316 
AMERICAN JOURNAL OF BOTANY 
[Vol. io, 
Epidermis. Because of differences in position, size, and form of the 
cells, only the outermost layer of cells has been designated here as epidermis. 
Great irregularity in form and distribution is exhibited by the cells of this 
layer. Many of them are tangentially stretched, or become so in due course 
of time. A considerable number, however, are conic in form, with the apex 
protruding into the cuticle. Tangential and radial measurements taken 
early and late may be considered as a fair index of the general change in 
form and size of the epidermal cells. The outer tangential wall is usually 
somewhat thickened, and the thickening may extend also to the fore part 
of the radial wall. No seasoned change in this respect, however, could 
be detected. 
Table 3. Measurement of Cells of Epidermis 
Ave. Diam., 
Tangential 
Ave. Diam., 
Radial 
Approx. Area 
Long. Sect. 
Increase 
No. of determinations. . . . 
42 
4 2 
August 20. 
13-5 M 
12.5 ix 
170 sq. [x 
September 17. 
15-5 /* 
12.5 ix 
194 sq. ix 
14-1% 
Two or more cells, often called ‘‘daughter” or “window” cells, may be 
grouped together to form a “mother” cell, as shown by McAlpine (13) and 
other investigators, thus making a comparatively large open space between 
adjacent groups. Early in the growth of the fruit the “window cells ” appear 
to be more irregular in outline and closer together. The average number of 
“window cells” per “mother cell” was found to have increased during the 
season as shown in table 4. Material collected on August 20 showed 
Table 4. Average Number of “Window Cells” per “ Mother Cell ” 
Date 
Aug. 20 Sept. 17 
No. of determinations.53 53 
Number. 2.92 4.00 
usually two “window cells” in each group, though some groups contained 
four, and a few one, three, or five cells. On September 17 the average 
number of “daughter cells” had increased to four, with a few groups con¬ 
taining two, six, and eight. Uneven numbers of cells—three, five, etc.—were 
now only rarely encountered. This numerical distribution may be con¬ 
sidered as a fair proof of the usual subdivision of the original cells during 
the time of enlargement of the fruit. Quite often single undivided cells 
were also present. 
Subepidermis: First layer. Though the tissue to a depth of five cells or 
more immediately below the epidermis may be designated as subepidermis 
(fig. 6), the first layer was considered separately. Cells of this layer differ 
