20 A Monograph of the Erysiphaceae 



practice followed by many authors of determining a plant specific- 

 ally in its conidial {Oidiuvi) stage. As yet, we do not know enough 

 about the conidial form (which is possibly as variable in some 

 species as the perithecial is known to be) to be able to determine 

 it specifically with safety, and it is certain, that the practice of doing 

 so had led to many errors. I have, therefore, as far as it has been 

 possible, omitted in my host-index those records in which the 

 fungus has been observed only in the conidial condition. The 

 treatment which the fungus on CucurBitaceae {Qiamis, Ciiairbita) 

 has received, affords evidence of the danger of naming the conidial 

 forms. In herbaria and published records the fungus is stated 

 almost without exception, to be Sphacrothcca Castagnci. Schroe- 

 ter, however, having observed the perithecial stage on Cuairbita 

 Pcpo, has named it Erysiphe polygoni. In collections and cxsiccati 

 this fungus on Cuciirbitaceae exists apparently only in the conidial 

 stage, and the only specimen that I have seen with pcrithecia 

 occurred on Ciicurbita Pcpo, in late autumn, at Rcigate, Surrey, 

 England. On this specimen the fungus is undoubtedly Erysiphe 

 cichoracearum, the species recorded on Cuciirbitaceae by American 



mycologists. 



3. 



±Lrysiph 



always wrongly determined. . In Europe there is a species, not 

 uncommon on Valeriana officinalis which, without exception, has 

 been. referred to E. communis. {E. polygoni) by authors; it is, how- 

 ever, undoubtedly E. cichoracearnvu The fungus not uncommon 

 on Anchnsa officinalis and Echinm vulgare has been named E. 

 cichoracearum, and no records exist of E. polygoni occurring on 

 these hosts, yet all the specimens I have seen in herbaria are really 

 to be referred to the latter species. The only fungus on Galium 

 that I have seen is E. cichoracearum, yet it is always named in 



European herbaria 



is so recorded in all works. 



polygoni) 



I have thought it advisable, therefore, on account of the con- 

 fusion that exists in many cases, to compile a host-index from 

 .personal observations, chiefly with the view of seeing if any facts 

 of interest could be gained on the subject of the general relation 

 between parasite and host. By means of the abundant material 

 received from many parts of the world, and through the examina- 



