•WHEELER— ANT LARV^. 325 



insects had in their evolution passed far beyond the stages repre- 

 sented by the termites, ants and social wasps. There can, indeed, 

 be little doubt that the bees are descended from wasp-like ancestors 

 and that they must therefore have passed from an animal to a vege- 

 table diet. If the change of diet took place after the social habit 

 had been established, as is possible and as is so clearly shown to be 

 the case in the harvesting and fungus-growing ants, the loss of a 

 resort to the larval secretions by the adult social bees could be readily 

 explained as due to the abandonment of a scarce animal food, pro- 

 cured with considerable difficulty, for nectar and pollen, which are 

 abundant and easily obtained. 



Another objection that may be urged against the view that 

 trophallaxis is so fundamental as I contend, is the behavior of the 

 ants towards their inert pupae, which though transported and de- 

 fended as assiduously as the larvae, yield neither liquid exudates nor 

 secretions. This does not seem to me to be a serious objection, be- 

 cause the pupae evidently have an attractive odor and may therefore 

 be said to produce volatile exudates like certain myrmecophiles. 

 Both the larvae and pupae, moreover, evidently represent so much 

 potential or stored nutriment available for the adult ants when the 

 food-supply in the environment of the colony runs very low or 

 ceases entirely. Infanticide and cannibalism then set in with the 

 result that the devouring of the young of all stages may keep the 

 adult personnel of the colony alive till the trophic conditions of the 

 environment improve. Certain predatory tropical species (Dory- 

 linae, Cerapachyini) regularly raid the colonies of other ants and 

 carry home and devour their brood. In northern Eurasia and North 

 America Formica sanguinea makes similar raids on colonies of 

 Formica fusca but permits a certain number of the pupae to hatch 

 and become " slaves." The latter, however, represent only a small 

 portion of the pupae secured during the course of the summer. Was- 

 mann believes that the fusca pupae are plundered for the sake of 

 being reared. This I doubt, but if true, we should have to account 

 for it by supposing that to the sanguinea workers the odor of the 

 fusca pupae is, if anything, even more attractive than that of their 

 own. 



If we confine our attention largely to the ants, I believe it can 



PROC. AMER. PHIL. SOC, VOL. LVII, W, AUGUST Q, I918. 



