HIVES FOR BEES. 
35 
cluster and work in approximately globular forms, 
rectangular hives are unsuitable for them. The in- 
ference has not much justification if the question be 
properly analysed. It is true that the globe has a 
theoretical advantage, since it is the solid which has 
the smallest amount of surface in proportion to its 
content. Thus, one cubic foot of bees in globular form 
would have 4^ square feet surface ; in a Stewarton 
hive (see “Stewarton 5-I square feet surface; and 
in a cubical box (the cube being the rectangular solid 
with the least surface in proportion to its solidity), 6 
square feet surface. But these differences are really 
insignificant when compared with the benefit deriv- 
able from utilising all the bees ; e.g.^ if a swarm be 
placed in the rectangular hive (B, Fig. 5), wax- 
workers are able to continue their operations into the 
very corners, if the walls are only sufficiently non- 
conductive. The disadvantage here, then, does not 
arise from the angles, but from excess of space, which 
is so great that one side of the colony is exposed, 
and many bees {wh) are kept in enforced idleness ; 
while (introducing the principle of elasticity), if the 
movable side, or dummy (^), had been brought up to 
the line a b, the external idlers would have been set 
free, and converted at once into gatherers and builders, 
the combs, as a result, growing much more quickly. 
The dummy here is capable of absolutely adjusting 
the hive to the size of the swarm, be the latter 
large or small ; so that, in every case, nearly all the 
bees may be actively utilised in adding to the general 
wealth. 
The position chosen for exit and entrance by all 
D 2 
