142 
Journal of Agricultural Research 
Vol. XXIII, No. 3 
the usual septum appears some distance back, forming the cell that is 
to give rise to the haustorium. 
With the entrance of the fungus into the host cell, however, making 
the first actual contact between the living protoplasm of host and para¬ 
site, abnormal changes begin. Plate 4, C, shows the first recoil from 
the contact. The haustorium-producing cell a is partially collapsed and 
has shrunken away from the host cell. There is a small red-stained spot 
on the host cell wall at the point of entry of the fungus. The scattered 
watery contents of the haustorium mother cell are in sharp contrast to 
the normal haustorium-producing cell. The latter at this stage (compare 
PI. 2, A, at d and PI. 2, G) would be turgid, closely appressed to the host 
cell, and the remainder of its protoplasm would form a dense layer close 
to the point of exit into the haustorium. The appearance here suggests 
that some substance diffuses out from the host cell, disorganizing the 
haustorium-producing cell. The harmful effect does not stop here, for 
the cytoplasm of the hypha below the mother cell has drawn back sharply 
from the septum, suggesting retreat from the advance of some plasmolyz- 
ing substance. It is not probable that this plasmolysis is due to fixation 
for, as pointed out earlier, the osmotic pressure in fungous tissues is high 
and they are rarely shrunken by ordinary fixing fluids. 
Changes in the host cell are equally marked. The first change is an 
increase in turgor. By a fortunate mistake, one of the fixing fluids used 
proved too strong for the younger seedlings of Kanred. Preparations 
from it show the healthy mesophyll cells plasmolyzed, while the cells 
attacked by the fungus remain turgid, thus indicating an altered chemi¬ 
cal condition in the host cell induced by the fungus. This excess of 
turgor in the attacked cell is rapidly succeeded by a collapse, as seen in 
Plate 4, C, and the protoplast may contract into separate masses, each 
of which may include plastids. Even the cell wall shrinks irregularly, 
leaving points on the outline of the cell. The nuclear contents are dis¬ 
solved, although the nuclear membrane is distinct. Only in the imme¬ 
diate vicinity of the fungus are the plastids altered. There they appear 
as an indistinct mass. The adjoining host cells are perfectly normal in 
appearance. 
A later stage in this reaction is seen in Plate 4, D, showing a longitudinal 
section through the epidermis and stoma. A portion of the collapsed 
appressorium is seen on the outer surface at a and the substomatal vesicle 
just inside the guard cell at b. Here, as before, only a single mesophyll 
cell at e was attacked. The disordered contents of this cell, the absence 
of the nucleus, the clumped, misshapen plastids, and the rough jagged 
points on the cell left by the irregularly collapsing wall give ample proof 
of the attack. It is to be noted in passing that in this case the collapse 
of the wall is sharply localized at e, and the end of the cell farthest from 
the fungus is relatively smooth in outline. The remaining stub of the 
infecting hypha, with its ragged discolored tip at d, lies at some distance 
from the host cell attacked. It may be that here, too, as in the earlier 
stage just described, the fungous cytoplasm retreated down the hypha to 
its base. Here a septum was laid in at c } walling it off. The relative 
abundance of cytoplasm and nuclei in the substomatal vesicle supports 
this view. The haustorium and the cell producing it and the portion of 
the hypha intervening between it and the stub have disappeared com¬ 
pletely. 
