X036 
Journal of Agricultural Research 
Vol. V, No. 22 
As soon as penetration of the stoma was gained by the germ tube, 
a marked change was noted to take place in the character of the fungous 
growth produced, as indicated by different staining qualities. The co- 
nidium and the slender germ tube external to the spore opening stained 
lightly, while the cells in the pore opening or beneath the stoma stained 
much more deeply and were comparatively large and round (PI. LXXXI, 
A , B y F ). It was only rarely observed that penetration into two different 
stomata took place by germ tubes from one conidium (PL LXXXI, 
By b). In the case observed, the two stomata were near each other and 
slight germ-tube growth was sufficient for the penetration of both. As a 
rule, however, only one germinating tube from a conidium has been found 
to penetrate the host tissue, although it is known that, if this tube does 
npt penetrate before its desiccation takes place, another cell of the 
conidium may germinate later before the entire conidium loses its via¬ 
bility and penetration might again be possible. At times the pore wall 
of a guard cell may be penetrated and the growth gradually spread to the 
adjoining epidermal cells (PI. LXXXI, F, c ). Normally, however, the 
germ tubes grow through the pore opening, probably receiving some 
stimulus from the guard cells and form round, heavily staining mycelial 
cells which pile up directly in the air chamber below the pore opening. 
The fungus then grows toward the parenchyma cells (PI. LXXXI, C, d) 
and flatten out against their walls, probably for nutritive purposes. 
At times, without further development within the host, the fungus grows 
back out through the stoma and produces conidiophores (PI. LXXXI, 
Dy e). In such a case new conidia might be produced before an extensive 
area of the host tissue had been killed. Usually, however, the fungus 
grows farther into the host before conidia are formed. It probably is 
true, as first suggested by Uzel (io), that the fungus causes asphyxiation 
and consequent collapse of the parenchyma cells, since only a slight 
intercellular growth of the fungus occurs. An attempt by the host cells 
to isolate the invading organism is seen in the massing of heavily staining 
substances (PL LXXXI E, /) in the parenchyma cells which adjoin the 
air chamber. Under certain conditions this isolation probably is accom¬ 
plished and the host cells then remain turgid and normal. Where this 
can not be done, the cells surrounding the fungous mycelium collapse 
(Pl. LXXXI, G) t the mycelium gradually produces tufts of conid¬ 
iophores, and the characteristic leafspot is formed. The host under 
normal growth conditions is able to isolate this infected area, though as 
a result of severe, abundant infections, entire leaves may be covered 
with the conidiophore tufts of the fungus. 
It then appears that there is no attractive force existing between the 
closed stomata and the conidial germ tubes of the fungus, and also that 
the latter do not possess enzymic power to directly penetrate the epi¬ 
dermal cells. However, with open stomata germ-tube penetration may 
occur, even though some length must be attained before the tube can 
