K.—BOTANY 203 
THE Type U. orBICULATA (Figs. 15, 16, 18). 
If a bizarre structure warrants using this species as a type, the choice is 
justified. The plant was first described from the point of view of the 
morphologist by Goebel in 1891. I have been able to include Goebel’s 
Ceylon collection in my studies and I have regarded this material as the 
typical U. orbiculata Wall. It is curious that no other collections sent to 
me from India and Ceylon have yielded species with the characteristic 
tubers of the Goebel material, which he described. All the other 
species sent me are very closely akin to the one before us and, at all 
events, the traps are alike. The trap is pear-shaped in lateral view, with 
the stalk affixed just in front of the middle point of the ventral side. 
Above the oblique entrance there is a short rostrum, giving rise to two 
thick downwardly and outwardly curved antennz, bearing numerous 
uniseriate branches or cilia, each ending in a rather large, globose, 
glandular capital cell. The aspect of this apparatus distinguishes this 
species and its related ones indisputably. ‘The trap is small, scarcely 
ever longer than 1 mm., exclusive of the antennz. 
When the sagittal section of the entrance is examined, it resembles 
closely that of U. capensis, but on study certain important differences 
appear. The threshold, as in U. cerulea, is sharply uptilted near its 
inner border, and has an unusually deep sulcus running transversely ; in 
this the door edge rests along the edge of the middle piece, which is narrow. 
The door edges along the lateral reaches slope up on either side to the 
inner angles of the threshold. ‘There is a velum supplied by the cells of 
the outer zone of the threshold, this being the primary velum, homologous 
with that found in other types. In front of the threshold the forecourt 
leading thereto is lined, as in U. capensis, with a clothing of stalked 
trichomes, graded in length, the shorter the nearer the threshold, in such 
fashion as to continue the general surface of the latter, thus forming 
a narrow channel leading from the outside to the threshold. These 
trichomes produce long, expanded cuticles which form a secondary velum, 
supplementing the primary (Fig. 15). ‘The membranes are voluminous 
and entirely fill the space fronting the door as far as its middle point. 
We have seen that a secondary velum is thus produced in the U. monanthus 
type, except that in U. orbiculata there is no secondary velum above the 
entrance, but merely for the stretch of the three-quarters cylindrical 
forecourt. Only in U. monanthos, etc., the trichomes supplying the 
secondary velum are scarcely distinguishable from those of the threshold 
proper. 
But it is the door which furnishes the bizarre feature. In general 
proportions it resembles that of U. cerulea ; it is not long and narrow, 
hor is it semicircular, but something between these extremes (Fig. 18). 
As seen in the sagittal section, its posture is oblique with reference to 
the threshold, and correspondingly the lateral hinge regions are extensive 
and well developed, and exert a downward thrust which applies the 
middle piece tightly to the middle zone of the threshold. The middle 
piece is laterally not extensive and is composed of smaller cells than those 
of the hinge regions on either side ; nor is it sharply differentiated from 
