K.—BOTANY 213 
evidence, would certainly obtrude itself. The two kinds of traps may be 
found on the same stolon, close together. 
In the one form the door resembles that of the reniformis type, save 
that the middle piece is not so highly specialised. ‘There is, however, 
but a single tripping bristle, straight, tapering and pointed. The basal 
cell and the next cell to it form a special stiff, hinge-like base. ‘There are 
a few short, clavate trichomes on the upper part of the door surface. 
Over the general surface, both on the ramp and elsewhere, are scattered 
globular, sessile trichomes. In the other form there is no tripping bristle 
at all. The door carries numerous long clavate trichomes, and similar 
still longer ones are found along the ridge of the ramp and under the 
antennz, which here resemble those of the globulariefolia type. ‘Thus one 
form of trap resembles reniformis and the other globulariefolia. Neither 
Merl nor I have been able to correlate surely the distribution of the two 
kinds of traps with position on the plant. 
The second species of the American triad, U. nana, also resembles 
reniformis, but is lacking antennz, and for this reason recalls cornuta. 
We need not particularise further than merely to point out that, instead 
of there being only one tripping bristle, as in one form of trap in 
U. Lloydit, or four, as in reniformis, there are just two, standing side by 
side. Their basal cells are somewhat enlarged, but show no striking 
form such as noted in U. Lioydit. The structure of the door is, however, 
precisely like that in U. Lloydii. 
The third species, U. longiciliata, was examined by Merl, who observed 
(1915) that there arises from the middle of the door, just above a massive 
middle piece, a single glandular trichome, consisting of a slender, cylin- 
drical basal cell, a short mid-cell, and an ovate capital cell. In view of 
the massive character of the middle piece and the smallness of the 
trichome, it is not easy to believe that the door is actuated by contact 
with the trichome. The form of this trichome allies the trap with that of 
our last type, U. purpurea. The likeness is perhaps rendered still more 
striking by the fact that there is below the entrance a strong protuberance 
bearing two widely spreading branches. A single protuberance is to be 
seen in some of the allies of U. purpurea, though none in others. There is, 
moreover, also a small rostrum above the entrance, somewhat as in an 
Old World species described above (my No. 131, unnamed). Though 
the general type of trap, aside from the curious appendages, allies 
U. longiciliata with the vulgaris type, I am inclined to look upon it as 
indicating an alliance with U. purpurea on account of the door trichome. 
Finally, here we must place a most intriguing type which I collected 
in Africa at the Victoria Falls, namely, U. Kirkit (Figs. 17, 19), identified 
for me by Dr. Rendle ; and in this connection | recall the kindness of 
Dr. Saunders, who supplied me with vials for my collection, having been the 
more provident. U. Kirkii is a small, blue-flowered plant of terrestrial 
habit, growing in ground where there was abundant seepage, with a trap 
looking to be almost identical with that of U. capensis, but that the rows 
of the more slender peristomal trichomes below the entrance are raised 
somewhat on a collar, as in U. albina. Like capensis, too, it has a broad 
threshold, the lateral thrust of the door being downwards. 
