522 
AMERICAN JOURNAL OF BOTANY 
[Vol. 9. 
Figure 69, Plate XXXII, represents a longitudinal tangential section of 
a very young carpophore in which there is no annular gill cavity, but the 
palisaded cells are arranged quite obviously into at least four distinct 
groups with three faintly stained trama areas (A) between them which are 
continuous with the stipe and pileus rudiments. As we shall see later, 
where the tips of the oriented palisaded cells of each pocket meet, a gill 
cavity will be formed. The strictly median longitudinal sections of the 
young carpophores which lie in the plane of the future gill cavity are, it is 
obvious, less favorable for the correct understanding of the gill origin than 
longitudinal tangential sections. Figure 70 represents such a longitudinal 
median section of the same button shown in figure 69; the group of cells 
faintly stained at the right is a section of the area of ce Is which form the 
boundary between two adjacent clusters of oriented palisaded hyphal cells. 
The cluster of deeply stained cells to the left represents a section through 
the palisade cells. Where the tips of the cells meet we have a radial view 
of the beginning of the gill cavity. It is already clear in this carpophore 
that the palisaded cells are well established, and yet there is no evidence of 
an annular gill cavity. 
Figure 71, Plate XXXIII, represents a longitudinal median section of a 
young carpophore a little older than the one previously described. Here, 
as in figure 70, we find the two characteristic patches, of palisade cells to the 
right and left, slightly above the center of the carpophore, toward the 
margins of the young pileus. There is still no indication of an annular gill 
cavity. As our sections become more tangential (fig. 72) we notice the 
grouping of the palisaded cells (at A) and also the more parallel hyphae 
which lie between the groups of oriented cells and will form the trama. 
In a section more eccentric shown in figure 73, there are four distinct groups 
of palisade cells and three pronounced tramal regions between them (at C). 
Within each group of palisaded cells there are already faint indications of 
gill cavities which cannot merge into one annular gill cavity unless the 
tramal cell connections are torn. The primordium of the hymenium is 
oldest nearest the stipe. Toward the margin it becomes progressively 
younger. Figures 71, 72, and 73 show a point of interest in that the layer of 
terminal cells seems to form a distinct covering and appears to correspond 
to what Brefeld called a general veil. This layer has been torn away 
from the pilear region in the process of sectioning. 
A somewhat older carpophore is shown in figure 74. The section is 
longitudinal and not quite median, through two groups of palisade cells 
with their small gill cavities in the earliest stages of appearance. More 
tangential sections of this carpophore (fig. 75) show four distinct groups of 
oriented palisaded cells (at A) with their rudimentary gill cavities and 
the intervening tramal cells. A distinct and conspicuous gill cavity is 
visible in this figure in the second group of palisade cells from the left side. 
In a section of the same plant still more tangentially placed and shown in 
