330 WHEELER— ANT LARV^. 



The individual ant whicli first finds the larva is always the one to carrj' 

 it off. Although during its attendance several other ants may find the larva 

 and stay by it a short time, and even milk it, they soon leave it to its original 

 attendant, who apparently informs them that their services are not needed. 

 [ !] Whether the ant signals to the larva for it to prepare itself for transit, 

 or the larva gives the signal that it is ready to be taken, seems doubtful ; but 

 from what we have seen both Capt. Purefoy and I are inclined to think that 

 the larva gives the* signal. No. 3 larva alluded to hunched itself both the 

 second and third time while the ant was about an inch away and facing 

 an opposite direction, and at the fourth hunching up the ant was standing 

 over the larva ready for the signal, and when this was given it was quickly 

 seized and carried. 



Chapman observed that after the caterpillar was taken into the 

 nest it fed on the Mynnica larvae. During this period of its Hfe it 

 was not seen to yield the secretion of its honey-gland but was 

 treated by the ants as what Wasmann would call an indififerently 

 tolerated guest, or synoekete. 



5. The fifth expansion of trophallaxis, namely the acquisition of 

 trophic relations with the myrmecophytes, or plants possessing 

 extra-floral nectaries or food-bodies, is also imperfect like ordinary 

 trophobiosis, since the ants merely obtain nutriment from the plants 

 and possibly afford them some protection. The nectar and other 

 plant-foods are for the purposes of the ants merely so many exu- 

 dates like the excrement of the Homoptera (honey-dew) and the 

 sweet secretions of the Lycsenid caterpillars which feed on the 

 foliage. 



As the foregoing study of trophallaxis has an important bearing 

 on Wasmann's and Holmgren's interpretation of symphily it will be 

 advisable to consider their views in greater detail. Wasmann has 

 elaborated his ideas in regard to the origin and meaning of symphily 

 in several papers, but as an article published in 1910 embodies his 

 mature and apparently final contentions, his earlier publications 

 need not be drawn into the discussion. Having fomid that particu- 

 lar symphiles live only with particular host ants and termites, he con- 

 cludes, first, that the latter have during their phylogeny acquired 

 particular symphilic instincts as dift'erentiations or modifications of 

 their original nursing and adoptive instincts, and second, that the 

 true ant and termite guests have been developed by these symphilic 

 instincts through a process called " amical selection," which he 



