THE SPEECH OF MONKEYS. 197 



entirely by landmarks. On the bark of a tree 

 from which they were gathering their winter 

 stores I observed that there were certain little 

 knots or protuberances by which they directed 

 their course, and which they always passed in a 

 certain order. Between these landmarks they 

 did not confine themselves to any exact path, and 

 the concourse would sometimes widen out over 

 the space of more than an inch ; but as they ap- 

 proached a landmark every ant fell into line and 

 went in the exact path of the others, which rarely 

 exceeded in any case more than an eighth of an 

 inch in width. Whenever an ant would lose its 

 way it would lift its head high into the air, look 

 around, and then turn almost at right angles 

 from the course it was pursuing toward the path 

 of the others. In scores of cases I observed that 

 the outward-bound ant, when it had been lost and 

 returned to the path, always came on the home- 

 ward side of the landmark and passed out. On 

 the other hand, if a homeward-bound ant was lost, 

 it would approach from the outward side of the 

 landmark and pass in. About five feet from the 

 ground were two small round knots about one- 

 eighth of an inch in height, and a space between 

 them of about the same width. This appeared 



* 



