78 OX THE HABITS OF ANTS. [LECT. 



This observation has been fully confirmed by other 

 naturalists. However small the prison, however large 

 the quantity of food, these stupid creatures will starve 

 in the midst of plenty, rather than feed themselves. I 

 have had a nest of this species under observation for a 

 long time, but never saw one of the masters feeding. I 

 have kept isolated specimens for weeks by giving them 

 a slave for an hour or two a day to clean and feed them, 

 and under these circumstances they remained in per- 

 fect health, while, but for the slaves, they would have 

 perished in two or three days. I know no other case in 

 nature of a species having lost the instinct of feeding. 



In P. rufescenSj the so-called workers, though thus 

 helpless and stupid, are numerous, energetic, and in 

 some respects even brilliant. In another slave-making 

 species, however, Strongylognathus, the workers are 

 much less numerous, and so weak that it is an unsolved 

 problem how they contrive to make slaves. 



Lastly, in a fourth species, Anergates atratulus, the 

 workers are absent, the males and females living in nests 

 with workers belonging to another ant, Tetramorium 

 ccBSpitum. In these cases the Tetramoriums, having no 

 queen, and consequently no young of their own, tend the 

 young of the Anergates. It is therefore a case analogous 

 to that of 'Polyergus, but it is one in which slave-owning 

 has almost degenerated into parasitism. It is not, how- 

 ever, a case of true parasitism, because the Tetramoriums 

 take great care of the Anergates, and if the nest is 

 disturbed, carry them off to a place of safety. 



M. Forel, in his excellent work on ants, has pointed 

 out that very young ants devote themselves at first to 

 the care of the larvae and pupse, and that they take no 



