GOVERNMENT OF ANTS. 303 



the reformers by seizing their legs or antennae in no 

 very gentle manner ; tor they sometimes go the 

 length of biting them through, and even, if we do 

 not greatly mistake, of devouring those they suc- 

 ceed in mutilating and mastering. We have seen 

 several instances of this in those we accidentally ob- 

 served in the fields ; but as we could not, in such 

 cases, be certain whether both individuals belonged 

 to the same community, and as those even of the 

 same species are always at deadly enmity, we placed 

 several nests of the red ant (Mijrmica rubra) in 

 glass frames, in order to watch their proceedings. 

 The result was, that when they were confined within 

 narrow limits, they all lived amicably enough, and 

 did not attempt any change, because there was only 

 a single place at all fit for their purpose : but when 

 a passage was allowed them to several places at a 

 distance, the desire for change put them all in agita- 

 tion, and we soon witnessed scenes of dragging and 

 mutilation such as we had previously seen in the fields. 

 We had, consequently, no doubt that the plans of 

 individuals often meet with opposition which lead 

 to violent feuds, sometimes ending in the death of 

 the proposer or of individuals who refused to agree 

 to his plans.* To many all this may appear an 

 imaginary sketch suggested by theory, but we have 

 not stated one circumstance which we have not ac- 

 tually seen in repeated instances. The following re- 

 marks by Huber partly corroborate the preceding 

 facts. 



' I have been enabled,' he says, ' to observe, 

 through the glasses of my artificial ant-hill, the great 

 care taken of the larvee. They were generally guard- 

 ed by a body of ants, who were raised upon their feet, 

 with their tail between their legs ready to cast their 

 venom upon all intruders, whilst, here and there, other 



* JR. 



