138 Chapter III. 



But neither the sanguinms nor their slaves {F. rufa 

 and fusca), Hving in the same nest, ever hit upon this 

 obvious method, although it would have sufficed to 

 raise the surface of their nest at the spot in question 

 just by I cm. ! 



A more wonderful result was obtained in another 

 experiment on the same nest of sanguinea. On June 

 1 6, 1884, I filled a large watch-crystal with water and 

 in the center upon a kind of island I placed a little 

 shell filled with ant-cocoons previously taken from the 

 same colony. This artificial pond with its island was 

 then introduced into the nest. The ants soon noticed 

 the cocoons and stretched out their feelers towards 

 the island ; but getting into the water at every attempt 

 to approach, they retreated again and again. I 

 began to think they would never be able to overcome 

 the difficulty, when suddenly a sanguinea began to 

 throw into the water pellets of earth, bits of wood, 

 dead ants and similar solid materials. Others followed 

 her example and they soon had built a road over the 

 zvater! In the space of an* hour, counting from the 

 minute I started the experiment, they had fetched all 

 the cocoons from the island by means of this ^'floating 

 bridge." The very last cocoon having been secured 

 by the ants, one of them returned to the island and, 

 finding it empty, she squatted on her haunches, passed 

 the spur of her fore-feet through her mouth and then 

 combed her feelers with the spur, sitting there for 

 several minutes in a most provoking attitude, as if she 

 were saying to me: "Ah, my dear, who has won the 

 game now?"^ Is this fact not a staggering proof, 



') This very last instance, as many others in this translation, was 

 added by the author from his original notes. 



