492 PROCEEDINGS OF THE AMERICAN ACADEMY. 



Pheidole pilifera, P. vinelandica and P. dentata, on the contrary, 

 belong to the group described by Emery ( ' 94, p. 55) as having 

 two classes of workers, the extremes not being connected by inter- 

 mediates. 



The large size of the larger workers of Camponotus, especially the 

 great size of the head and its disproportionate width, which are corre- 

 lated with greater strength of jaw, adapts them particularly for com- 

 batting enemies and for work requiring great strength. The small 

 workers, owing to their size, are better adapted to pursuits which re- 

 quire agility. However, their jaws and the general proportions of 

 their bodies probably fit them better for carrying loads, such as larvae, 

 pupae, earth, wood, food, etc. In the workers, the position of the 

 eyes, the depth of the head, and the shape of its posterior margin are 

 probably caused by the greater development of the clypeus and the 

 mandibles. But in all these structures there is a gradual transition iu 

 size from one end of the series to the other. 



In Pheidole piUfera, however, the case is different. Though the same 

 parts of the head that are variable in C herculeanus pictus are also sub- 

 ject to variation in this species, there is not a graded series ; there 

 are only three well-marked conditions, a queen, a large worker and a 

 small worker. The small worker, especially because of its weak man- 

 dibles, which are toothed, seems adapted to only the lighter work of 

 the colony, while the soldier, with its peculiar mandibles, is better 

 adapted for heavy work or fighting. The queen resembles in some 

 respects the small worker, in others the large one. The characters of 

 the queen which least resemble those of the small worker, become 

 especially prominent in the soldier, which has, in addition, certain 

 modifications of structure depending on the enormously developed 

 mandibles. It seems that we have a case in harmony with the view of 

 Emery (94) and of Wheeler ( : 07'' ; :10), that the large workers resem- 

 bling the queen were the primitive form, that the small workers were 

 developed later as the result of gradual diminution in size, and that 

 the intermediate forms then disappeared, leaving the extremes as two 

 separate classes, and that in addition the large workers became enor- 

 mously developed along certain lines. 



2. These forms, Camponotus and Pheidole, have been compared with 

 each other for the purpose of discovering, if possible, whether there is 

 any difference in the distribution of functions in the classes of the two 

 genera. With this end in view a larger number of activities has 

 been studied in each group than has hitherto been examined in any 

 one species. Though it was found that artificial conditions made some 

 difference in the proportion of active ants, there does not seem to be 



