160 CONDITIONS OF THE PROBLEM. 



Hymenoptera are really so highly gifted, it ought 

 not to be necessary for us to rely on accidental observa- 

 tions ; we ought to be able to test them by appropriate 

 experiments. 



Tliose which I have made with reference to bees 

 will be described in a subsequent chapter. 



Every one knows that if an ant or a bee in the course 

 of her 1-ambles has found a supply of food, a number of 

 others will soon make their way to the store. This, 

 however, does not necessarily imply any power of de- 

 scribing localities, A very simple sign would suffice, 

 and very little intelligence is implied, if the other ants 

 merely accompany their friend to the treasure which she 

 has discovered. On the other hand, if the ant or bee 

 can describe the locality, and send her friends to the 

 food, the case is very different. This point, therefore, 

 seemed to me very important ; and I have made a 

 number of observations bearing on it. 



The following may be taken as a type of what hap- 

 pens under such circumstances. On June 12, 1874, I 

 put a Lasiiis niger, belonging to a nest which I had 

 kept two or three days without food, to some honey. 

 She fed as usual, and then was returning to the nest, 

 when she met some friends, whom she proceeded to 

 feed. When she had thus distributed her stores, she 

 returned alone to the honey, none of the rest coming 

 with her. When she had a second time laid in a stock 

 of food, she again in the same way fed several ants on 

 her way towards the nest ; but this time five of those 



