Formation of the Cells of Bees and Wasps. 127 
If cell a had been previously commenced, we can see that the 
work afterwards performed in cell b might be restricted on the 
side next that cell; but on the opposite side, which is not so re- 
stricted, the work is not laterally extended beyond the same point; 
and I infer from this, that when the insect is in the position men- 
tioned (and there is no disturbing cause*), its power of stretch 
controls the diameter of the cell, as appeared to be the case with 
the Osmia. That from a certain point, which shall be, we will say, 
the centre of cell 6, the stretch being in all directions equal from 
the centre, must describe a circle, and at the same time it must be 
equal, in each direction from the centre, to one side of a hexagon, 
the angles of which would touch that circle. 
Cell-building consists of a constant repetition of similar work, 
and of similar parts of the work. Let us see what would be the 
effect of such repetition upon the cell 3. Cells 1, 2 and 3 are 
partly built and are marginal cells; cells 4.and 5 are within these, 
and are more advanced. The next operation, according to the 
usual course, would be to commence the building of cells a and b, 
the first atom of material laid down for these cells being in the 
angle between cells 2 and 1, or 2and 3. The position of the insect 
in building cell a I suppose to be in the direction of the partition 
between cells 1 and 2 and across cell 5; that for building cell b 
would be in the direction of the partition between cells 2 and 3 
and across cell 4, Cell 2, then, is enclosed at all parts, and has 
assumed a truly hexagonal form, and to complete this hexagon the 
insect has worked alternately in cell 2 and in cells a and 6, But 
a repetition of part of the work would have produced the same 
effect on cell 2. The axis of the insect being in the direction a 5, 
and suitable for building cell a, the insect might work alternately 
inside and outside of a part of cell 2, and thus produce the 
straight partition between cell a and cell 2. Again, the position 
of the insect being altered, as it would be, to build cell b, that is 
in the direction b 4, by the same operation the partition between 
cells 6 and 2 would become straightened. At the same time the 
same position enables the insect to work alternately likewise in 
cells 3 and 6, and to produce a flat partition between them. The 
inner half of the circumference of cell 3 has been converted into 
three sides of a hexagon; externally it furnishes two entering 
angles d and e, Let the work be repeated as before described, 
* Under different conditions to those above indicated—as, for instance, in the 
construction of the nest-covering—the insect evidently does not keep its body 
fixed in position while at work. Here the work is performed with shifting centres, 
and segments of very large circles are exhibited in all parts of the work. 
