334 Copeland. — The Mechanism of Stomata. 
cell. Such folds occur only where the cells are in con- 
tact with stomata, and more often from the sides of the 
latter. The folds frequently leave what appear to be inter- 
cellular spaces next the guard-cells or in the end of the fold : 
I have been unable to determine the extent of these spaces 
in cross-section, even with serial sections cut in paraffine, but 
they certainly do not pass the external wall. In cross-section 
(Fig. 2) the stomata are conspicuously shallow. The walls are 
equally thin throughout, except for a narrow ridge of entrance 
and a place which cannot nearly always be detected, represent- 
ing the ridge of exit. In very thin sections it appears that 
the lumen projects somewhat toward the ridge of entrance. 
To study the changes in guard-cells with variations in the 
width of the pore, I cut surface sections of the leaf with the 
razor. Such sections keep the stomata subject to any tensions 
the underlying mesophyll may exert upon it, directly or 
indirectly ; and spare them the violence inflicted by stripping 
the epidermis off alone. The leaf of Medeola is so spongy 
throughout and the stomata are so large that a little less 
accuracy in measurement is the only objection to studying 
them in place. Their appearance and behaviour are exactly 
the same in the sections. This is not necessarily true of all 
stomata, but it helps me to believe that the objections of 
Benecke and Kohl to work done with sections are not prac- 
tically important. At any rate, on plants where it cannot be 
determined that there is or is not a difference it is because 
they cannot be studied in place, so that resort to sections, 
when one would measure all dimensions, is compulsory. The 
sections were made from leaves in bright diffuse light. They 
were mounted in water, drawn very carefully with a camera 
lucida, closed by running glycerine under the cover, and 
drawn again : the drawings were measured. All measure- 
ments are stated in mikrons. Even approximately accurate 
measurements of depth, by focussing with the micrometer 
screw adjustment as is feasible with many stomata, are im- 
possible with Medeola for want of marks to focus on. And 
I was unable to make satisfactory use of Schellenbergs 
