Insectivorous Plants (Part //). 405 
colour, thus contrasting with the surrounding epidermal cells. 
In all, but specially in the goblet-cell, large refractive 
globules are present. On vertical section, it is evident that 
the cuticle, present as a protective layer over the ordinary 
epidermal cells, thins out round, and is absent from, the 
surface gland-cell, which, having only a thin cellulose wall, 
is free to discharge its secretion. At the upper arching part 
of the tube are the skylights, the nature of which has been 
fully explained by Zipperer \ There are no hairs over the 
exterior. The outer surface of the involute edge of the 
pitcher is richly studded with honey-glands, which are more 
complex than those just described, each being formed of 2-4 
surface cells, and 4-6 times as many compose the entire gland 
(Plate XIX, Figs. 3, 4). They are also sunk in depressions 
of the epidermis, and are slightly covered by beautiful down- 
growing tooth-like processes of it, so that the whole greatly 
resembles the involute rim of Nepenthes . The inner surface 
of the involute edge is quite smooth to the depth of a quarter- 
inch, and destitute alike of hairs and glands, but below this 
is a narrow cincture of long, stiff, outstanding hairs, different 
from all other hairs of the pitcher. 
As noted by Asa Gray, the bilobed flap secretes honey 
copiously on its inner surface from glands which are one-, 
rarely two-celled on surface view. But further at the free 
extremity hairs begin to appear, at first as minute protu- 
berances of the epidermis, but towards the orifice these have 
so increased that there is a perfect forest of short, strong hairs 
directed towards the orifice. 
The internal arched c attractive ’ surface has long, stout, 
striated, downward-directed hairs, interspersed with two-, 
rarely one-celled glands (on surface view). There is an 
evident line of demarcation between the c attractive ’ and 
‘ conducting ’ surfaces, the latter consisting of cells each 
prolonged into a fine, sharp, striated hair. The junction of 
conducting and detentive surfaces is not clearly demarcated, 
the short hairs of the former being for some distance 
1 Op. cit. p. 25. 
E e 2 
