EXPERIMEXTS WITH ANTS. 267 



which were \^•lli^led round and round went straiglit 

 home. 



Moreover, it appears that after all, as a matter of fact, 

 they did not fly straight home. If they liad doue so they 

 would have been back in three or four minutes, whereas 

 they took far longer. Even then, if they started in the 

 right direction, it is clear that they did not adhere 

 to it. 1 have myself tried experiments of the same 

 kind with hive bees and ants. For instance, I put 

 down some honey en a piece of glass close to a nest 

 of Lashis niger, and wlien the ants were feeding I 

 placed it quietly on the middle of a board one foot 

 square, and eighteen inches from the nest. I did 

 this with thirteen ants, and marked the points at 

 which they left the board. Five of them did so on 

 the half of the board nearest the nest, and eight on 

 that turned away from it. I then timed three of 

 them. They all found the nest eventually, but it took 

 them ten, twelve, and twenty minutes respectively. 

 Again, I took forty ants which were feeding on some 

 honey, and put them down on a gravel-path about fifty 

 yards from the nest, and in tlie middle of a square 

 eighteen inches in diameter, vrhich I marked out on 

 the path by strav/s. 



I prepared a corresponding square on paper, and, 

 having indicated by the arrow the direction of the nest, 

 I marked down the spot where each ant passed the 

 boundary. They crossed it in all directions; and 

 dividing the square into two halves, one towards the 

 nest and one away from it, the number in each were 

 almost exactly the same. 



After leaving the square, they wandered about with 

 every appearance of having lost themselves, and crossed 



