Oct., 1922] KLAPHAAK AND BARTLETT RESISTANCE TO MILDEW 45 1 
enous; mycelium very variable, persistent to evanescent, thin and effused; 
perithecia few and scattered, 85-95^1 in diameter, cells distinct, ii-i4)U 
wide; appendages simple and long, variable in number (3-6), partly inter- 
woven with the mycelium and colorless; asci few (3-5), small and ovate, 
50-60 fi X 30-35 M, containing 3-6 spores, 20-23 M X 9-12 Conidia 
(oidium stage) cylindrical to ellipsoid, 33-36 ji X 15-18 m- The mildew 
grows very superficially, feeding by means of haustoria extending into the 
epidermal cells. 
Powdery mildew may infect the host plant at any time. Infection 
in Oenothera is generally first noticed in leaves approaching maturity. 
Neither very young nor very old leaves will show any infection when a 
healthy plant is first attacked. At a later stage in the growth the mildew 
may cover the entire plant. Infections were found as early as April in the 
greenhouse, and in the fields as soon as the plants were set out. The 
heaviest infection in the field is commonly found in the summer from June 
to September, depending upon weather conditions, rainfall being conducive 
to the spreading of the disease. It is often found that during the summer 
time susceptible plants are entirely covered with mildew, so as to appear 
whitish. No evidence has been found that mildew does any great damage 
to Oenothera plants, as it does to the pea. Though entirely covered by 
Erysiphe, infected Oenotheras appear to grow normally, to come to bloom 
and to ripen seeds in exactly the same way, and presumably in the same 
amount, as healthy plants. Even though highly susceptible, species of 
Oenothera seem to be very tolerant to the disease. 
In our experiments the plants have been under observation during the 
whole season, and have been classified as immune to powdery mildew if 
they have shown no infection at any time. It may be objected that in 
some cases immunity may have been only apparent and due to a position 
in the field preventing infection. This objection is easily answered. The 
plants were set out in the field in rows of from 150 to 160 each. Of each 
of the parent strains chosen, 25 plants were grown to maturity, and of the 
hybrids about 100 plants. Cultures differing in their susceptibility to 
mildew were grown near together, so that in many cases an immune strain 
or species was grown among highly susceptible strains, often so as to be 
entirely surrounded by them, and with intertwining branches. When a 
form remains free of mildew under such favorable conditions for infection, 
it may be called immune, especially when the disease spreads as easily as 
in the case of powdery mildews. 
A second objection to our experimental procedure has been based on 
the supposition that somewhere there might exist strains of Erysiphe 
Polygoni which would infect the so-called ''irnmune" strains of Oenothera. 
This is quite possible but hardly concerns us, since we have not been in- 
terested particularly in the production of disease-free Oenothera strains, 
but rather in the fact that immunity to certain strains of Erysiphe exists, 
