214 SECTIONAL ADDRESSES 
The upper moiety of the door, as seen in sagittal section, is uniform in 
thickness, with the outer course of cells thin, but somewhat thinner 
toward its lower limit, where it meets the middle piece. The character 
of its tissues marks the upper region as hinge. This hinge extends 
around the sides, forming much thicker lateral hinges. It is evidently 
convex (at least in the relaxed position), and is clothed rather densely with 
short, clavate trichomes. Just at the lower limit and close to the middle 
point (here the door is thin) there are two stiff bristles arising out of the 
outer course of cells. ‘These extend downward and forward. The 
middle piece is quite thin along its sagittal line, just below the upper part 
of the door, but on each side there is a large bulbous protuberance. The 
two masses become thinner toward the door edge, to which they extend, 
and appear to be stiffening agents, giving rigidity to the middle piece in 
one direction without limiting its flexibility in the other. They are 
derived entirely from the inner course of cells (Fig. 19, 5), merely by 
enlargement without additional cell-divisions, in the fashion in which 
the knob of the door of U. purpurea is formed (here from the outer cell 
course) (Fig. 25). It is evident that the thin line of door tissue between 
the bulbous masses is a longitudinal hinge. 
The threshold is like that of U. orbiculata, with a deep transverse 
depression which receives the middle piece edge. ‘The precise posture of 
the door when the trap is in the set position still eludes us, as the material 
could not be studied at the time of collection. It may be suspected that 
the posture represented in the diagram, while probably correct for the 
relaxed posture, should, for the set condition, be less convex, giving the 
middle piece a larger angle of contact with the threshold. We may be 
sure that this occurs, but precisely to what extent we do not know. In 
the relaxed position the tripping bristles appear to lie in the trichomatous 
clothing of the approach to the threshold. Less convexity of the door 
would result in bringing them up into a position which would seem to 
be a better one for their functioning. But in any event it seems fairly 
certain that their position is such that the prey should press down on 
them in approaching the door, their downward swing resulting in prying 
up the upper part of the longitudinal hinge. 
As in the case of U. orbiculata, U. Kirkit combines some of the mechanical 
features of the cornuta type with those of vulgaris. 
In view of the anatomical facts displayed, it would seem that actuation 
of the trap takes place as follows : Pressure on the two juxtaposed trichomes, 
inserted just above the two protuberances and so placed that impact will 
usually be from above, will push the upper part of the middle piece 
inwards along the middle thin line between the protuberances. This 
initial flexure allows the water pressure to act in the usual manner. The 
flexure travels upward along the middle of the upper region, where the 
door is relatively thin, and backward to the door edge. 
THE Type U. PURPUREA (Figs. 25, 26). 
We have come to the last type to be considered, wholly American and 
chiefly confined to the tropics, or at least to South America. North 
America has the one species of which I have been able to study living 
