SPECIES OF ANTS. 79 



larvae as food. It is as if we had small dwarfs, aboxA 

 eighteen inches to two feet long, harbouring in the 

 walls of our houses, and every now and then carrying 

 off some of our children into their horrid dens. 



Most ants, indeed, will carry off the larvae and pup» 

 of others if they get a chance ; and this explains, or at 

 any rate throws some light upon, that most remarkable 

 phenomenon, the existence of slavery among ants. If 

 you place a number of larvae and pupae in front of a 

 nest of the Horse ant (^F. rufa), for instance, they are 

 soon carried off; and those which are not immediately 

 required for food remain alive for some time, and are 

 even fed by their captors. 



Both the Horse ant (^Formica rtifa, PI. II. fig. 5) 

 and the slave ant {F. fusca, PI. I. fig. 3) are abun- 

 dant species, and it must not unfrequently occur 

 that the former, being pressed for food, attack the 

 latter and carry off some of their larvae and pupae. 

 Under these circumstances it no doubt occasionally 

 happens that the pupae come to maturity in the nests 

 of the Horse ant, and it is said that nests are some- 

 times, though rarely, found in which, with the legiti- 

 mate owners, there are a few F, fuscas. With the 

 Horse ant this is, however, a very rare and exceptional 

 phenomenon ; but with an allied species, F. sanguinea 

 (PL I. fig. 6), a species which exists in some of our 

 southern counties and throughout Europe, it has be- 

 come an established habit. The F. sanguineas make 

 periodical expeditions, attack neighbouring nests, and 



