1 42 THE NA TU RE-STUD Y RE VIE W 



1 3 : 5 — MAY, 1907 



rely are here too weak to do any injury and when a conflict does 

 ensue between them and their slaves, it is the masters that 

 invariably succumb. Secondly, and as might be expected, 

 these degenerate and useless soldiers are very much reduced in 

 numbers. In this species, therefore, the worker-caste has 

 disappeared, and except in name only, the soldiers also. They 

 have become totally dependent. 



We are now confronted with the difficult problem of the 

 origin of this mixed colony. Surely the explanation given in 

 the case of the Polyergus colonies cannot apply here. For in 

 many cases at least the slaves are not queenless, and this rules 

 out the possibility of the adoption of the Strongylegnathus queen 

 by Tetramorium orphans. Still more improbable is the per- 

 petuation of the mixed colony through the later plundering of 

 near-by slave nests. It is therefore, necessary to seek some other 

 explanation. In this case it is quite possible that we are con- 

 cerned with the third type of mixed colony, that is, one that 

 arises through the alliance of two distinct colonies. Two young 

 and fertilized queens representing the two species come to- 

 gether and in a common nest lay their eggs and raise their first 

 batch of young, the bulk of the labor falling on the Tetramorium 

 slaves. But with the growth of the colony a strange fact be- 

 comes apparent. The slave-workers apparently affect the 

 small Strongylognathus kings and queens much more than 

 their own larger ones, and as a result we find that while the 

 number of sexual forms of their own decreases, that of the others 

 increases. In this way is very probably produced this mixed 

 colony presenting the peculiar numerical relations of the castes 

 of the castes of both species; the sexual forms of the masters 

 preponderating, while the workers and soldiers are almost 

 exclusively slaves. 



This account of the slave-making habits among ants must 

 now be concluded with the final and most complete stage of depend- 

 ence of one species upon another. This is typified by the 

 European genus Anergates (Fig. 3) and by the recently dis- 

 covered American species Epoecus pergandei and Epipheidole 

 inquilina. In these forms all the neutral females, that is, the 

 workers and soldiers, have entirely disappeared, leaving only 

 very peculiar males and females to continue the race. In the 

 case of Anergates these are found associated with small numbers 



