124. Mr. G. R. Waterhouse on the 
have but one plate; each of such cells being exactly opposed to 
one other cell on the opposite side of the comb.* 
and it is only at the mouth of the cell that the seven sides appear. 5-sided cells 
are common, this being the usual form of the cells of the first-formed tier in the 
hive, cells which are on one side in contact with the roof of the hive, and are 
surrounded at other parts by four cells. 
* In the British Museum is a nest of one of the slender-bodied wasps which 
is well worth examination in connexion with any theory intended to explain the 
form of the cells. The nest in question has its upper part dome-shaped, and is 
built round a stout stick. The under part is closed in by a thin covering, the 
outer surface of which is gently convex (excepting at one part, where a small 
area is depressed and slightly concave). It furnishes an extensive surface, being 
between seven and eight inches in diameter. A small opening is left in this 
lower envelope of the nest for the ingress and egress of the wasps. Nearly the 
whole of its surface is covered by a network of rudimentary cells, the average 
depth of the cells being about the eighth of an inch ; those on the concave part, 
however, are much deeper. The cells covering the chief part of this area are 
truly hexagonal, and the bottoms of the cells are flat or very nearly so. Notwith- 
standing the beautiful regularity which prevails in this work, there are, in certain 
parts, some remarkable aberrations from the normal conditions. The most ir- 
regularly formed cells are those on the margins of the nest-covering. Here the 
lower edges of the dome-shaped upper and outer covering of the nest descend and 
hang down like a curtain, with an average width of about half an inch, which 
serves to protect the comb being constructed on the under covering of the nest, 
and which in time, had the structure proceeded, would itself have been inclosed. 
The cells which come in contact with this curtain are so built that the curtain 
forms their outer boundary, and most of them are pentagonal, but with very 
unequal sides, the partitions which separate them from each other being often 
twice, and sometimes three times as long as those which separate the marginal 
from the sub-marginal cells. These cells, moreover, are frequently confluent, 
some of the partitions which should have separated them not being carried up, 
but sketched out, as it were, by an indistinct ridge; in short, so variable are 
these marginal cells, that one may safely say there are no two alike. In other 
parts are some singular modifications in the forms of the cells. I will notice 
one particular cell and a few others that are near toit. This cell has six sides ; 
two longer sides which meet as nearly as possible at a right angle, and one 
very short side. From this short side extend two sides of an adjoining cell, 
and these form part of the boundary of a cell with five unequal sides, two of 
which sides do not meet so as to form a true angle, for the angle is rounded (I so 
express it for brevity sake). From this rounded angle runs out one of those rudi- 
mentary partitions which merely sketch out the boundaries of what I have termed 
confluent cells. Again, joining the cell first noticed, and also adjoining each 
other, are two cells which are very nearly square, each having four long sides. 
and one extremely short one; these again join a cell which has seven unequal 
sides. I will only further add, with respect to this nest that there is a con- 
siderable area round the thick stick which passes through the nest on which cells 
have not been commenced, and that nearly the whole of the cells which abut 
upon this area (and they are numerous) have their free margin rounded; in one 
or two only is the outer margin rendered angular. 
