NOB DUE TO THE USE OF A PASSWORD. 151 



There was no dead ant either in the nest or the outei 

 box. 



Lastly, on October 1 5, 1 put in four more at 7 A.M., 

 and watched them all day at short intervals. They 

 exhibited no sign of fear, and were ne\'er attacked. 

 In fact, they made themselves quite at home, and were 

 evidently, like the preceding, recognised as friends. For 

 the sake of comparison at noon I again put in a stranger. 

 Her behaviour was in marked contrast. The preceding 

 ants seemed quite at home, walked about peaceably 

 among the other ants, and made no attempt to leave 

 the nest. The stranger, on the contrary, ran uneasily 

 about, started away from any ant she met, and made 

 every effort to get out of the nest. After she had 

 three times escaped from the nest, I put her back with 

 her own friends. 



Thus, then, when a nest of Formica fusca was 

 divided early in spring, and when there were no young, 

 the ants produced in each half were in twenty-eight 

 cases all received as friends. In no case was there the 

 slightest trace of enmity. 



These observations seem to me conclusive as far as 

 they go, and they are very surprising. In the previous 

 experiments, though the results were similar, still the 

 ants experimented with had been brought up in the 

 nest, and were only removed after they had become 

 pupae. It might therefore be argued that the ants 

 having nursed them as larvae, recognized them when 

 they came to maturity ; and though this would cer- 



