all the more interesting the instances in three 

 separate kinds of social insects of the apparent evo- 

 lution of group control of castes after the hatching 

 of the egg. To this hasty sketch of the operation of 

 group determination of caste in wasps and bees may 

 be added that of termites. 



The bees and their allies belong to one of the 

 most specialized of insect orders, so that they are 

 assigned a high position in the evolutionary tree of 

 that class of animals. The termites, miscalled white 

 ants, belong to a relatively unspecialized insect order 

 related to the cockroaches, and stand low in the 

 evolutionary scale among the insects. They have, 

 however, reached a high state of social development. 



Unlike bees, ants and wasps, the colony, as we 

 have said, is at all times composed of males and fe- 

 males in approximately equal numbers. There are 

 male and female reproductives. of which three dif- 

 ferent kinds are known; these are the so-called first 

 form which have wings for a time and engage in a 



