10 
herd’s-purse seedlings in the spring of 1887, and it produced a large 
crop of spores which did not spread the trouble rapidly to other plants 
as in an ordinary season. In June there was very little of the mildew 
and only a few spots could be seen on the seedlings in the autumn. 
Only a small amount of the fungus was found the next spring. The 
mildew was fairly abundant upon the common pepper grass ( Lepidium 
Virginicum ), but soon disappeared. Three years ago in a moist sea¬ 
son the blossoms, flower stalks, and seed vessels of the garden radish 
were generally attacked and often distorted beyond recognition. Very 
late in October, 1887, the mildew was found prevalent upon seedling 
plants of Sisymbrium officinale, which had developed after the Sep¬ 
tember rains in a shady place. Last year there was very little of the 
mildew upon the Shepherd’s-purse. The two dry years had done much 
toward eradicating it along with the host. 
Cystopus cubicus, Ley., is the least common species of the genus, 
and for the last three years has been found upon Ambrosia artemisicefo- 
lia , (7., only at rare intervals. 
Cystopus bliti, Lev., occurs upon an increasing list of hosts. 
Amarantus bUtoides , was badly infested in all three seasons, and as 
this species flourishes in dry places, upon paths, and roadways it 
at first seems an exception to the rule. However, this Amaranth is 
thick-leaved and succulent, and like the purslane is full of moisture, 
even though the surroundings are arid. It was a fact of observation 
that the greatest development of the mildew was upon plants which 
were most favored with moisture and shade. 
Cystopus portulac^e, Ley., is the last species of the list and one 
of no little interest in the present connection. It seems to have been 
more abundant during the two dry seasons than before or since. The 
host is a juicy plant and the appearance of greater thrift on the part of 
•1 
the parasite may be due to the fact that it was more than usually de¬ 
ductive, the purslane not being able to withstand its attacks. There¬ 
fore a patch of the host while showing more of destructive effects of 
the mildew might in fact have less of the parasite. It is impossible 
to say there was more or less of the mildew the last year than the sea¬ 
son preceding. These observed facts seem to show that with the Per- 
onosporece there is no doubt that the species are best suited to a moist 
season. The members of tlie genus Peronospora have in no instance 
been so abundant during the two dry years as before or since the 
drought. In general the mildews were found in early spring in 1887, 
after this, through the long dry summer, in limited quantities, upon 
plants growing in moist places along streams and edges of pools. In 
1888 the greatest abundance was in June after the May rains. 
The genus Cystopus seems less influenced by drought, but as a rule 
the infected specimens were those best situated for obtaining moisture. 
In all cases where Peronosporece flourished in drought, they were upon 
succulent hosts, and even with these there was probably less growth of 
