184 A Monograph of the Erysiphaceae 



alien haben die Conidien die (im Profil elliptische) Gestalt einer 

 schmalen an den Enden abgerundeten Tonne." 



Winter, Schroeter, and others, have followed De Bary in main- 



'llifi 



P 



shape of its conidia. I have, however, found generally in study- 

 ing specimens of Oidiuui in the fresh state that the shape (and size) 

 of the conidium is subject to so much variation that I am strongly 

 inclined to doubt the advisability of employing such characters for 

 systematic purposes, and especially of considering them of specific 

 importance. In the present case, although only examination of 

 living material of the forms on the different host-plants can satis- 

 factorily setde the whole question, it appears to me very doubtful 

 from the study of herbarium material, if the cylindrical form of 

 conidium is exclusively confined to examples on Umbellifcrae and 

 whether conidia of both shapes do not occur in the same speci- 

 men. Certainly the conidia of E. polygoni on some host-plants, 

 e.g.. Clematis alpina, are more or less cylindrical. No differences 

 are to be found in the peritheclal form of fruit of examples on 

 Umbellifcrae to those of E. polygoni on other hosts. 



In dealing with the variation in mycelial characters we come 

 to two very interesting forms, viz, E. liriodendri Schwein and E. 

 densa Berk. 



E. liriodendri is maintained as a species by Burrill (60), who 

 remarks that it " may be identified by its abundant, white myce- 

 lium, especially on the young stems." 



I have examined many specimens of this form on Liriodendron, 

 and have found no characters except the persistent mycelium by 

 which it can be separated from typical E. polygoni. When occur- 

 ring on the stems, E. liriodendri has a thin persistent myce- 

 lium, but it is to be noted that on the leaves the mycelium is sub- 

 evanescent, and then the fungus cannot possibly be distinguished 

 from many of the forms of E. polygoni on other hosts. Nor can 

 the. persistent mycelium on the stem be considered absolutely pe- 

 culiar to the form on Liriodendron ; in the herbarium of the Up- 

 sala Museum there is a specimen of E. polygoni on the stems of 

 Thalictruni aquilcgifolium in which the mycelium is persistent, and 

 in some places almost pannose in consistency. This specimen 



