88 EXI^LANATION OF THE PKESENT STATE OF 



of Tetrarrwriwm^ and in some manner contrive to assas- 

 sinate their queen. I have shown that a nest of ants 

 may continue, even in captivity, for six years, without 

 a queen. If, therefore, the female of Anergates could 

 by violence or poison destroy the queen of the Tetra- 

 moriums, we should in the following year have a com- 

 munity composed of the two AnergateSy their young, 

 and workers of TetramoHuTrif in the manner described 

 by Van Hagens and Forel. This would naturally not 

 have suggested itself to these naturalists, because if 

 the hfe of an ant had, as was formerly supposed, been 

 confined to a single season, it would of course have 

 been out of the question ; but as we now know that 

 the life of ants is so much more prolonged than had 

 been supposed, it is at least not an impossibility. 



It is conceivable that the Tetramoriume may have 

 gradually become harder and stronger; the marauding 

 expeditions would then be less fruitful and more dan- 

 gerous, and might become less and less frequent. If, 

 .then, we suppose that the females found it possible 

 to establish themselves in nests of Tetramoriuni, the 

 present stale of things would almost inevitably be, by 

 degrees, established. Thus we may explain the re- 

 markable condition of Strongylognathua, armed with 

 weapons which it is too weak to use, and endowed with 

 instincts which it cannot exercise. 



At any rate, these four genera offer us every grada- 

 tion from lawless violence to contemptible parasitism. 

 Formica sanguinea, which may be assumed to have 



