TO FRIENDS AND STRANGERS. 105 



were neglected. On the other hand, some of the ants 

 were always watching over the bottle containing the 

 strangers, and biting at the muslin which protected 

 them. The next morning at 6 a.m. I found five ants 

 thus occupied. One had caught hold of the leg of one 

 of the strangers, which had unwarily been allowed to 

 protrude through the meshes of the muslin. They 

 worked and watched, though not, as far as I could see, 

 with any system, till 7.30 lq the evening, when they 

 eflFected an entrance, and immediately attacked the 

 strangers. 



September 24. — I repeated the same experiment 

 with the same nest. Again the ants came and sat over 

 the bottle containing the strangers, while no notice was 

 taken of the friends. 



The next morning again, when I got up, I found 

 five ants round the bottle containing the strangers, 

 none near the friends. As in the former case, one of 

 the ants had seized a stranger by the leg, and was trying 

 to drag her through the muslin. All day the ants 

 clustered round the bottle, and bit perseveringly, 

 though not systematically, at the muslin. The same 

 thing happened all the following day. 



These observations seemed to me sufficiently to test 

 the behaviour of the ants belonging to this nest under 

 these circumstances. I thought it desirable, however, 

 to try also other communities. I selected, therefore, 

 two other nests. One was a community of Polyergua 

 rufeacens with numerous slaves. Close to where the 



