IMPOETANCE OF SCENT. 



259 



then twisted as shown in Fig. 19. The bricks and tunnel 



being also rearranged so that they were exactly in the 



same direction as before, but the board having been moved, 



the line d e was now outside them. This change, however, 



didnotatalldiscom- Fig. 19. 



pose the ants ; but 



instead of going, as 



before, through the 



tunnel and between 



the rows of bricks 



to a, they walked 



exactly along the 



old path to e. 



I then arranged 



matters as before, 



but without the tunnel and with only three pairs of 



bricks (Fig. 20). When an ant had got quite used to the 



path d to e,l altered 



the position of the 

 bricks and food, as 

 shown in Fig. 2 1 , mak- 

 ing a difference of 8 

 inches in the position 

 of the latter. The ant 

 came as before, walked 

 up to the first brick, 



touched it with her antenna, but then followed her old 

 line to a. From there she veered towards the food, and 

 very soon found it. When she was gone, I altered it 



S 2 



