K.—BOTANY 187 
6. Appendages.—The trap is usually provided with appendages, though 
there are a few exceptions, e.g. U. cornuta (Fig. 1), U. nana. In the 
submersed, floating, or semi-terrestrial forms resembling U. vulgaris, 
there are two antennz which are more cr less fringed with long bristles, 
and similar bristles radiate from the sides and top edge of the entrance, 
but these may be almost entirely absent. Or, instead of antennz, there 
may be two (U. Jateriflora, Fig. 13) to about seventeen (U. Kirkit) rows of 
bristling trichomes, forming a funnel-shaped lead to the entrance, which 
may also be provided with a proboscis projecting from the upper 
(U. albina) or from the lower lip of the entrance (U. longiciliata) ; or the 
latter may occur in the absence of any other appendages (U. elephas). 
On the surface of it, these appendages have been interpreted as guides to 
the entrance, but in many instances it is difficult to regard them as of 
any importance whatever. For our present purpose these may be dis- 
regarded, except as they may have some bearing on the character of the 
entrance. 
In order to avoid confusion arising out of the great mass of material, 
I shall choose types of various groups for special consideration. 
Tue Type U. cornuta (Figs. 1-8). 
I begin with U. cornuta because it is the simplest in structure, though 
not by this token a primitive form. Schimper published in 1882 a descrip- 
tion which was wrong in most particulars, so far as the structure of the 
door and threshold are concerned. My own account (1931), brief and 
very incomplete as it was, was offered before I had had the advantage of 
studying living material, and it is wrong in regarding the relaxed position 
of the door in the set position. I can now make amends for my 
inaccuracies. 
The trap in lateral view is rounded with a protruding beak beneath 
which is the entrance, and a stalk. ‘The large traps are about 1 mm. in 
diameter (Fig. 1). Viewed on edge, the sides are seen to be concave 
when the trap is in the set condition, and convex after actuation. Owing 
to the approximation of the stalk and beak, an edge view, showing the stalk, 
does not display the entrance proper, but this I have shown in another 
figure (Fig. 2). In this we see that the entrance is bounded by a lower 
lip in the form of an inverted arch, not quite circular. Above the edge of 
the lip we see some irregular cells projecting radially. ‘These are the 
forward courses of cells of the pavement epithelium. Hanging down- 
ward from the beak and extending inward is the door. This figure is 
a thick section, beginning at f, and extending as far back as, say, d, Fig. 3. 
By consulting this latter figure one sees that the arched entrance leads to 
a curved platform (approximately semi-cylindrical) lined with the pave- 
ment epithelium, made up of glandular cells of elongate form, being the 
capital cells of a closely packed tissue of glandular trichomes. Schimper 
recognised these trichomes as being of structure similar to the glands of 
the outer surface, as also did Goebel, Hovelacque and others for other 
Species. But, hitherto, students have supposed that the pavement 
epithelium is uniform in structure, which indeed, when regarded super- 
ficially, it seems to be. If we dissect away the adjacent parts, so_as to 
