K.—BOTANY 207 
Tue Type U. vuccaris (Figs. 21-24). 
Of the members of the genus, no species has been more under 
examination than U. vulgaris. F. Cohn and Darwin were the leading 
students of an earlier day (previous to 1882): it was then accepted that 
the door was a simple, inwardly moveable valve, which the prey opened 
easily by pushing against it. Its recurrence to its original position pre- 
vented escape. We need not recount at length the views that the bladders 
were floats, about which there was a lot of discussion finally closed by 
Goebel (1889). It was not till Brocher (1910) made the important 
observations that the bladders engulf air when a plant is raised from the 
water, and was led to see that only when the trap is set, that is, when it 
is in a condition of unstable equilibrium brought about by the exhaustion 
of the water content, that it can do so, he appreciated that the trap is 
watertight, but thought it merely plugged with mucilage, and that the 
trap could act but once. Merl (1921) found that the action of entrapping 
prey could be repeated, and determined the time necessary for the renewal 
of that condition of unstable equilibrium, namely, 15 to 30 minutes— 
observations which were made also by Czaja at nearly the same time. 
It was not quite certain to Merl that the whole action (aside from the 
exhaustion of the water from the interior of the trap) is purely mechanical, 
but Czaja took this position definitely. Recently M. Kruck (1931) has 
resuscitated the view, never very firmly held, that the action of the door is 
a sequel of the transmission therethrough of the stimulus from the protuber- 
ant, stiff,‘ irritable’ hairs : but this has nothing to support it. It isa curious 
fact that none of the above-mentioned observers, nor any others, had 
observed accurately the position of the door and its mode of contact with 
the threshold, nor had anyone save Withycombe suspected the inadequacy 
of Brocher’s idea of the way in which the watertightness of the door is 
procured. For my part, I have shown that this is due to the presence of the 
velum, that the contact of the door and threshold is a delicate adjustment 
involving a tripping mechanism, and that the whole action is mechanical 
and depends for its efficiency on the physical properties and adjustments 
of the various parts. What these are may now be briefly summarised. 
The door may be a continuation of the upper wall above the entrance 
(Fig. 24), or it may arise from a projecting overhang, e.g. U. gibba 
(Fig. 21). This produces no observable difference in the sensitivity of 
the mechanism. In any event, the door is very delicately constructed, in 
some species surprisingly thin, e.g. my No. 27 from Tropical Africa— 
a U. gibba-like form. The shape is nearly that of a quarter-spherical 
surface, one edge being attached to the walls of the trap, the other con- 
stituting the free edge of the door, the convex surface being turned out- 
wardly. A wide outer zone is flexed in front (that is to say, in the middle 
third of the door), so that it is here concave and may be regarded as 
a hinge, or, at all events, a region where the maximum bending can occur, 
as when the door is opened. ‘The outer course of cells is very thin ; the 
inner, thick and richly provided with transverse corrugations supported 
on the ends of props (observed by Meierhofer, 1902) in the radial walls 
(Lloyd, 1932). These cells are elongated radially in the door and, in 
