6o 
Journal of Agricultural Research 
Vol. V. No. a 
r 61 e in bridging over the winter. This paper gives the results of experi¬ 
ments and observations which show that in the Northern States species 
of the Peronosporaceae which have perennial mycelium are common 
and that the mycelium may live from one growing season to another in 
the living diseased host tissues. 
In several of these experiments the locality where infected plants were 
growing was marked in the autumn and the plants collected from time 
to time during the winter and early spring, after which they were allowed 
to revive in the greenhouse and a careful watch kept for any evidence of 
fruit of the fungus. In other cases the underground parts of infected 
plants were taken in the spring and planted in steam-sterilized soil in the 
greenhouse, and when the shoots came through the ground conditions 
were made favorable for the sporulation of the fungus. In still other 
cases the presence of the mycelium in perennial parts of the host was 
determined microscopically. 
PERONOSPORA PARASITICA 
Tate in the fall of 1910 and 1911 it was observed that young plants of 
Lepidium virginicum in the vicinity of Madison, Wis., were very generally 
infected with Peronospora parasitica and that the tissues of these plants 
contained few or no oospores, although they were produced in abundance 
in the summer when the host tissues were dying. Plants of Lepidium sp. 
always form a rosette of leaves in the late fall, and some of these remain 
alive through the winter. 
In the fall of 1911 two patches of Lepidium plants, about 50 per cent 
of which were infected with Peronospora parasitica f were marked so that 
they might be easily found during the winter. One was on the side of a 
short incline made by dumping several loads of soil in a heap and the 
other on the parking of a city drive in Madison. Both patches were 
well exposed during the winter of 1911-12, which was unusually severe, 
there being no covering of snow on the former at any time and the 
latter being covered only a part of the time. 
After the first killing frost, which, according to the Weather Bureau, 
occurred on October 24, infected plants of Lepidium virginicum were 
collected at various times during the winter. Beginning on October 30, a 
test was made of the germination of the conidia of Peronospora parasitica 
growing on Lepidium virginicum . Although when alive the conidia of this 
fungus usually germinate profusely within 2 to 3 hours and always within 24 
hours, no germination occurred in this test, although exposed to favorable 
conditions for 48 hours. This coincides with what is known of the behavior 
of the spores of other species—e. g., Cystopus candidus (Melhus, 10)—and 
excludes the possibility of these conidia becoming a source of further 
infection. A careful search for oospores was made after October 30 in a 
large number of infected plantlets, but none was found. 
