POSSESSION OF A SIGN OK PASSWOED. 235 



Lastly, on April 9 I again put twelve ants, six 

 from each of these nests, together, and kept them so 

 till October. I then took four of those from 71, 

 put three into 81 and the fourth into 71. I also took 

 four of those from 81, and put three into 71, and 

 the fourth back into 81 among her old friends. The 

 two ants thus restored respectively to their old nests 

 were as usual recognized as friends and left quite 

 unmolested. As regards the other six, the results were 

 as follows. The ants were introduced into the nests at 

 8.15 a.m. 



I do not give these results as by any means proving 

 that ants do not recognize their friends by means of 

 smell. They do seem, however, at any rate, to show that 

 not even six months of close companionship under pre- 

 cisely similar conditions will so far assimilate the odour 

 as to lead to confusion. If the recognition is due in any 

 degree to this cause, the odour is therefore probably an 

 hereditary characteristic. 



In the interesting memoir already cited, Forel says,* 

 "Lubbock {loc. cit.) a cru demontrer que les fourmis 

 enlevees de leur nid a I'etat de nymphe et ecloses hors 

 de chez elles etaient neanmoins reconnues par leurs 



* Eecueil Zool. Suisse, 1887, 



