TvR'NER, Homing of Ants. 4II 



cm. high. This stack of shdes was placed near enough to the 

 incHne for the ants, by stepping across a narrow horizontal gap, 

 to pass easily from it to the incline. At once the ants mounted 

 the stack of slides and went up the mcline. 



Does not this indicate that the ants have an impression of verti- 

 cal direction ^ Does it not also indicate that they have an impres- 

 sion of distance .? Otherwise, what makes them stop at that place 

 and feel upward for the incline .'' It will not do to say that they 

 stopped because the scented trail terminated there, for, even if it 

 were not true that tracks of those ants cross that place in all direc- 

 tions, it has already been shown that the termination of a scented 

 trail does not impede the progress of ants moving in a direction 

 that is thoroughly known. Nor will it do to say that they halted 

 and acted that way because they sensed the incline above them; 

 for they do not react that way towards bodies held above them on 

 other parts of the stage, nor do untrained ants react that way when 

 they pass underneath the mcline. In each of these test cases, ants 

 may reach upward with the antennae and then pass on; but they 

 do not return to the spot over and over again and reach up as 

 though they were hunting for something that was lost. 



This view is supported by yet other experiments. In another 

 connection, I have mentioned the case of a worker ant that learned 

 to come to a certain point on the island, mount my forceps and 

 be conveyed to the stage. After this mode of conduct had been 

 thoroughly learned, the ant, on leaving the nest, would immedi- 

 ately go to this position on the island and meander in a small circle 

 until my forceps were presented. 



It was not an uncommon thing for an ant that knew the way up 

 and down a certain incline, occasionally to miss arriving at the 

 foot of the incline. In such cases, the ant would not, as a rule, 

 roam here and yon, but would usually turn quickly and hunt for 

 the incline. If it missed it again, it would go back to some point 

 in its path, often as far as the nest opening itself, and, taking a 

 new start, arrive at the incline without further trouble. This 

 mode of conduct not only indicates that ants use certain land- 

 marks as such, a subject which will be referred to in the next sec- 

 tion, but It also seems to harmonize with the views stated in this 

 section. 



Another common occurrence which puzzled me not a little 

 may, perhaps, be best explained in this connection. In some 



