Copeland . — The Mechanism of Stomata. 337 
leaving the movement to be performed by the ventral and 
outer walls. The closure is incomplete, and is effected about 
equally by the central part of the ventral wall and by the 
ridge of entrance. I measured the pore from ridge to ridge. 
These measurements were made on a rather small stoma. 
Open. 
Closed, 
Length of stoma . 
45 
45 
Width of stoma . 
35 
35 
Width of pore 
5 
3 
Depth of guard-cell 
13*7 
8.4 
The chief peculiarity of this stoma, as is well known 1 , 
consists in the partial resorption of the wall between the 
guard-cells, leaving them in open contact at both ends of 
the rift. As Haberlandt pointed out (1. c. p. 465 ), and direct 
observation shows, this does not interfere with the mechanism 
of the stoma. As practically all stomata are symmetrical 
both shut and open, the variations in turgescence must be 
the same in both guard-cells. So far as movements are 
regulated by variations in the turgescence of the guard-cells 
alone, it can therefore make absolutely no difference in their 
movements whether or not the guard-cells communicate. 
It is surprising that the relation between the shape of the 
stoma in surface view and the mechanism should hitherto 
have been overlooked. In those that we have been con- 
sidering, the width of the guard-cells was necessarily in 
considerable excess over the depth. The length was limited 
because any possible variation in volume at the ends is wasted, 
and so interferes with the sensitiveness of the stoma. For 
these reasons, stomata with this mechanism are conspicuously 
broad : the limit to the ratio of breadth to length is fixed 
by the danger that the bulging of the outer and inner walls 
should be interfered with by the end walls. There are a vast 
number of stomata so shaped that the depth increases with 
1 W. Ph. Schimper, Recherches anatomiques et morphologiques sur les mousses, 
Strassburg, 1848; Haberlandt, 1 . c., 18S6. Similar stomata occur on vaiious 
other mosses. 
