THE SOCIAL INSECTS 



different habitats or to an increase in the size of the colony. 

 Some ants add saliva to vegetable matter, to construct a 

 stiff " carton ". A few use silk, which their grubs spin, to 

 join leaves together. Usually, however, no scarce materials 

 are required to make the nest. Perhaps one reason for the 

 success of the ants has been that their nests do not depend on 

 hardly won materials. They are at any rate more ubiquitous 

 than any other group of social insects, and the only habitats 

 on land in which they are not found are those of the Arctic. 

 They are also much the easiest of the social insects to keep 

 under observation in captivity. Many species can be main- 

 tained for years in small chambers of plaster of Paris with 

 a glass lid. It is necessary only to keep them damp and to feed 

 them on sugar and on occasional dead insects Much of our 

 knowledge of ants, especially of the early stages of the colony, 

 depends on observations made on captive colonies. 



There are good grounds for thinking that ants have led 

 a social life for a much longer period than either bees or 

 wasps. Certainly the castes were fully developed at the time 

 the Baltic amber was fossilised, thirty to forty million years 

 ago, and even long before this. We know this, because we 

 find ants perfectly preserved in the amber. It is probable that 

 their social organisation will be found to be more elaborate 

 than that of the bees, but at the moment it is much less under- 

 stood. Although it is not difficult to keep ant colonies in 

 observation nests, no ant species has been as intensively 

 studied as the honey bee. An immense amount has been 

 recorded about the general natural history of ants, but most 

 of this information is purely descriptive. That is, though we 

 know in a general way what many species do under natural 

 conditions, we do not know how they do it. A real under- 

 standing of ant behaviour would require much more investi- 

 gation by means of experiment. 



120 



