86 ON THE HABITS OF ANTS. [lect. 



her legs. 2.20. The friend was now almost clean, so 

 that I could only just perceive any colour. The 

 stranger, on the contrary, was almost as much coloured 

 as ever. She was now near the door, and I think 

 would have come out, but two ants met and seized 

 her. 3.0. Two ants were attacking the stranger. The 

 friend was no longer distinguishable from the rest. 

 3.30, Do. 4.0, Do. 5.0, Do. 6.0. The stranger now 

 escaped from the nest, and I put her back among her 

 own friends. 



The difference of , behaviour was therefore most 

 marked. The friends were gradually licked clean, and 

 except for a' few moments, and that evidently by 

 mistake, were never attacked. The strangers, on the 

 contrary, were not cleaned, were at once seized, dragged 

 about either by one, two, or three assailants, and at 

 length either made their escape from the nest> or were 

 killed. 



It is certainty most remarkable that ants should 

 thus recognize their friends, after an interval of more 

 than a year. I have since repeated these experiments 

 with similar results. 



Thus, I separated a nest of F. fusca into two 

 portions, on the 20th October, 1876, and kept them 

 entirely separate. 



On the 25th February, 1877, at 8 a.m., I put an ant 

 from the smaller lot back among her old companions. 

 At 8.30 she was quite comfortably established among 

 them. At 9, ditto. At 12, ditto, and at 4, ditto. 



June 8th. — I put two specimens from the smaller lot 

 back, as before, among their old friends. At 1, they 

 were all right among the others, At 2, ditto. After 



