INFECTION EXPERIMENTS WITH THE MILDEW ON 
CUCURBITS, ERYSIPHE CICHORACEARUM DC. 
GEORGE M. REED. 
All the recent work on the so-called physiological species of 
the mildews has tended to show that each genus and, in many 
cases, each species of host plant has. its own particular special¬ 
ized form. So far as investigated, the mildew from one host 
plant is not able to infect a species belonging to another genus. 
Marchal (15), it is true, asserts that the same special form oc¬ 
curs on both oats (Avena sativa ) and Arrhenatherum elatius. 
Salmon’s (24) results, however, contradict those of Marchal 
on this point. 
Heger (20) has further found that conidia of Erysiphe cich- 
oracearum DC. taken from Artemisia vulgaris will not infect 
A. absinthium, although both are reported as host plants of 
this species of mildew. A similar result has been obtained by 
Salmon (24) with the mildew of the clover, Erysiphe polyg¬ 
ons DC. He found that conidia taken from Trifolium pra- 
tense will not infect any other species of this genus, although 
plants of this species are readily infected. My work (22) with 
the grass mildew, E. graminis DC., upon rye (Secale cereale ) 
and blue grass (Poa pratensis) in general confirm these results. 
The degree of the specialization of Erysiphe graminis DC. 
within the genus Hordeum , has also been investigated by Sal¬ 
mon (26). He found (1) that conidia from barley (Hordeum 
vulgare) readily infected barley, H. distichum , H. zeocriton, 
H. intermedium and H. hexastichum; (2) that conidia from 
