[Vor. 3 
50 ANNALS OF THE MISSOURI BOTANICAL GARDEN 
Melhus ('11) showed that chilling the conidia of Cystopus 
candidus to a temperature of 8-10°C. produced the optimum 
germination and also infection in the case of the radish, and 
thus proved that this fungus is dependent on such chilling for 
its best development in its attack on the host. 
Similar relations have been found to hold true in the cases 
of other Phycomycetes. From field observations covering a 
period of twenty years Lutman (711) concluded that Phytoph- 
thora infestans required a fall in temperature for its best de- 
velopment on the potato. Melhus (212) showed that the opti- 
mum temperature for spore germination in this species, both 
conidia and zoospores, was 8-14°C., thus corroborating the 
previous field observations. After the fungus has entered, 
however, he noted in a later paper (18) that the disease was 
produced more readily at comparatively high temperatures, 
thus showing that, in this ease at least, there was a difference 
between the temperatures favorable to parasitie growth, de- 
pending upon whether the infecting material be spores or 
mycelium. 
Reed (712), working with Phytophthor a infestans on tomato 
plants, found that here again its attack was dependent upon 
low temperatures. The attack only occurred above the alti- 
tude of 2000 feet, and then only at times when there were cool 
nights. 
Opposed to Phytophthora, Plasmopara Viticola has been 
found to be dependent on a rather high degree of temperature. 
Sajó (’01) observed this in 1900 as compared with 1899, the 
temperature in 1900 being higher throughout the summer than 
in the previous year. Again, in 1912, Ravaz and Verge (712, 
124 2120) showed that for quickest germination a temperature 
of 22-27? C. was necessary for this fungus and that, since 
the fungus found water also necessary for infection, the host 
eould only be attaeked at periods of sustained high tempera- 
tures and humidity. Istvánffi and Pálinkás (713) showed 
that not only were Ravaz and Verge correct, but that the de- 
velopment of conidiophores and conidia from the infected host 
was also somewhat dependent on these same temperatures. 
