RECOGNITION OF FRIENDS. 121 



The observation recorded by Huber was made 

 <?asually, and he did not take any steps to test it by 

 subsequent experiments. The fact, however, is of 

 so much importance that I determined to make 

 further observations on the subject. In the first place, 

 I may repeat that I have satisfied myself by many 

 experiments, that ants from one community introduced 

 into another, — always be it understood of the same 

 species, — are attacked, and either driven out or killed. 

 It follows, therefore, that as within the nest the most 

 complete harmony prevails- — indeed, I have never seen 

 a quarrel between sister ants — they must by some means 

 recognise one another. 



When we consider their immense numbers this is 

 sufficiently surprising ; but that they should recognise 

 one another, as stated by Huber, after a separation of 

 months, is still more astonishing. 



I determined therefore to repeat and extend his 

 observations. 



Accordingly, on August 20, 1875, I divided a colony 

 of Myrmica ruginodis, so that one half were in one 

 nest, A, and the other half in another, B, and were 

 kept entirely apart. 



On October 3, I put into nest B a stranger and 

 an old companion from nest A. They were marked 

 with a spot of coloui-. One of them immediately 

 flew at the stranger ; of the friend they took no 

 notice. 



October 18. — At 10 a.m. I put in a stranger and a 



