460 ANNUAL KEPOET SMITHSONIAN INSTITUTION, 1912. 



in Formica rufa, F. pratensis, and F. exsedoides the females are very 

 large and do not yet shoAV any trace of parasitic modifications. All 

 the Formica species just mentioned, whose females either faculta- 

 tively or obligatorily found their colonies with the aid of the workers 

 of an ahen species of auxihary ant, form only temporarily mixed 

 colonies. After the death of the original auxiliary ants — about three 

 years after their foundation — these colonies again become simple, 

 unmixed ant colonies and as such may still continue to increase for 

 decades. Permanently mixed colonies we find, on the contrary, with 

 the dulotic Formica species, which, after the dying off of the original 

 auxiliary ants, procure new ones for themselves by the capture of 

 pupse. How do these forms link with the preceding and especially 

 with a rufa-like initial stage of social parasitism ? 



When with a large strong acervicolous species of Formica, which 

 had already passed over from the independent to the dependent 

 foundation of colonies, a change occurs in the mode of nutrition of 

 the workers, conditioned by clunatic changes,^ so that it lives more 

 and more exclusively by preying upon insects and furthermore par- 

 ticularly upon the pup* of strange ants, then the basis is ofl'ered for 

 the origin of slavery; for provision is already made, through the 

 dependent foundation of these colonies, that among the captured 

 strange pupas precisely those of the auxiliary species shall be reared. 

 Even Formica truncicola and F. exsecta, which in nature are not slave 

 makers, in those of their colonies which have again become simple 

 retain the inclination to rear the worker pupse of the species of their 

 former auxiliary ant, if they are given them in artificial nests, while 

 they devour the pupae of other alien species or at least kiQ the work- 

 ers emerging from them. The origin of slavery in a Formica form 

 like F. sanguinea (fig, 6) depends, then, upon two agents: (1) Upon 

 the dependent foundations of colonies by their females with the assis- 

 tance of a strange species of ant; (2) upon the inclinations of their 

 workers to capture strange pupae as prey. According to this view 

 the hypothetical origin of slave making within the genus Formica 

 is thus to be followed back to a common root with the origin of 

 social parasitism within the same genus, namely, to an incipient 

 stage of dependent foundation of colonies which is to some degree 

 comparable to the present state of F. ruja. In any case, the mclina- 

 tion to prey upon pupse does not in itself sufiice to explain the origin 

 of dulosis in Formica or in any other genus of ants; for there are 

 many species of ants, especially in the subfamily Dorylinae, which 

 pillage the pupae of strange ants and notwithstanding do not rear 

 slaves from them, because precisely the first of the two above-named 

 agents, the dependent foundation of their colonies by means of an 



1 How the change of a forest climate to a prairie chmate can ofier this occasion, I have shown especially 

 for Formica sanguinea. 



