K.—BOTANY 199 
to longitudinal flexure of the door upon actuation. The upper part of 
the door is wider than the lower part, forming a large convex surface, 
easily flexed by contact through the circular port formed by the velum. 
When this has been caused by the movement of some animal, the pressure 
of water continues the flexure till it reaches the transverse bend near the 
door edge. The sudden reversal of this bend results, in my opinion, in 
the snap action, which has been observed, following my suggestions, by 
Mrs. Johnson. She assures me that such snap action occurs, the changes 
in contour of the lateral walls and the engulfing of air having been observed 
by her, as also by Mr. McIntyre. 
The door is peculiar in shape in having an extensive upper region, 
wider than in the lower region, and thin. Near its attachment to the wall 
of the trap, and where the maximum bending occurs when the door 
opens, the outer course of cells is thin. Elsewhere the two courses are 
approximately equal in thickness, except in the lateral reaches of the 
lower region, where these areas exert downward thrust on the middle 
piece. ‘The transverse curvature of the door in this region is not as sharp 
as in such forms as U. capensis, but this is compensated for by the trans- 
verse flexure of the door in bending down over the back part of the 
threshold. In order to understand how important this flexure is we must 
appreciate that it is not a straight flexure, as a bend in a piece of flat paper, 
but a curved one, like the bend between the rim and crown of a bowler 
hat. The overcoming of the thrusts afforded by this arrangement is the 
peculiar feature of the action of this type of door. The effectiveness of 
the watertightness in view of the extensive outwardly bowed upper region 
of the door is procured by the extensive velum, supplemented by the 
adjunctive or secondary velum arising from the trichomes on all sides of 
the entrance in front of the threshold. The whole arrangement must be 
of a high degree of efficiency, and the in-pull of water great, as the sides of 
the trap are thicker than elsewhere, whereas in the other types the walls 
are of equal thickness on the flanks, with the exception of Polypompholyx, 
as we shall shortly see. 
The actuation of this trap appears to be of the same manner as the 
previous forms described—namely, pressure on a thinner area of the 
door, occupied by a group of sessile glands on the surface and so placed 
as to fall nicely within the hole formed by the ring-shaped velum. This 
rather definitely points to the function of the sessile trichomes on the 
upper part of the door, no less for the previous groups of species described 
than for this—namely, as an area of contact by entering prey, possibly 
attracted thereto. 
POLYPOMPHOLYX. 
In spite of the presence of most of the usual structural stigmata (for 
I failed to find the velum), in my previous publication (1932) I leaned 
toward the view that the door of this plant (of which two species, much 
alike, exist) acts as a simple valve, in the sense meant by Darwin for 
Utricularia. At my request Mrs. Johnson was kind enough to make 
careful observation of the living traps. Her notes, accompanied by 
sketches, leave no room for doubt that snap action occurs. I have made 
