364 
AMERICAN JOURNAL OF BOTANY 
[Vol. 9, 
at C. By cutting oblique sections of somewhat larger stems, rows of 
terminal buffer cells can be found, raising the cuticle for some distances. 
In still older stems which are provided with a cork layer, the pseudo- 
parenchymatous primordium arises from 
hyphae mostly confined to the phellogen 
layer (PI. XXII, fig. 4). The cork is 
raised and br'oken by the buffer tissue in 
the manner which I have described (/.c.) 
for the effuse trunk forms of Gymnospo- 
rangium nidus -avis. 
However limited may be the attack by 
the sporophytic stage of this species on 
Juniperus, it should be noted that it is 
quite the contrary case with the gameto- 
phytic mycelium in Crataegus. Here the 
hyphae penetrate deeply into the cortex 
down to the cambium, which loses its iden- 
tity and ceases to function in an orderly 
fashion so that stereome and sieve tubes 
are often entirely wanting on the side of 
the stem invaded. The outer margin of 
the wood ring is irregular and indefinite. Haustoria are found in pith 
rays and in parenchyma cells surrounded by xylem. There seems to be 
no special tendency to obtain food from cells of the epidermis or phello- 
gen, although in material cut late in the season haustoria are found in nearly 
every living cell. They are long, irregular, and hypha-like, branching 
and coiling about in the cell, quite unlike the trim, constantly binucleated 
haustoria of the sporophytic stage. 
In cedar leaves or young stems sporophytic hyphae surround epidermal 
cells, attacking the cuticularized layers, and sending haustoria even into 
the guard cells, and it is interesting to find them in cells whose walls have 
become somewhat suberized. Substances related to those from which cutin 
is finally formed naturally occur more abundantly in these regions. In 
older stems provided with a cork tissue, the cells of a phellogen are certainly 
rich in substances undergoing a series of transformations in the layers of 
their walls such that by their final elaboration they may be deposited as 
suberin. These facts noted above might be taken as further evidence that 
the transition substances entering into cutin and suberin are closely allied. 
It is not strange, then, that this species is very limited in its attack on 
the host and that the first sori, i.e., those on leaves and young stems, are 
subcuticular. A fundamental morphological feature such as the formation 
of buffer cells from the terminal cells in the sorus primordium is one fixed 
in the course of evolution and not to be easily affected by environmental 
conditions. Further cytological study will undoubtedly result in the dis- 
FiG. 7. Buffer cells in a sorus 
on a young stem at the margin of 
a leaf. At A and B, various stages 
in buffer cell formation. Telio- 
spores arise from subterminal cells. 
At C, a young spore arising from the 
subterminal cell, above which can 
still be seen the buffer cell. 
