128 Mr. G. R. Waterhouse on the 
the insect working with the axis of its body in the same relative 
direction to the cells; say, first, in the direction d 2, and, secondly, 
in that of e 2; then two more flat sides to cell 3 may be pro- 
duced by simple repetition of similar work ; and, lastly, the side 
f 4& being now straight, furnishes a position through which the 
insect may operate upon the side which still remained untouched. 
Thus it is possible, by mere repetition of similar work under 
similar conditions, to alter a cell having a circular section into an 
hexagonal cell; the cell in question being only in contact with two 
other cells. 
In the foregoing it is assumed that the wasp is at work ona 
plane surface—on a piece of comb in which the openings of the 
cells are all on the same level, and it brings about conditions 
which verY NEARLY accord with those which actually present 
themselves. 
The surface of the forming comb, however, is not a plane one, 
but is more or less convex, and the young cell on the margin of 
the comb, as I have had to point out, opens out obliquely, dipping 
away from the level of the next inner series of cells; and it is 
obvious in this case that the stretch of the insect would be more 
or less restricted. ‘The insect, for example, which is at work in 
cell a, might reach to the right and to the left on the outer 
surface of cells 1 and 2, but it would only be to some point short 
of the centre of the exposed surface of cells 1 and 2, and some- 
thing shorter than the side of the hexagon ultimately attained. 
And if the wasp were to work (as it very often will do) on the 
outer surface of a cell having the position of cell 2, but without 
having the adjoining cells a and 6 in contact, it would have the 
power of flattening the once-rounded outer wall, from the entering 
angle on either side towards, but not quite to, the centre of the 
free margin, and there would, therefore, be no true angle formed 
at that part; and this corresponds exactly with the conditions 
ordinarily presented. 
I have now described the cells constructed by the wasp, and 
those of allied insects; and I have likewise described the cells of 
certain bees, and compared together the works of these two tribes 
of insects. In so doing I have endeavoured to set down the 
leading facts exactly as they present themselves, and not to be 
influenced by any theoretical views. Subsequently, I have 
endeavoured to harmonise those facts, and this part of the com- 
munication only, I think, can in any way be regarded as theo- 
retical. Here I may, and probably have, erred in endeavouring 
to explain too much; but setting aside certain minor points, I 
