THE HEXAGONAL PRINCTPLE. 571 



to prove that each cell is forced into the hexagonal shape by the 

 pressure of six cells surrounding it, and the other that the cell 

 is made hexagonal by the cutting away of material from six 

 surrounding cells — they both agree in one point, namely, that 

 the normal shape of the cell is cylindrical, and that it only 

 assumes the hexagonal form by mechanical means. 



These questions were briefly mentioned, because an entire 

 omission of them would appear negligent, but they were not 

 followed up because the nests that would set them at rest be- 

 longed to another group. "We will first take the central nest. 



The specimen from which this was drawn was fortunately in 

 an unfinished state, only eight cells being made, and some of 

 these but partly finished. As the reader may see by reference 

 to the illustration, all the cells are hexagonal, whether finished 

 or incomplete, and moreover, that the edges of the hexagon arc 

 quite sharp and well defined. 



Now, if either of the two theories were true, the cells would 

 not have assumed this shape. Where are the six surrounding 

 cells that are needed to compress the outermost cell into an 

 hexagonal ? Or where are the six surrounding cells from which 

 the hexagon was excavated ? There are none. The outermost 

 cell, for example, is perfectly free on five of its sides, being only 

 attached to the neighbouring cell by the sixth side. Com- 

 pression, therefore, has not been employed, because there is 

 nothing that can compress it ; neither has excavation been used, 

 because there is no material to be excavated. No one, on look- 

 ing at this group of cells, can deny that the hexagonal form is 

 produced by the direct labours of the insect, and not by any 

 secondary mechanical means. 



Perhaps some one who has not examined the actual object 

 might say that the materials of which the cells are made are 

 sufficiently stiff to need no support of contiguous cells. Now 

 the substance of this remarkable nest is singularly slight, the 

 walls being not thicker than the paper on which this account is 

 printed, and the material is quite soft, as may be seen by the 

 curvature produced by the mere weight of the structure. Yet 

 none of the cells are united by more than three sides, the 

 greater number by two only, and the external cells merely by 

 a single side, leaving five sides and four angles perfectly free. 



In this particular specimen the material has evidently been 



