BUCKINGHAM. — DIVISION OF LABOR AMONG ANTS. 491 



nations for the apparent greater activity of the soUiers of this species 

 in feeding : First, the amount of difference between the two classes 

 may not be real, both being about equally concerned in this activity. 

 Secondly, the greater activity of the minors in the field may be due to 

 the fact that they collect more food than they need for themselves, 

 and later regurgitate it from their crops to other individuals. In the 

 case of colony 58 they may have taken only enough for themselves, so 

 that they were here less active than usual, and in consequence, the 

 soldiers appeared relatively more active. In view of the facts that 

 the minors usually regurgitate and soldiers receive, that soldiers at 

 best give out food only very seldom, and that soldiers of Pheidole pili- 

 fera, and probably also of this species, appear only rarely in the field, 

 I am inclined toward the second explanation. 



3. The soldiers were more licked than were the minors. 



4. Though, as in P. piUfera, the minors were most concerned in 

 tending the young, on one occasion a soldier was seen with young in 

 its mouth. 



5. In surrounding the queen, both classes took part, the minors be- 

 ing somewhat more active than the soldiers. This was the only colony 

 of Pheidole in which this activity was tested. 



6. Though the nature of guarding resembled that found in P. piU- 

 fera, the soldiers in P. vinelandica seemed much more active than the 

 minors. 



7. Though carrying and being carried were looked for, no cases were 

 observed. 



8. No individuals of either class crushed or broke up grass-seed. 



VII. Discussion. 



Having given in the historical part of this paper some account of 

 the work of other writers on division of labor among ants, and having 

 recorded my own observations on several species, we may now state such 

 conclusions concerning the division of labor as seem warranted by the 

 facts already established. 



1. It will be remembered that Forel ('95, p. 146) and Escherich 

 ( : 06, p. 45) state for ants in general that polymorphism goes hand in 

 hand with division of labor. Many examples of this are cited in the 

 historical part (II) of this paper. 



Of the species which I have studied, Camponotus americanm and 

 C. herculeanus pictus are examples of a graded series of workers ex- 

 tending from the smallest to the largest with no break in the line, and 

 in many characteristics the queen closely resembles the largest worker. 



