Copeland, . — The Mechanism of Stomata . 335 
method, using thick cross-sections. It is very easy to make 
certain that there is a decrease in depth when the pore closes, 
but not to measure it. The dimensions determined were : — 
I. 
II. 
Open. 
Closed. 
Open. 
Closed. 
Length of stoma . . . 
. . 62 
62 
59 
59 
Width of stoma . . . 
. . 46 
46 
47*5 
47-5 
Width of guard-cell . . 
. . 20 
22 + 
21 
23 + 
Width of pore .... 
. . 6 
1-5 
54- 
1 
Width of outer vestibule 
. . 11 
10 
The pore of most of the stomata does not quite close : so 
transpiration through the stoma must take place at night. 
Still it is a motile stoma the width of whose pore varies more 
than 4 / x. The length and width of the entire stoma do not 
change. But the closure of the pore is effected entirely by 
a flattening out of the guard-cells. In this process the outer 
vestibule is narrowed ; and, as may be seen on sections viewed 
from within the leaf, the lines representing ridges of exit 
move nearer together. The whole process when the pore 
opens is therefore a becoming round of each guard-cell, in 
which all parts of the wall except the very narrow back seem 
to be equally concerned. 
Since the pore opens at most only as much as the guard- 
cells become narrower — the dorsal wall being subject not to 
a push but to a pull from the guard-cell— it is essential that 
this wall be quite rigid. Though the walls are thin, this is 
accomplished by the outgrowths already described, which are 
therefore a most important feature of the apparatus. The 
rigidity of the ends is insured in part by the wall between 
the guard-cells, and by the walls between the epidermal cells, 
which strike the guard-cells irregularly, and in part by folds of 
the wall, which, however, are of much less frequent occurrence 
here than on the dorsal walls. The rigidity of the ends 
seems more essential to stomata of some other mechanical 
types than to these, and will be found provided for in various 
ways. But reinforcement of the dorsal wall is a want peculiar 
