SOCIETAL EVOLUTION I47 



environment, the adult life of the mother or of both parents is 

 considerably prolonged so that it overlaps the larval period 

 or even a part or the whole of the adult Hfe of the offspring 

 and thus furnishes an opportunity for close relations between 

 the members of two successive generations. I find these 

 conditions reahzed in at least thirty insect groups, which 

 are often so remotely related to one another or related in 

 such a manner that the family must be supposed to have 

 arisen independently (polyphyletically) on at least as many 

 separate occasions during the long racial history of the 

 Hexapoda, which extends over some 300,000,000 years 

 from the Upper Carboniferous to the present time. These 

 famihes vary greatly in complexity and stabihty. In most 

 cases the parents are deserted by the progeny while the 

 latter are still young, and the rudimentary society dissolves, 

 a condition observed in what I have called the subsocial 

 insects (certain beetles, wasps, bees, Embiids, earwigs, etc.). 

 In the termites, ants, higher wasps and bees, however, the 

 affiliation of the progeny with the mother (Hymenoptera) 

 or with both parents (termites) becomes much more intimate 

 and prolonged, so that at least the worker caste, which 

 constitutes the great majority of the personnel of the 

 society, never dissolves its consociation with the parents. 

 The single family is thus enabled to remain a society, 

 though capable in some cases of growth to a population of 

 hundreds of thousands of individuals. This is accompHshed 

 by partial starvation of most of the offspring so that they 

 fail as larvae to develop their reproductive organs (ahmen- 

 tary castration) and even as adults, in their capacity as 

 nurses, inhibit the further development of their gonads 

 by starving themselves as a result of feeding the successive 

 broods of larvae, the queen and the other adult members 

 of the colony (nutricial castration). 



In the foregoing account of insect societies nothing is 

 said about the nature of the bonds which unite the parents 

 and offspring and thus initiate the family or about the 

 nature of the social medium which regulates the social 

 behavior. I beheve that we may detect these recondite 

 factors in what I have called "trophallaxis," or exchange 

 of food. 



