128 SUPPOSED KECOGNITION BY SCENT. 



Presently I took up the struggling heap. Two of the 

 assailants kept their hold; one finally dropped, the 

 other I could not tear loose, and so put the pair back 

 upon the tree, leaving the doomed immersionist to her 

 hard fate.' 



After recording one or two other similar observa- 

 tions, he adds : ' — ' The conclusion, therefore, seems 

 warranted that the peculiar odour or condition by which 

 the ants recognise each other was temporarily destroyed 

 by the bath, and the individuals thus " tainted " were 

 held to be intruders, alien and enemy. This con- 

 clusion is certainly unfavourable to the theory that any 

 thing like an intelligent social sentiment exists among 

 the ants. The recognition of their fellows is reduced 

 to a mere matter of physical sensation or " smell." ' 

 This conclusion does not, I confess, seem to me to be 

 conclusively established. 



We can hardly suppose that each ant has a pecu- 

 liar odour, and it seems almost equally difficult, con- 

 sidering the immense number of ants' nests, to suppose 

 that each community has a separate and peculiar smell. 

 Moreover, in a previous chapter I have recorded some 

 experiments made with intoxicated ants. It will be 

 remembered that my ants are allowed to range over a 

 table surrounded by a moat of water. Now, as already 

 mentioned, out of forty-one intoxicated friends, thirty- 

 two were carried into the nest and nine were thrown 

 into the water; while out of fifty-two intoxicated 



' Mound-making Ants of the AlleghaniM, p. 2i8L. 



