450 
AMERICAN JOURNAL OF BOTANY 
[Vol. 9. 
different host plants, among which the common garden pea, Pisum sativum, 
is one of the best known. This powdery mildew often bears the name of 
"mildew of the pea." When the seedling plants of the Oenothera crosses 
were grown in the greenhouse it was found that the pea mildew did not 
infect them. The seedhng plants were placed in two separate greenhouses, 
in each about 1,500 plants. In one of the greenhouses there were many 
pots of garden peas covered with powdery mildew {Erysiphe Polygoni DC). 
Whenever these plants were moved or shaken, small clouds of spores arose 
from leaves and stems. None of the Oenothera seedlings standing in the 
same house showed infection at the time of planting in the field about three 
months after being placed in the greenhouse. In the other greenhouse a 
few small plants of Oenothera nutans Atkinson & Bartlett, fall seedlings, 
which were kept during the winter and which were abundantly infected by 
mildew, were standing near the young seedlings. Before these seedlings 
were set out in the field, all the susceptible plants among them showed 
infection. A general exposure to conidia of Erysiphe Polygoni from pea 
did not infect any Oenothera seedlings, while a much less general exposure to 
conidia of Erysiphe Polygoni from Oenothera nutans resulted in the infection 
of a large number. This seems to support the statements of Salmon that, 
although no distinct morphological differences are found among the strains 
of Erysiphe Polygoni from different host plants, physiological differences 
may exist, upon which are based the so-called biologic strains." Searle 
(11) also proved the existence of biologic strains of Erysiphe Polygoni 
among various hosts. 
It was not considered necessary to make any extensive trials at cultivat- 
ing the Erysiphe of Oenothera on artificial media, since the powdery mildews 
in general have been amply proved to be obligate parasites by Salmon and 
others. Salmon (9) found in his experiments that he could grow powdery 
mildew (oidium) on leaves of Euonymus japonicus L. placed on moist filter 
paper in a damp chamber for as long as 14 days, in which time the leaves 
were badly affected. In his experiments with biologic strains, Salmon 
kept his strains growing on living plants. 
Examinations were made of prepared slides of leaves from the five 
different species and strains chosen, in order to determine if any morpho- 
logical differences might account for the differences in susceptibility and 
immunity. No such differences were found among the morphological 
characters of the leaves, suggesting in this case that immunity must have 
a physiological or chemical basis. Salmon (8) reached the conclusion 
'Uhat susceptibility and immunity were due to constitutional (physiological) 
peculiarities and not to any structural ones.'' 
Description of the Mildew, Erysiphe Polygoni 
A study of the Oenothera mildew showed that in general it conforms to 
the description of Erysiphe Polygoni given by Salmon (7). It is amphig- 
