126 Mr. G. R. Waterhouse on the 
angle to angle, and I assume, therefore, that they are cells having 
the full diameter permitted by the stretch of the insect, as it 
ordinarily works in cell-building ; such cells would merely touch 
each other were they cylindrical (whereas, in the smaller cells, the 
natural outlines would intersect), and there would be in their 
interstices a certain amount of superfluous material, which the 
instinctive propensities of the insects would lead them to remove ; 
for, as I have before pointed out, their cells are carried up from 
the original foundation by adding material to their margins, and 
this is followed by excavation on the inner side. 
I have now to speak of those cells on the margin of the wasp 
comb which are more or less angular on the outer side, and yet, 
at that part, are not in contact with other cells. 
It has been stated that after the marginal cells have been carried 
up to a small extent the next work of the wasp is to commence 
other cells in the angles between each pair of these cells; thus 
converting the cells which were marginal into inner cells, and that 
then the sub-marginal cells become truly angular and hexagonal. 
Let 1, 2 and 3 (PI. XIII. fig. 4) be three partially constructed 
marginal cells; and let a and b be two cells commenced in the 
angles between them; it is clear that the work performed on the 
cells 1, 2 and 8 can no longer belong entirely to those cells,—they 
being continued by the insect almost simultaneously with the cells 
a and b. So much work as is bestowed upon cell 2 on its outer 
side must equally form part of the walls of cells a and b; and the 
work of cells a, 6 must in part belong to cells 1, 2 and 3. As these 
cells (a and b) moreover increase in size, part of the work in each 
must be common to the two, viz., at the point d, where they will 
become united. The force, then (I will so put it), that is thrown, 
for instance, on that part of cell a that comes in contact with cell 
2, is met by another force in cell 2 at that part in which it is in 
contact with cell a; and the two forces are equal. If the cells 
2, a and Db are to be carried up together, there is no reason why 
any one should encroach on any other, but the contrary. 
When the insect is at work with its head in cell }, there must 
be a certain time during which its work is bestowed on the part 
which is common to cell b and cell 2, and at another time it will 
be at work upon the part which is common to cell 6 and cell 3. 
Tt works, then, to the left and to the right equally, and we must 
suppose that the axis of its body is in the direction of the partition 
which separates cells 2 and 3; and, taking the size of the wasp 
and of its cell into consideration, we may add that its abdomen 
would be extended across another cell (4) behind these. 
