348 C. DA WSON AND S. A. WOODHEAD [November 



formed in any case, but their size is undoubtedly regulated by the 

 thickness of the plate of wax, the rule being the thinner the plate 

 the more minute the individual hexagon. The same result may be 

 obtained on a much smaller scale so as to produce only one or two 

 hexagonal forms, but the operator will then find that the difficulty 

 lies in the rapid cooling at the sides of so small a mass of wax. 



The explanation of the formation of these hexagonal bodies is as 

 follows : — 



On cooling, the wax at first forms into nuclei of nearly equal size. 

 On the shrinking of the wax by further cooling, these nuclei or 

 spheroids are pressed together, forming planes at their points of 

 contact. Should the wax be rapidly chilled before these spheroidal 

 bodies are formed to their full extent, they are then prevented from 

 coming into contact one with another by the intervening nebulous 

 masses of " uncentralised " particles of wax. It would appear by 

 microscopic examination that these particles are' also smaller nuclei 

 which become absorbed in the larger. They also, like the larger, 

 assume hexagonal form. In this state the nuclei appear when cold 

 as solid circular bodies. 



The hexagons appear very distinctly above and below the surface 

 while the wax is cooling. "When it is actually solid, their forms are 

 often very indistinctly seen, or may be altogether invisible, but they 

 are none the less present. The bases of these hexagons, which lie 

 mid-way between those visible at the top and those at the bottom, 

 are pointed and are arranged so that the point of the base of the 

 upper hexagon coincides with the points of contact of the lower 

 hexagons as in the honeycomb. These bases can be observed by 

 making a very thin microscopic section, but several hundred sections 

 had to be examined before they were made out with certainty. 



When a small amount of resin and turpentine is added to bees- 

 wax and melted, and the mixture is allowed to get cold, the outlines 

 of the planes of contact on the hexagons are more distinct and are 

 to be seen raised upon the surface. Under these circumstances they 

 may be easily rubbed with black lead, which still further increases 

 their visibility. 



Our chief experiment was next to put our theory to a practical 

 test, and observe in what manner the bees would deal with a cast 

 sheet of pure beeswax, which, when viewed by a side light, distinctly 

 showed traces of these natural hexagons over its surface. 



Before introducing it to the bees, we had traced upon it with 

 vermilion a group of the hexagons which appeared near the centre of 

 the plate. (Another group we black-leaded.) This was then photo- 

 graphed, after which the wax plate was placed in an observatory-hive 

 on a bar-frame. The bees soon started upon it, proceeding to excavate 

 round hollows in the centres of the hexagons, at the edges of the 

 plate, pushing out on all sides the iWbris around the edge of each 



