COOPERATION 153 



In many desert ants, both division of labor and conservation of 

 food, in the original as well as the derived sense, have been brought 

 about by employing certain workers or soldiers (Wheeler, 1908; 1923: 

 179,335) as "honey jars." The honey is stored in the crops of such 

 individuals, which become so distended that the ants are unable to 

 walk; when stroked by the workers, they regurgitate droplets to serve 

 as food for them. Such reciprocal feeding, or trophallaxis, as Wheeler 

 terms it, usually concerns larvae and adults in ants, as well as in the 

 social wasps, but is entirely lacking in social bees. In wasps, the 

 larvae are fed with pellets made of caterpillars, flies, etc., and the 

 nurse adults imbibe the sugary saliva that exudes from the mouth of 

 the larva. Ant larvae may be given solid food such as pieces of in- 

 sects or pellets or liquids regurgitated by the workers that nurse them. 

 In turn, the larvae yield secretions that are eagerly sought by the 

 adult; these may be sweet saliva or fatty substances excreted through 

 the integument. With termites, trophallaxis is primarily an adult co- 

 action, in which they feed each other with saliva, with regurgitated 

 food, and with feces. In addition, they also produce fatty exudates, 

 which are consumed by other individuals, the outcome being a com- 

 plex food bond beyond that of any other family group of insects. 



Outside of the groups of social insects, cooperation occurs but 

 sporadically until the fishes and higher vertebrates are reached. 

 Though there are scattered instances in other groups of care of eggs 

 or young on the part of one parent, it is rare that the two parents 

 cooperate in this respect. In birds, and many groups of mammals, 

 the cooperation of male and female is the rule, as a consequence of 

 which there results much division of labor in terms of behavior. 

 Among the ungulates, the general absence of nest building and the 

 necessity of constant foraging for food affords much less scope for 

 family cooperative behavior. In the large aggregation families, such 

 as flocks of birds and herds of ungulates, there may be a certain de- 

 gree of division of labor in terms of leaders, scouts, and sentinels, but 

 this apparently does not often assume a definite pattern. 



Finally, cooperation in families is promoted by a means of com- 

 munication among its members. There is no reason to discuss here the 

 moot question of language among bees, ants, and other insects, especi- 

 ally since this hinges largely upon definition. It appears probable 

 that a large number of species possess methods of communicating ob- 

 servations, warnings, and intentions to one another and that these 

 play a large part in the integration of the family group, as well as its 

 protection. One of the most remarkable examples of such cooperation 

 is furnished by squadrons of white pelicans, which perform a number 



