18 A Monograph of the Erysiphaceae 



Gattung Phoma, sind im Folgenden je nachdem sie auf Holzge- 

 wachsen, krautartigen Dicotyledonen, Monocotyledonen, oder 

 Kryptogamen wachsen, und in jeder Abtlieilung nach den alpha- 

 betisch geordneten Nahrpflanzen angeordnet/' When in place of 

 such an arrangement as this, where the connection between host 

 and parasite is considered of primary specific importance, and 

 where a mere alphabetical list is substituted for a natural classifica- 

 tion, a monograph of the genus Plioma appears in which the species 

 are defined by morphological characters, a more scientific method 

 in the classification of fungi will have commenced. 



After the forms of the Erysiphaceae have been grouped around 

 what may be termed *' morphological centers," the next most im- 

 portant work is to experiment biologically with them. Such ex- 

 perimental work remains, unfortunately, almost entirely to be 

 done. A few instances are on record of attempts to sow the 

 spores of one species on the host-plants peculiar to another species, 

 e, g,y the spores of E. graminis on the host-plants of E, polygojii^ 

 and those of E. polygon on the vine, the host-plant of Uncinula 

 necator ; in all cases without success. 



An experiment performed by Magnus (mentioned in detail under 

 Sphae rathe cci), although unfortunately incomplete, is of great inter- 

 est. Conidia of Sphaerotheca hnimili growing on the hop, were 

 taken and sown on Taraxactun officinale, a common host-plant of S. 

 liiumili var. fuliginea. This variety (considered a distinct spe- 

 cies by Burrill and most American authors) is distinguished 

 by the much larger cells' of the perithecium, usually dif- 

 ferent character of the appendages, etc. The conidia of S, 

 Jiitmiili germinated on the Taraxaciun^ and produced a mycelium 

 bearing conidia. Observations were not continued to see if peri- 

 thecia would be produced. From this experiment we are led to 

 one of three conclusions. We must suppose either : (i) That the 

 conidia of S. hiumtli on this host would have produced ultimately 

 the perithecia at present considered characteristic of the var. 

 fidigineay in which case we should have to consider the latter as 

 merely a form of S, humidi induced by growing on certain host- 

 plants. This view is rather favored by the fact that in some 

 species we have certain forms associated with certain hosts which 

 present slight morphological differences, although these are not 



