THE SOCIAL INSECTS 



the possibility of providing itself with the necessary variety 

 of food. 



The theory put forward by Grasse is that caste-determina- 

 tion is essentially a " group effect ", acting through the 

 nervous system. This implies that the physical development 

 and behaviour of a termite is determined in part by the 

 sensory impressions which it receives from the other members 

 of the colony. This might be automatically reinforced by 

 differences in feeding which become established as the 

 behaviour of different individuals diverged. Group effects 

 of a kind are, of course, common in human beings. We learn 

 the language and often the vocabulary and trade of our 

 immediate neighbours. But this is the result of imitation, of 

 learning and of the legacy of traditions, none of which is 

 operative amongst termites. A closer but still false analogy 

 is with the revolt, often subconscious, which children some- 

 times make against the ideas or way of life of their parents. 



Group effects are, however, well known in many insects, 

 though none is quite comparable with what is assumed to 

 happen in termites. One example has already been described, 

 namely the inhibition of oviposition in humble-bee and 

 Polistes workers as long as the queen is present. In Polistes, 

 the removal of the queen releases egg-laying within a day or 

 two. Group effects have also been noted in various beetles 

 and flies when bred in laboratory cultures. Grain weevils, 

 for instance, do not lay at the highest rate of which they are 

 capable unless each female has at least fifty grains of wheat, 

 so that twenty females on six hundred grains lay less well 

 than the same number on one thousand grains. Moreover, 

 the weevils bred from a moderately crowded culture are 

 larger and are capable of laying more eggs than are those 

 reared from cultures which were either less or more crowded. 

 This last effect seems to be due to physical changes in the 



1 86 



