INFECTION EXPERIMENTS WITH ERYSIPHE 
GRAMINIS DC. 
GEORGE M. REED. 
Assistant in Botany, University of Wisconsin. 
Neger is to be credited with the discovery that among the mil¬ 
dews, as in the rusts, our ordinary morphological species may 
consist of a number of physiologically specialized forms which 
are limited in their occurrence to a, single host plant or to a 
group of closely related host plants. Neger’s (8) first experi¬ 
ments were made during the summer and fall of 1901. His gen¬ 
eral conclusion was that, in the mildews he studied, specializa¬ 
tion has gone to such an extent that conidia from one species 
will not infect a species of any other genus. In some cases the 
specialization has gone still further so that conidia from one 
species is incapable of infecting another species of the same 
genus. His results may be summarized as follows: 
1. Conidia of Erysiphe cichoracearum DC. from Artemisia 
vulgaris will infect A. vulgaris but not A. Absinthium, Alchem- 
illa vulgaris, Galium silvaticum, G. rotundifolia, Hicracium 
murorum, Lactuca muralis, Lcontondon taraxacum, Lithosper- 
mum arvense, Plantago lanceolata, Ranunculus repens, Senecio 
vulgaris, nor Sonchus oleraceus. 
Conidia from Lactuca muralis will infect L. muralis but not 
Galium silvaticum, Hieraceum murorum., nor Pulmonaria offici¬ 
nalis. 
Conidia from Hieracium murorum will infect IL murorum but 
not Artemisia vulgaris, Galium silvaticum, Hypericum monia- 
num, Lactuca muralis, Leoniodon taraxacum nor Sonchus olera¬ 
ceus. 
Conidia from Senecio vulgaris will infect S. vulgaris but not 
