BUCKINGHAM. — DIVISION OF LABOR AMONG ANTS. 497 



vided her body still contains sufficient food-tissue. I find that the 

 large workers and the soldiers resemble the queens in being active in 

 time of danger, for then they come to the front and enter vigorously — 

 in most species even more vigorously than the small workers — into 

 the defence of the colony, and in small colonies they share more in the 

 work. 



This passivity of the large workers, except in case of need, seems to 

 be a valuable adaptation for the safety of the colony as a whole, in that 

 there are thus always a certain number of well-adapted individuals 

 which are ready to put all their efforts into the defence in a sudden 

 emergency. The other workers may be more or less spent by their oc- 

 cupations, but the large workers, or soldiers, as the case may be, are 

 continually fresh and ready for the fray. Another way in which this 

 passivity on the part of the soldiers seems a profitable adaptation is in 

 guarding, especially in such species as Colobopsis. Were the soldiers 

 of a lively disposition, it would be impossible for them to remain for a 

 long time quietly keeping their heads in the entrances of the nests, 

 and hence a much larger number of individuals would be necessary to 

 carry on this work. 



8. Males were never active in Camponotus, and none were found in 

 Pheidole. 



9. All observers seem to agree, more or less completely, that feed- 

 ing in the larval stage is in some way responsible for differences of 

 classes in ants, the chief reasons for such a view being as follows : 

 (a) Such differences have been shown to be the result of feeding in both 

 bees and termites, (b) When a young queen ant of a species in which 

 the workers are polymorphic establishes a colony, she necessarily rears 

 her workers with a small amount of food ; this first brood always con- 

 sists of small individuals, (c) Poorly nourished larvae do not starve, 

 but pupate and become small ants. As some organs begin to develop 

 at one stage of larval growth and others at other stages, it follows that 

 some organs are likely to become influenced by lack of food more than 

 others. In ants provided with meagre nourishment, nutrition is often 

 not sufficient to stimulate certain organs to develop at all. The con- 

 sequent difference in structure results in the establishment of classes. 

 This may be seen in cases where there is a long series of intermediate 

 forms, as well as in cases where the classes are more obviously separated 

 from one another, for Wheeler (:02) has shown that there is less varia- 

 tion in cases where there is not a plentiful food supply. He (Wheeler, 

 :10, p. 106) also states that "the grossly mechanical withdrawal by 

 parasites of food substances already assimilated by the larva produces 

 changes of the same kind as those which distinguish the worker from 



VOL. XLVI. — 32 



