ASCOMYCETES 2 1 7 



Host plants. The various species and varieties of mildew have 

 been reported upon about fifteen hundred species of phanerogams. 

 The list of hosts includes plants of numerous orders and families. 

 A few notable exemptions among plants of normal terrestrial 

 habits are Liliaceae, Iridaceae, and some other monocotyledons. 

 Furthermore, there are many exceptions among such families, for 

 instance, those having the habit of growing under unusually moist 

 conditions. Moreover, herbs, shrubs, and trees are more or less 

 equally affected, and sometimes a single species of these mildews 

 may be found upon plants of all three sorts. 



The leaves are usually the chief parts affected, although some 

 species may attack also the twigs, stems, and fruits. As a rule, 

 those having the densest mycelia are more persistent and more 

 likely to infest all portions of the plant. The Erysiphaceae seldom 

 cause conspicuous distortions of the host plant. The anatomical 

 modifications are therefore secondary in interest to the physio- 

 logical effects. 



Cross inoculations. In a very recent summary of the general 

 results of cross inoculation in the mildews, Reed states : l 



One or more species of five of the genera of the Erysiphacese have been 

 tested for their capacity for infecting host plants other than the one from which 

 they came. Podosphaera is the only genus which thus far has not been tested. 

 With reference to four genera, Microsphaera, Sphaerotheca, Phyllactinia, and 

 Uncinula, the data are very meager. The bulk of the work has been done with 

 three species of Erysiphe, E, Cichoracearum, E. graminis, and E. Poly- 

 goni. Even with these species the number of trials is very small in many 

 cases, the evidence often resting on a single experiment. Still, sufficient data 

 have been accumulated to form the basis of certain at least tentative general 

 conclusions. 



So far as investigated, the mildew on the cucurbits, Erysiphe Cichorace- 

 arum D. C, is the only one which is shown to be capable of infesting plants 

 belonging to more than one genus. My results with this mildew are based on 

 a large number of trials, many of them repeated at different times during three 

 years, and cannot be questioned. 



There are other cases where the mildew is limited closely to plants of a single 

 genus. For example, the mildew on rye is limited to species of the genus Secale. 

 The same is true with reference to the bluegrass mildew on species of Poa. 



Several cases also are recorded where the mildew from one species will not 

 infect other species of the same genus. Most of these claims, however, rest on 



1 Reed, Geo. M. Infection Experiments with Erysiphe cichoracearum PC- 

 of Wisconsin Built. 350 : 340-416. 10,08, 



