28 Mitchell.—Contributions towards a Knowledge of the 
thickened surrounding cells. They are easily ruptured on the drying of the 
sporangium. The basal portion is much more thickened than the flaps, the 
cells at the edge of the valve being specially large and thick (PI. Ill, Fig. i). 
Connecting the lower parts of the valves, and visible only as a deep furrow 
in the closed sporangium, is a definite layer of thin-walled tissue, easily 
seen on pulling the valves apart (PI. IV, Fig. 4), and, as will be noted later, 
subserving a definite function. This tissue also extends between the upper 
flaps, but is not so evident in that situation. The cells in the basal region 
are not isodiametric, their long axes being orientated in different directions 
(PI. Ill, Fig. 4), those near the edge of the valve being approximately parallel 
to that edge, whilst those between the cracks are exceptionally thick, and 
specially noticeable in travelling across in a directly transverse direction. 
Above this region, and in the upper flap, occurs the hollow prepared for 
the reception of one of the upper spores, and the cells round this become 
smaller and thinner walled, as they approach the edge of the flap. In this 
region the cells are arranged in rows approximately parallel to the margin 
of the flap. Transverse and longitudinal sections show (PL III, Figs. 1 and 2) 
that the structure of the active thick-walled cells is precisely the same 
as that of the cells of the fern annulus, i. e. the outer walls are thin, the 
inner ones strongly thickened, the thickening narrowing in the radial direc¬ 
tion towards the outer margin ; whilst the cells connecting the basal parts 
of the sporangiuum are thin, shallow, and entirely unthickened. The smaller 
cells figured at 1 c give additional flexibility to the sporangium wall. 
The orientation of the cells in the basal portion is of importance. 
On examining the two valves of one sporangium it is evident, as Stein¬ 
brinck has already pointed out, that the transverse band of cells between 
the cracks is found only on one valve, viz. the lower one. This acts as 
a hinge on which the flap of the lower valve is turned back. The upper 
valve cannot move any considerable distance on account of the rigid axis of 
the cone ; the lower valve, on the contrary, by its vigorous movement, 
presses back the sporophyll (which is narrowed at the base to allow of this), 
and thus gives greater space for the shooting out of the spores. The 
relative rigidity of the upper valve is necessary in order to give the lower 
valve full play. 
When the sporangium opens apically, the flaps diverge, but at the same 
time continually change their shape. The edges of the flap curl over and turn 
back, thus widening the space between the valves and creating more space for 
the distribution of the spores. The changes take place gradually, so that 
when the flaps reach a stationary condition the upper spores are lightly poised 
on a flat or even sometimes convex surface (i. e. concave outwardly) and 
readily respond to any movement of the lower part of the sporangium. 
(The fact that the two upper spores are sometimes shed first, and that 
the flaps are frequently convex instead of concave after ejection, led me at 
