202 SECTIONAL ADDRESSES 
ments which take place during the setting of the trap after actuation is 
a question which may well be asked, but there is no sure answer ; yet it 
may be recalled that Brocher (1911) entertained some such idea in regard 
to U. vulgaris, but I think without objective justification. I think the 
bolster of tissue in such forms as U. vulgaris is capable of resisting changes 
of shape, but if this is the case it does not preclude the possibility from 
other forms. 
The histology of the door deserves remark. The inner course cells of 
the middle piece, or rather the region of the middle piece and the area 
above it, where the sessile glands on the outer face of the door are situated, 
are of long cells, running longitudinally to the edge of the door, as far 
as the beading. These cells are, moreover, very richly provided with 
folds and props, indicating great flexibility in precisely that region, the 
middle piece, where in other forms there is a tendency in the other 
direction, namely, of rigidity. This condition is to be met with also in 
the cells of the central hinge in U. vulgaris and similar forms, a point 
where great flexibility in various directions is present (Lloyd, 1932). 
In the globulariefolia type, the degree of complexity arising from folds or 
corrugations in all directions is very great indeed. The effectiveness of 
the application of the door edge on the threshold lies in part in the con- 
siderable curvature of the door edge, so that the line of contact is itself 
nearly the arc of a circle. 
With this type we conclude our remarks concerning these foregoing 
types which may be collectively subsumed under one generalisation— 
namely, that they are all forms in which the actuation of the trap is pro- 
cured by the inbending of a more flexible area of the door, thus upsetting 
the unstable equilibrium set up. by the physiological activity of the walls 
in expelling water, a state of reduced water pressure within being preserved 
by the somehow engaged door rendered watertight by a velum. 
In this type, however, we have departed from those in which there is 
a small angle to one in which that angle is quite wide, and the application 
of the door to the threshold is along its edge. In this, the trap is like that 
of U. vulgaris, but is otherwise quite of its own kind. Not in this type 
is the pose of the door surface against the threshold insured by the down- 
thrust of its lateral regions. This thrust is now oblique with relation to 
the threshold, pushing the door edge forward to engage the front of the 
middle zone. The rigidity of the door in its efficient closed position is 
assured by the slope of the sides of the threshold, which faces outward ; 
thus the threshold is a short half-funnel into which the door (being 
longer along its edge than the threshold) is cramped, so that the greater 
the water pressure the tighter the contact of the lateral regions of the door 
against the resisting threshold. 
We pass on to consider another series of types, contrasting with the 
foregoing by having a special tripping mechanism, consisting of long 
bristles or glandular hairs protruding from some point of the door 
surface and correlated with the structure of the door itself. We shall 
consider first a curious Asiatic species which appears to stand in a position 
intermediate between those which we have already considered and those 
which are distinctly typical, e.g. U. vulgaris. 
