TENOPYR: CONSTANCY OF CELL SHAPE 61 
I found that, if the sections of the lower epidermis to be meas- 
ured were prepared by stripping this tissue away from the leaf, 
it was difficult to determine which axis of the cell had been parallel 
to the midrib. The possibility of error in this respect was greatest 
in cells having irregular outlines. I therefore prepared these sec- 
tions by removing the upper epidermis and the green tissues from 
that portion of the leaf which was to be examined, leaving the 
lower epidermis intact. The entire leaf, or the exposed epidermis 
with a portion of the midrib, was then mounted in water and 
placed under the microscope. Before measuring the palisade 
cells, both the upper and the lower epidermis of the region to be 
examined were stripped off, care being taken to remove none of 
the green tissue. When the leaf was placed flat on a slide the 
remaining tissues were sufficiently transparent to enable me to 
make measurements of the two dimensions of the upper ends of the 
palisade cells. All the measurements were made on the living 
cells, thus avoiding the possibility of shrinkage or distortion due 
to fixation. 
I found an ocular micrometer preferable to a stage micrometer, 
as, by moving the slide or by rotating the micrometer, I could 
more easily bring the specimen to be examined in the proper posi- 
tion in relation to the scale, without handling the specimen itself. 
Throughout the investigations I used a Leitz ocular microm- 
eter. The scale consists of a large square, divided into four 
smaller squares, one of which is again divided into twenty-five 
equal squares. The value of these divisions of course depends 
upon the magnification, and it was ascertained with the aid of a 
stage micrometer for each eye-piece and objective used. For 
example, when a one-inch ocular and two-thirds objective were 
used with a Bausch and Lomb microscope, each side of any one 
of the smallest squares was equal to 0.116 millimeter. All cell 
counts on all the plants examined are stated in terms of cell 
diameters per millimeter. 
The section to be measured was placed on the stage of the 
microscope in such a position that the midrib of the leaf was paral- 
lel to one side of the ocular micrometer scale. 
In measuring the palisade cells, a favorable area of this tissue, 
uncrushed and free from veins, was chosen, equal to one of the 
* 
