THEIR ARCHITECTURE. 179 



may have some mode of guiding themselves inde- 

 pendently of light. 



It has been said that they know the structure of the 

 nest, and can trace their way by touch, using their 

 antennae as a blind man does his stick. But they cannot 

 know the construction of other nests, and yet, when one 

 body of ants attacks the nest of another, the invaders 

 plunge boldly into the galleries of the strange nest, and 

 seem to have no difficulty in chasing their inmates 

 through them. 



Their eyes have nothing about them especially worthy 

 of attention, except the extremely variable number of the 

 facets. In one species the insects have a thousand facets 

 to each compound eye ; but this number is, however, far 

 less than is found in many other insects, especially those 

 butterflies which belong to the genus Fanessa. 



There seems, indeed, to be little doubt that the ants 

 depend much more on touch than on sight for their 

 guidance. Sir John Lubbock made a series of experi- 

 ments with some ants, giving them food, and placing it 

 on a wooden cylinder eight inches in height. When an 

 ant had been accustomed to make her way to the honeyed 

 bait, the cylinder was moved only six inches from its 

 former position. The ant, on her return from the nest, 

 was quite bewildered, not appearing to see the cylinder, 

 and only came against it by chance. 



The evident inference was that sight was a less valu- 

 able guide to the food than some other faculty, which 

 was probably the sense of smell. No experiments 



