270 INSECT ARCHITECTURE. 



or formicaries. The preceding figure represents a 

 view of one of his formicaries of mason-ants. 



We have ourselves followed up his observations, 

 both on natural ant-hills and in artificial formicaries. 

 On digging cautiously into a natural ant-hill, esta- 

 blished upon the edge of a garden walk, we were 

 enabled to obtain a pretty complete view of the in- 

 terior structure. There were two stories, composed 

 of large chambers, irregularly oval, communicating 

 with each other by arched galleries, the walls of all 

 which were as smooth and well polished as if they 

 had been passed over by a plasterer's trowel. The 

 floors of the chambers, we remarked, were by no 

 means either horizontal or level, but all more or less 

 sloped, and exhibiting in each chamber at least two 

 slight depressions of an irregular shape. We left 

 the under story of this nest untouched, with the 

 notion that the ants might repair the upper galleries, 

 of which we had made a vertical section ; but instead 

 of doing so they migrated during the day to a large 

 crack, formed by the dryness of the weather, about 

 a yard from their old nest.* 



We put a number of yellow ants (Formica fiava), 

 with their eggs and cocoons, into a small glass frame, 

 more than half full of moist sand taken from their 

 native hill, and placed in a sloping position, in order 



